CONFERENCE

PROGRAM

2015 International Conference on

Advanced Mechatronic Systems

August 22-24, 2015, Beijing, China
2015 ICAMechS
International Conference on Advanced Mechatronic Systems

August 22-24, 2015

Beijing, China

PROGRAM

Organizers:
International Journal of Advanced Mechatronic Systems
Beijing University of Posts and Telecommunications, China
Tokyo University of Agriculture and Technology, Tokyo, Japan
IEEE Systems, Man, and Cybernetics Society

Sponsors:
Zhongyuan University of Technology, China
Agricultural Robotics and Automation Technical Committee, IEEE Robotics & Automation Society
The Institute of Complex Medical Engineering
International Journal of Modelling, Identification and Control
International Journal of Innovative Computing, Information and Control

Cooperation with:
The Society of Instrument and Control Engineers
The Institute of Systems, Control and Information Engineers
Group C of The Institute of Electrical Engineers of Japan
The Japan Society of Mechanical Engineers
Organizing Committee (1)
General Chairs: Mingcong Deng, Tokyo University of Agriculture and Technology, Japan

Xiaoguang Zhou, Beijing University of Posts and Telecommunications, China

Hongnian Yu, Bournemouth University, UK

Mengchu Zhou, New Jersey Institute of Technology, USA

Shinji Wakui, Tokyo University of Agriculture and Technology, Japan

Program Chairs: Yixin Yin , University of Science and Technology Beijing,China

Xinkai Chen, Shibaura Institute of Technology, Japan

Luige Vladareanu, Romanian Academy, Romania

Dongyun Wang, Zhongyuan University of Technology, China

Local Arrangement Chairs: Shengjun Wen, Zhongyuan University of Technology, China

Dianwei Qian, North China Electric Power University, China

Special Session Chairs: Z. Q. Lang, University of Sheffield, UK

Jose Ragot, Nancy-Universite, France

F. Smarandache, University of New Mexico, USA

Shiro Masuda, Tokyo Metropolitan University, Japan

J. X. Zhang, Swinburne University of Technology, Australia

Toru Yamamoto, Hiroshima University, Japan

S.G. Ponnambalam, Monash University, Malaysia

B. Bandyopadhyay, Indian Institute of Technology Bombay, India

Yan Shi, Tokai University, Japan

Publication Chairs: Changan Jiang, Ritsumeikan University, Japan

Hossam A. Gabbar, University of Ontario Institute of Technology, Canada

David Raisz, Budapest University of Technology and Economics, Hungary

Yoshihiro Ohnishi, Ehime University, Japan

Publicity & Exhibition Chairs: Shuoyu Wang, Kochi University of Technology, Japan

Giancarlo S. Gajani, Politecnico di Milano, Italy

Manabu Yamada, Nagoya Institute of Technology, Japan

Tomohiro Henmi, Kagawa National College of Technology, Japan

Organizing Committee (2)
Registration Chairs: Shuhui Bi, Shandong Academy of Science, China

Zijiang Yang, Ibaraki University, Japan

Jinglong Wu, Okayama University, Japan

Students Activity Chairs: Yegui Xiao, Prefectural Univ. of Hiroshima, Japan

Yang Shi, University of Victoria, Canada

Takao Sato, University of Hyogo, Japan

Best Paper Award Committee Hansheng Wu, Prefectural University of Hiroshima , Japan
Chairs:
Grigore Gogu, Institut Franais de Mcanique Avance , France

Zabih Ghassemlooy, Northumbria University, UK

Advisory Chairs: Hiroyuki Ohno, Tokyo University of Agriculture and Technology, Japan

Tianyou Chai, Northeastern University, China

Xiaoping Liu, Beijing University of Posts and Telecommunications, China

Akira Inoue, Okayama University, Japan

Tsu-Tian Lee, Chung Yuan Christian University, Taiwan

Honorary General Chairs: Hong Wang, University of Manchester, UK

Chun-Yi Su, Concordia University, Canada

Ken Nagasaka, Tokyo University of Agriculture and Technology, Japan

Ikuro Mizumoto, Kumamoto University, Japan

Zongxiao Yang, Henan University of Science and Technology, China

Quanmin Zhu, University of the West of England, UK

Secretary: Shin Wakitani, Tokyo University of Agriculture and Technology, Japan

Aihui Wang, Zhongyuan University of Technology, China

Program Committee (1)
Chairs: Yixin Yin, University of Science and Technology Beijing, China

Xinkai Chen, Shibaura Institute of Technology, Japan

Luige Vladareanu, Romanian Academy, Romania

Dongyun Wang, Zhongyuan University of Technology, China

Members List: Adrian Olaru, University Politehnica of Bucharest, Romania

Ajiboye Osunleke, Obafemi Awolowo University, Nigeria

Chenzhong Li, Florida International University, USA

Cunchen Gao, Ocean University of China, China

Dong Yue, Huazhong University of Science & Technology, China

Dongya Zhao, China University of Petroleum, China

Eldert van Henten, Wageningen University, Netherlands

Feng Qiao, Shenyang JianZhu University, China

Guangdong Tian, Northeast Forestry University, China

Haibin Yu, Chinese Academy of Sciences, China

Hao Chen, China University of Mining and Technology, China

Hiroyuki Takanashi, Nihon University, Japan

Hongbo Wang, Yanshan University, China

Huijun Gao, Harbin Institute of Technology, China

Huimin Xiao, Henan University of Finance and Economics, China

Huiping Li, Northwestern Ploytechnical University, China

Ionel Staretu, Transilvania University of Brasov, Romania

Jiacun Wang, Monmouth University, USA

Jianqiang Yi, Institute of Automation, China

Jing Yao, Tongji University, China

Jinliang Ding, Northeastern University, China

Jin-Oh Hahn, University of Alberta, Canada

John Billingsley, University of Southern Queensland, Australia

John Reid, John Deere Moline Technology Innovation Center, USA

Program Committee (2)
Members List: Jose Ragot, Nancy-Universite, France

Junming Xiao, Zhongyuan University of Technology, China

Kazuki Umemoto, Kanagawa University, Japan

Kazuyuki Ito, Hosei University, Japan

Liangyong Wang, Northeastern University, China

Longguo Jin, Qingdao Technical College, China

Marcel Bergerman, Carnegie Mellon University, USA

Mitsuaki Ishitobi, Kumamoto University, Japan

Nader Barsoum, Cutrin University of Technology, Malaysia

Ni Bu, Qingdao University of Science and Technology, China

Nobutaka Wada, Hiroshima University, Japan

Pandian Vasant, University Technology Petronas, Malaysia

Puya Ghasemi Afshar, Yokogawa UK Ltd., UK

Shan Liang, Chongqin University, China

Shaoyuan Li, Shanghai Jiaotong University, China

Shigang Yue, University of Lincoln, UK

Shuichi Wakimoto, Okayama University, Japan

Tomohiro Henmi, Kagawa National College of Technology, Japan

Wanli Yang, Academy of Armored Force Engineering, China

Wei Wang, Dalian University of Technology, China

Wenxin Liu, New Mexico State University, USA

Xiangjie Liu, North China Electric Power University, China

Xiaolei Wang, Zhongyuan University of Technology, China

Yang Liu, University of Aberdeen, UK

YangQuan Chen, University of California, USA

Yinlai Jiang, The University of Electro-Communications, Japan

Yonghong Tan, Shanghai Normal University, China

Yukinori Nakamura, Tokyo University of Agriculture and Technology, Japan

Zhengguang Hou, Institute of Automation, China

Zhengtao Ding, The University of Manchester, UK

Zofia Lukszo, Delft University of Technology, Netherlands

 2015 International Conference on

Advanced Mechatronic Systems

August 22-24, 2015
Beijing, China

GREETINGS FROM THE GENERAL CHAIRS

On behalf of ICAMechS 2015 Organizing Committee, it is our great pleasure and honor to welcome
you all to the 2015 International Conference on Advanced Mechatronic Systems. The conference is
held on August 22-24, 2015 in Beijing, China, sponsored by International Journal of Advanced
Mechatronic Systems (ICAMechS), Beijing University of Posts and Telecommunications, Tokyo
University of Agriculture and Technology, Tokai University, IEEE Systems, Man, and Zhongyuan
University of Technology, Cybernetics Society, IEEE Robotics & Automation Society Agricultural
Robotics and Automation Technical Committee, International Journal of Modelling, Identification
and Control and The Institute of Complex Medical Engineering and cooperation with The Society of
Instrument and Control Engineers, The Institute of Systems, Control and Information Engineers
and The Japan Society of Mechanical Engineers.

ICAMechS 2015 is an all-volunteer conference, and it provides an international forum for

professionals, academics, and researchers to present latest developments from interdisciplinary
theoretical studies, computational algorithm development and applications of mechatronic systems.
Reflecting these methodological and technological trends, this conference includes presentations of
excellent and interesting papers in all theoretical studies and applications of advanced mechatronic
systems. Novel quantitative engineering and science studies are also considered.

Beijing is the capital of the People's Republic of China located in the north of China. It is the
political, cultural, educational and diplomatic center of China. Beijing is one of the most populous
ancient cities in the world with a long history. There are many magnificent historical interesting
places, such as The Great Wall, Palace Museum (Forbidden City), Summer Palace, Tiananmen
Square and Temple of Heaven Park which attract many tourists from all over the world. Beijing is
also a food-lovers dream, the food is delicious. In this summer, we expect you to fully experience the
unique charm and vastness of the city. Hope to see you in Beijing.

In this conference we have invited three distinguished speakers, Prof. Anthony A. Maciejewski
(Colorado State University. USA), Prof. Tianyou Chai (Member of Chinese Academy of Engineering,
Northeastern University, China), and Prof. Chun-Yi Su, (Concordia University, Canada), as plenary
speaker, and we also have four organized sessions and one invited session. We would like to present
our sincere thanks to plenary speakers and organizers of special sessions for their valuable
contribution.
Finally, we would like to present our special thanks to authors and participants of this conference,
and the members of Advisory Committee, Program Committee and Organizing Committee.
Especially, we would like to give our sincere acknowledgement to local arrangement chair, Profs.
Shengjun Wen and Dianwei Qian, for their devoted work to success of the conference.

We hope that you all enjoy the conference and beautiful Beijing.

Best regards,
General Chairs

Mingcong Deng Xiaoguang Zhou Hongnian Yu Mengchu Zhou Shinji Wakui

CONFERENCE HIGHLIGHTS

Under three reviews for each paper by Program Committees and Organizing Committees, a total of 113
papers from different parts of the world were accepted and included into the final program of ICAMechS
2015. Among which 79 are the oral papers and 34 are poster papers. There are three plenary speeches, one
plenary panel discussion. Each oral session consists of 5-6 papers, where we have one invited session, 4
organized sessions, one poster session and 10 regular sessions. Papers were assigned with the sole purposed
of forming coherent sessions.

CONFERENCE REGISTRATION

The conference registration desk, located at the lobby of Beijing Fujian Hotel, will be opened during the
following time:
14:30-17:30, August 21, 2015 (Friday)
08:10-17:30, August 22, 2015 (Saturday)
08:20-17:30, August 23, 2015 (Sunday)

The full registration includes Welcome Reception, Conference Banquet, Technical Tour and Conference USB
Proceedings.

Additional sets of USB proceedings may be purchased at the registration desk ($50 USD for USB
proceedings).

SOCIAL EVENTS

Welcome Reception (Beijing Fujian Hotel, 18:30 - 20:30, August 21, 2015)
Conference Banquet (Beijing Fujian Hotel, 18:30 - 20:30, August 23, 2015)
Technical Tour (The Great Wall, Whole day, August 24, 2015)
CONFERENCE LOCATION

Beijing Fujian Hotel

Address: No.11, Anzhen Xili, Chaoyang District, Beijing, China
All conference rooms are at second floor of Beijing Fujian Hotel, the room guide map will be
available at the conference registration desk.
The conference site map is given as follows.

Beijing North
Railway Station

Beijing
Railway Station

Beijing West
Railway Station
Beijing South
Railway Station

Transportation

From the Airport (About 25.2km)

Take airport bus Gongzhufen line () to Anzhen Bridge bus stop and walk to Beijing Fujian
Hotel, About 1h50mins
Subway Line: Airport LineNo.10 No.8, About 1h30mins
Take a taxi for about 31 minutes

From Railway station

Beijing Railway Station-
Subway Line: No.2 No.8, About 50mins
Take a taxi for about 29 minutes
Beijing West Railway Station-
Subway Line: No.9 No.6 No.8, About 50mins
Take a taxi for about 36 minutes
 Beijing North station-
Bus No.387, About 50mins
Take a taxi for about 18 minutes
Beijing South Railway Station-
Subway Line: No.2 (Daxing line) No.2 No.8, About 50mins
Take a taxi for about 47 minutes

D B C
A
A A
A
A
A A
A A
A A A

3
1
A
A A
A A 2 1
A A A
4
A A
A
A

Nearby stations:
Subway station Bus stop

A Anhua Bridge subway station ()
1 Anzhen Bridge ()

B Bus North Tucheng subway station ()
2 Anzhen Bridge west ()

C Anzhenmen subway station ()
Anhua Bridge north()

D Jiandemen subway station ()
4 Anhua Bridge south ()

HOTEL RESERVATIONS

1. Beijing Fujian Hotel (Conference site)

Address: No.11, Anzhen Xili, Chaoyang District, Beijing, China
Phone:+86-10-64428833
URL: http://beijingfujianhotel.com/en/
2. Nearby Hotels
Yuanshan Hotel Beijing (About walk 1.4 km)
Address: No. 2 Deshengmen Yumin Street Xicheng district- Beijing-China
Phone: 4007-889-889
 URL: http://www.dfzsh.com/hotel/more-9355
Buding Hotel water cube (Beijing Anhua bridge) (About 870m)
Address: No. 3 North Third Ring Road jia Xicheng district- Beijing-China
Phone: 010-62058128
URL: http://www.budingjiudian.com/
Buding Hotel Beijing Anzhen bridge (About 960m)
Address: No. 9 BuildingAnzhenli 2 district-Chaoyang district- Beijing-China
Phone: 010-64448788
URL: http://www.budingjiudian.com/
Buding Hotel Beijing Anzhen bridge (About 960m )
Address: No. 9 BuildingAnzhenli 2 district-Chaoyang district- Beijing-China
Phone: 010-64448788
URL: http://www.budingjiudian.com/
Beijing GeHua Kaiyuan Hotel (About 990m)
Address: GuLou outer street Chaoyang district- Beijing-China
Phone: 4007-889-889
URL: http://www.dfzsh.com/hotel/4616
Zhejiang Tower Hotel (About 250m )
Address: No. 26 Anzhen xili 3 district- No.3 North ring-Chaoyang District-Beijing-China
Phone: 010-5166 3288
URL: http://www.huilv.com/H/cantact12843/

* Red circles indicate the hotels. Please check the corresponding hotel numbers and locations.
PLENARY SPEAKERS

Prof. Tianyou Chai, Northeastern University, China

Prof. Tianyou Chai received the Ph.D. degree in control theory and engineering in 1985 from
Northeastern University, Shenyang, China, where he became a Professor in 1988. He is the
founder and Director of the Center of Automation, which became a National Engineering
and Technology Research Center and a State Key Laboratory. He is a member of Chinese
Academy of Engineering, IFAC Fellow and IEEE Fellow, director of Department of
Information Science of National Natural Science Foundation of China. His current research
interests include modeling, control, optimization and integrated automation of complex
industrial processes. He has published 150 peer reviewed international journal papers. His paper titled Hybrid
intelligent control for optimal operation of shaft furnace roasting process was selected as one of three best papers
for the Control Engineerig Practice Paper Prize for 2011-2013. He has developed control technologies with
applications to various industrial processes. For his contributions, he has won 4 prestigious awards of National
Science and Technology Progress and National Technological Innovation, the 2007 Industry Award for
Excellence in Transitional Control Research from IEEE Multiple-conference on Systems and Control.

Title:

Intelligent Feedback Control for Operation of Complex Industrial Processes

Abstract:

Process control should aim at not only ensuring controlled variables to best follow their set points, but also
requiring the optimal control for the operation of the whole plant to make the operational indices (e.g. quality,
efficiency and consumptions during the production phase) into their targeted ranges. It also requires that
operational indices for quality and efficiency should be enhanced as high as possible, whilst the indices related to
consumptions are kept at their lowest possible level. Based upon a survey on the existing operational optimization
and control methodologies, this talk presents a data-driven hybrid intelligent feedback control for operation of
complex industrial processes and a hybrid simulation system. Simulations and industrial applications to a roasting
process for the hematite ore mineral processing industry are used to demonstrate the effectiveness of the proposed
method. Issues for future research on the optimal operational control for complex industrial processes are outlined
in the final section.
Prof. Anthony A. Maciejewski, Colorado State University. USA

Anthony A. (Tony) Maciejewski received the B.S., M.S., and Ph.D. degrees in Electrical
Engineering in 1982, 1984, and 1987, respectively, all from The Ohio State University. From
1988 to 2001, he was a Professor of Electrical and Computer Engineering at Purdue
University. In 2001, he joined Colorado State University where he is currently the Head of
the Department of Electrical and Computer Engineering. He has co-authored over 250
technical publications in the areas of robotics and high-performance computing and served
on eight journal editorial boards and over 100 conference program committees. He is a Fellow of IEEE and
currently serves as the Vice-President of Financial Activities for the Robotics and Automation Society. A complete
up-to-date vita is available at www.engr.colostate.edu/~aam
Title:

Designed to Fail: Incorporating Kinematic Redundancy into Robotic Systems

Abstract:

Operations in remote or extremely hazardous environments, e.g., in space, underwater, or in nuclear/chemical

waste sites, are frequently performed by robotic devices. Unfortunately, the very nature of these environments
makes the probability of a failure in the mechanical or electronic hardware of the robot more likely. Because by
definition the operating environment precludes the intervention of a human to repair or retrieve the device, fault
tolerance becomes a critical issue. By fault tolerance it is meant that damage to any portion of the robotic system,
such as the mechanical actuators, control electronics, sensors, etc., will not render the system uncontrollable. A
damaged fault tolerant system can be operated, albeit in a somewhat degraded capacity, to complete the assigned
task or at a minimum to retrieve the robotic device without compromising the working environment. This
presentation will focus on two approaches for designing failure tolerant robotic systems. The first is to develop
software that can control robots of an arbitrary design to achieve maximum fault tolerance. This would allow the
use of conventional manipulators to their maximum capability, regardless of the extent of the damage to the
device. The second is to design robotic arms with kinematic redundancy that maximizes the dexterity and
maneuverability of the arm after the occurrence of a fault.
Prof. Chun-Yi Su, Concordia University, Canada

Dr. Chun-Yi Su received his Ph.D. degrees in control engineering from South China
University of Technology in 1990. After a seven-year stint at the University of Victoria, he
joined the Concordia University in 1998, where he is currently a Professor of Mechanical
and Industrial Engineering and holds the Concordia Research Chair in Control. He has also
held several short-time visiting positions including a Chang Jiang Chair Professorship by
Chinas Ministry of Education and JSPS Invitation Fellowship from Japan, and Qian Ren
Talents Professor from China. His research covers control theory and its applications to various mechanical
systems, with a focus on control of systems involving hysteresis nonlinearities. He is the author or co-author of
over 300 publications, which have appeared in journals, as book chapters and in conference proceedings. In
addition to his academic activities, he has worked extensively with industrial organizations on various projects.
Dr. Su has been an Associate Editor of IEEE Transactions on Automatic Control, IEEE Transactions on Control
Systems Technology, Mechatronics, Control Engineering Practice, and several other journals. He has served as
Chair/Co-Chair for numerous international conferences.

Title:

Modeling and Control of Hysteresis Nonlinearities in Smart Actuators

Abstract:

Hysteresis nonlinearities are very common in magnetic materials, smart materials and industrial systems. For
decades, the existence of such nonlinearities have provided one of the most difficult challenges to control design
engineers since the entire Laplace domain and most state space control design techniques were developed
exclusively for differentiable linear or nonlinear systems. Hence, the existence of hysteresis nonlinearities in smart
material based actuators and systems were neglected and the controllers were designed based on the nominal
smooth systems. When the systems are considered with non-differentiable nonlinearities, these methods
encountered substantial difficulties in the analysis, model fitting and control design stages. It was extremely
difficult, if not impossible, to design or prove stability of such systems. The development of techniques for the
identification of such nonlinearities in realistic industrial plants has emerged as a significant problem in itself.

This talk is intended to raise awareness of modeling and control techniques and to provide an opportunity to
discuss state-of-the-art solutions for the problems. The presentation and discussion will range from modeling of
hysteresis, to the design of corresponding control schemes, especially in the absence of complete information
concerning the system model and state.
 2015 International Conference on
Advanced Mechatronic Systems

August 22-24, 2015

Beijing, China

CONFERENCE PROGRAM

Friday, August 21, 2015

14:30 - 17:30 Registration (Beijing Fujian Hotel, Beijing)

18:30 - 20:30 Welcome Reception (Beijing Fujian Hotel, Beijing)

 Saturday, August 22, 2015

08:40-08:50 Opening Remarks Main Conference Hall

08:50-09:50 Plenary Speech 1 (PS01) Main Conference Hall

Title: Intelligent Feedback Control for Operation of Complex Industrial Processes

Speaker: Tianyou Chai, Northeastern University, China

Chair: Hongnian Yu, Bournemouth University, UK
09:50-10:00 Break
10:00-11:00 Plenary Speech 2 (PS02) Main Conference Hall

Title: Modeling and Control of Hysteresis Nonlinearities in Smart Actuators

Speaker: Chun-Yi Su, Concordia University, Canada

Chair: Xinkai Chen, Shibaura Institute of Technology, Japan
11:00-13:00 Lunch Break

13:00-14:40 Parallel paper sessions

Room No.01 Room No.02 Room No.03

SatM01 SatM02 SatM03

Renewable sources of energy Advanced intelligent control in Optimization methods and

robotics and mechatronics applications I
14:40-14:50 Break

14:50-16:30 Parallel paper sessions

Room No.01 Room No.02 Room No.03

SatP01 SatP02 SatP03

Optimization methods and Advanced nonlinear control in Control of mechatronic systems

applications II robotics I
16:30-16:40 Break

16:40-18:20 (18:40) Parallel paper sessions

Room No.01 Room No.02 Room No.03

SatE01 SatE02 SatE03

New developments on ESP Information processing Nonlinear systems modeling,

control and applications
 Sunday, August 23, 2015

08:50-09:50 Plenary Speech 3 (PS03) Main Conference Hall

Title: Designed to Fail: Incorporating Kinematic Redundancy into Robotic Systems

Speaker: Anthony A. Maciejewski, Colorado State University, USA

Chair: J. X. Zhang, Swinburne University of Technology, Australia

09:50-10:20 Break

10:20-11:30 Poster Session Main Conference Hall

11:30-13:00 Lunch Break

13:00-14:40 Parallel paper sessions

Room No.01 Room No.02 Room No.03

SunM01 SunM02 SunM03

Advanced control systems Mechatronic systems analysis Complex systems control and
design - theory and applications and control I applications I
14:40-14:50 Break

14:50-16:50 Parallel paper sessions

Room No.01 Room No.02 Room No.03

SunP01 SunP02 SunP03

Control of mechatronic systems Mechatronic systems analysis Complex systems control and
II and control II applications II

16:50-17:00 Break

17:00-18:00 Plenary Panel Discussion Main Conference Hall

Title: New Developments in Advanced Mechatronic Systems

Organizers: Mingcong Deng, Tokyo University of Agriculture and Technology, Japan

Hongnian Yu, Bournemouth University, UK

18:30-20:30 Banquet (Beijing Fujian Hotel, Beijing)

Monday, August 24, 2015

Technical Tour to The Great Wall of China

 2015 International Conference on
Advanced Mechatronic Systems

August 22-24, 2015

Beijing, China

TECHNICAL PROGRAM

ICAMechS 2015 ICAMechS 2015

Friday, August 21, 2015 Saturday, August 22, 2015

14:30 - 17:30 08:40 - 08:50 Main Conference Hall

Registration Opening Remarks
(Beijing Fujian Hotel, Beijing)
08:50 - 09:50 Main Conference Hall
18:30 - 20:30 Plenary Speech 1
Welcome Reception Intelligent Feedback Control for Operation of
Complex Industrial Processes
(Beijing Fujian Hotel, Beijing)
Tianyou Chai
Northeastern University, China
End of the 1st Day Chair: Hongnian Yu Bournemouth Univ., UK

09:50 - 10:00 Break

10:00 - 11:00 Main Conference Hall

Plenary Speech 2
Modeling and Control of Hysteresis Nonlinearities
in Smart Actuators
Chun-Yi Su
Concordia University, Canada
Chair: Xinkai Chen Shibaura Institute of Tech., Japan

11:00 - 13:00 Lunch Break

Early Afternoon Technical Sessions

13:00-14:40, Saturday, August 22, 2015

SatM01 Session Room No.01

Renewable sources of energy

Organized Session
Organizer: Zongxiao Yang Henan Univ. of Sci. and Tech., China mechatronics
Chair: Zongxiao Yang Henan Univ. of Sci. and Tech., China Organized Session
Co-Chair: Aihui Wang Zhongyuan Univ. of Tech., China Organizer: Luige Vladareanu Romanian Acad., Romania
Chair: Luige Vladareanu Romanian Acad., Romania
13:00 - 13:20 SatM01-01
Co-Chair: Zengguang Hou Chinese Acad. of Sci., China
Cluster analysis on the service oriented alliance enterprise
manufacturing resource 13:00 13:20 SatM02-01
Yunxia Wang Nanjing Institute of Tech., China Active training research of a lower limb rehabilitation robot
Shenghai Qiu Nanjing Institute of Tech., China based on constrained trajectory
Chengchong Gao Nanjing Institute of Tech., China Hongbo Wang Yanshan Univ., China
Jun Wei Nanjing Institute of Tech., China Dong Zhang Yanshan Univ., China
Hao Lu Yanshan Univ., China
13:20 - 13:40 SatM01-02 Yongfei Feng Yanshan Univ., China
Performance and exhaust emissions of a turbocharged common Peng Xu Yanshan Univ., China
rail DI diesel engine fueled with alcohol Razvan-Viorel Mihai Romanian Acad., Romania
Bin Xu Henan Univ. of Sci. and Tech., China Luige Vladareanu Romanian Acad., Romania
Yibin Liu Henan Univ. of Sci. and Tech., China
Jian Wu Henan Univ. of Sci. and Tech., China 13:20 13: 40 SatM02-02
Chuang Li Henan Univ. of Sci. and Tech., China Fuzzy controller design of the wire feeder of invasive vascular
Weiwei Shang Henan Univ. of Sci. and Tech., China interventional surgery robot
Zhihao Ma Henan Univ. of Sci. and Tech., China Hongbo Wang Yanshan Univ., China
Shuaishuai Wang Yanshan Univ., China
13:40 - 14:00 SatM01-03 Wen Zhang Yanshan Univ., China
Design and simulation of passive filter based on fundamental Liyu Xie Yanshan Univ., China
magnetic flux compensation Xue Yang Univ. of Jinan, China
Qun Chen Henan Univ. of Sci. and Tech., China Luige Vladareanu Romanian Acad., Romania
Zongxiao Yang Henan Univ. of Sci. and Tech., China Radu Munteanu Romanian Acad., Romania

14:00 14:20 SatM01-04 13:40 14:00 SatM02-03

Research on large supermarket fresh food supplier evaluation Advanced intelligent walking robot control through sliding
and selection based on SWOT-entropy weight fuzzy motion control and bond graphs methods
comprehensive model Ionel Alexandru Gal Romanian Acad., Romania
Ying Yuan Henan Univ. of Sci. and Tech., China Luige Vladareanu Romanian Acad., Romania
Yi Zhang Henan Univ. of Sci. and Tech., China Hongbo Wang Yanshan Univ., China
Long Di Henan Univ. of Sci. and Tech., China Hongnian Yu Bournemouth Univ., UK
Ganggang Wu Henan Univ. of Sci. and Tech., China Mingcong Deng Tokyo Univ. of Agri. and Tech., Japan
Zongxiao Yang Henan Univ. of Sci. and Tech., China
14:00 14:20 SatM02-04
14:20 14:40 SatM01-05 Versatile intelligent portable robot platform for flexible robotic
A novel outer rotor axial primary magnetic circuit permanent cells with AGV
magnet generator Luige Vladareanu Romanian Acad., Romania
Ximei Li Henan Univ. of Sci. and Tech., China Cristian Spirleanu Romanian Acad., Romania
Zongxiao Yang Henan Univ. of Sci. and Tech., China Mihaiela Iliescu Romanian Acad., Romania
Dongliang Song Henan Univ. of Sci. and Tech., China Hongnian Yu Bournemouth Univ., UK
Mingcong Deng Tokyo Univ. of Agri. and Tech., Japan
Weizhong Guo Shanghai Jiao Tong Univ., China
SatM02 Session Room No.02
Feng Gao Shanghai Jiao Tong Univ., China

Advanced intelligent control in robotics and

 14:20 14:40 SatM02-05
Haptic intelligent interfaces for NAO robot hand control
Octavian Melinte Romanian Acad., Romania 14:40 - 14:50 Break
Luige Vladareanu Romanian Acad., Romania
Ligia Munteanu Romanian Acad., Romania Middle Afternoon Technical Sessions
Hongnian Yu Bournemouth Univ., UK 14:50-16:30, Saturday, August 22, 2015
Shuang Cang Bournemouth Univ., UK
Zengguang Hou Chinese Acad. of Sci., China
SatP01 Session Room No.01

Guibin Bian Chinese Acad. of Sci., China

Hongbo Wang Yanshan Univ., China
Optimization methods and applications II
Regular Session
Chair: Guangdong Tian Northeast Forestry Univ., China
SatM03 Session Room No.03 Co-Chair: Zongxiao Yang Henan Univ. of Sci. and Tech., China

14:50 15:10 SatP01-01

Optimization methods and applications I
Scheduling of rescue vehicles to forest fires via multi-objective
Regular Session
particle swarm optimization
Chair: Shuhui Bi Univ. of Jinan, China
Yaping Ren Northeast Forestry Univ., China
Co-Chair: Huimin Xiao Henan Univ. of Economics and Law, China
Guangdong Tian Northeast Forestry Univ., China
13:00 13:20 SatM03-01 Mengchu Zhou New Jersey Institute of Tech., USA
Weak reachability of probabilistic boolean control networks
Zhiqiang Li Henan Univ. of Economics and Law, China 15:10 15:30 SatP01-02
Huimin Xiao Henan Univ. of Economics and Law, China An effective VNSSA algorithm for the blocking flowshop
scheduling problem with makespan minimization
13:20 13:40 SatM03-02 Chaoyong Zhang Huazhong Univ. of Sci. & Tech., China
Oscillation criteria for a class of nonlinear impulsive parabolic Zhanpeng Xie Huazhong Univ. of Sci. & Tech., China
system under Dirichlet boundary condition Xinyu Shao Huazhong Univ. of Sci. & Tech., China
Junhong Liu Acad. of Armored Force Eng., China Guangdong Tian Northeast Forestry Univ., China
Wenxue Liu Acad. of Armored Force Eng., China
Lifeng Li Acad. of Armored Force Eng., China 15:30 15:50 SatP01-03
Qi Jin Acad. of Armored Force Eng., China Simulation modeling of shuttle racking system based on automod
Bingfeng Sang Beijing Univ. of Posts and Telecommunications, China
13:40 14:00 SatM03-03 Yukun Liu Beijing Univ. of Posts and Telecommunications, China
Operator-based left coprime factorization for nonlinear systems
Fazhan Tao Tokyo Univ. of Agri. and Tech., Japan 15:50 16:10 SatP01-04
Mingcong Deng Tokyo Univ. of Agri. and Tech., Japan An expected value multi-objective optimization model to locate a
vehicle inspection station
14:00 14:20 SatM03-04 Guangdong Tian Northeast Forestry Univ., China
Passivity-based robust control for uncertain nonlinear feedback MengChu Zhou New Jersey Institute of Tech., USA
systems
Ni Bu Qingdao Univ. of Sci. and Tech., China 16:10 16:30 SatP01-05
Mingcong Deng Tokyo Univ. of Agri. and Tech., Japan Experimental study on nonlinear vibration control of an L-shape
arm
14:20 14:40 SatM03-05 Yanfeng Wu Tokyo Univ. of Agri. and Tech., Japan
Correlated neural activity in spiking networks with topographic Mingcong Deng Tokyo Univ. of Agri. and Tech., Japan
couplings Longguo Jin Qingdao Technical College, China
Jinli Xie Univ. of Jinan, China
Qinjun Zhao Univ. of Jinan, China
Jianyu Zhao Univ. of Jinan, China
 Regular Session
SatP02 Session Room No.02
Chair: Shinji Wakui Tokyo Univ. of Agri. and Tech., Japan
Co-Chair: Kazuyuki Ito Hosei Univ., Japan
Advanced nonlinear control in robotics
Regular Session
14:50 15:10 SatP03-01
Chair: Hongnian Yu Bournemouth Univ., UK
An interpretation of unbalance vibration compensator for
Co-Chair:Yukinori Nakamura Tokyo Univ. of Agri. and Tech., Japan
five-axes active magnetic bearing systems based on internal model
14:50 15:10 SatP02-01
principle
The mechatronic system of eco-dolphin - a fleet of autonomous
Taiki Nakamura Tokyo Univ. of Agri. and Tech., Japan
underwater vehicles
Shinji Wakui Tokyo Univ. of Agri. and Tech., Japan
Hong Liu Embry-Riddle Aeronautical Univ., USA
Yukinori Nakamura Tokyo Univ. of Agri. and Tech., Japan
Xudong Shi Civil Aviation Univ. of China, China
Junzhen Shao Embry-Riddle Aeronautical Univ., USA
15:10 15:30 SatP03-02
Qi Zhou Embry-Riddle Aeronautical Univ., USA
Vibration suppression of galvano mirror considering whirling of
Stacey Joseph-Ellison Embry-Riddle Aeronautical Univ., USA
shaft
Johnathan Jaworski Embry-Riddle Aeronautical Univ., USA
Keisuke Nakade Tokyo Univ. of Agri. and Tech., Japan
Ci Wen Embry-Riddle Aeronautical Univ., USA
Taiki Nakamura Tokyo Univ. of Agri. and Tech., Japan
Shinji Wakui Tokyo Univ. of Agri. and Tech., Japan
15:10 15:30 SatP02-02
Yukinori Nakamura Tokyo Univ. of Agri. and Tech., Japan
Extended development of LinuxCNC for control of a delta robot
Feng Huo National Univ. of Singapore, Singapore
15:30 15:50 SatP03-03
Geoksoon Hong National Univ. of Singapore, Singapore
A consideration of parameter identification of a linear stage using
Aunneow Poo National Univ. of Singapore, Singapore
particle swarm optimization
Marino Watanabe Tokyo Univ. of Agri. and Tech., Japan
15:30 15:50 SatP02-03
Yukinori Nakamura Tokyo Univ. of Agri. and Tech., Japan
Minimizing energy consumption in a U-shaped robotic assembly
Shinji Wakui Tokyo Univ. of Agri. and Tech., Japan
line
J. Mukund Nilakantan Monash Univ. Malaysia, Malaysia
15:50 16:10 SatP03-04
S.G. Ponnambalam Monash Univ. Malaysia, Malaysia
Development of a torque-sensorless viscometer for food
George Q. Huang The Univ. of Hong Kong, China
processing applications
Seiji Hashimoto Gunma Univ., Japan
15:50 16:10 SatP02-04
Yuqi Jiang Gunma Univ., Japan
Charged system search algorithm for robotic drill path
Amat Basari Gunma Univ., Japan
optimization
Wei Jiang Yangzhou Univ., China
G. Kanagaraj Thiagarajar College of Eng., India
S.G. Ponnambalam Monash Univ. Malaysia, Malaysia
16:10 16:30 SatP03-05
C.K. Loo Univ. of Malaya, Malaysia
Study on detection of earthquake using DI value and its
application to switching control for air type anti-vibration
16:10 16:30 SatP02-05
apparatus
Implementation of remotely controllable mobile robot system
Masayuki Kubota Tokyo Univ. of Agri. and Tech., Japan
over mobile Ad-hoc network
Hirokazu Mineo Tokyo Univ. of Agri. and Tech., Japan
Mohammad Shahidul Hasan Staffordshire Univ., UK
Shinji Wakui Tokyo Univ. of Agri. and Tech., Japan
Hongnian Yu Bournemouth Univ., UK
Yukinori Nakamura Tokyo Univ. of Agri. and Tech., Japan

SatP03 Session Room No.03

16:30 - 16:40 Break

Control of mechatronic systems I
 Last Afternoon Technical Sessions
16:40-18:20, Saturday, August 22, 2015
SatE02 Session Room No.02
SatE01 Session Room No.01
Information processing
New developments on ESP Regular Session
Invited Session Chair: S.G. Ponnambalam Monash Univ. Malaysia, Malaysia
Organizer: Jie Shi Univ. of Electro-Communications, Japan Co-Chair: Ken Nakasaka Tokyo Univ. of Agri. and Tech., Japan
Chair: Jie Shi Univ. of Electro-Communications, Japan
16:40 17:00 SatE02-01
Co-Chair: Yan Yu Univ. of Electro-Communications, Japan
Modelling a 20MW scale solar farm in an unused angled area
16:40 - 17:00 SatE01-01 near Fukushima nuclear power plant
An overview of supplementary teaching materials for advanced Amin Mohammadirad Tokyo Univ. of Agri. and Tech., Japan
EST students in Japan Ken Nagasaka Tokyo Univ. of Agri. and Tech., Japan
Yan Yu Univ. of Electro-Communications, Japan
Jie Shi Univ. of Electro-Communications, Japan 17:00 17: 20 SatE02-02
Temporal and spatial projection of wind speed based on modular
17:00 - 17:20 SatE01-02 network SOM for installation of off-shore wind generation
Profiling Japanese EST students vocabulary ability using the Mitsuharu Hayashi Tokyo Univ. of Agri. and Tech., Japan
new general service list test (NGSLT) and the new academic Ken Nagasaka Tokyo Univ. of Agri. and Tech., Japan
word list test (NAWLT)
Shinichi Hashimoto Univ. of Electro-Communications, Japan 17:20 17:40 SatE02-03
Yan Yu Univ. of Electro-Communications, Japan A household load priority scheduling model based on real-time
Jie Shi Univ. of Electro-Communications, Japan electricity price
Haiyan Zhan Univ. of Electro-Communications, Japan Bo Hu State Grid Energy Research Institute, China
Xiao Zhong North China Electric Power Univ., China
17:20 - 17:40 SatE01-03 Song Liu North China Electric Power Univ., China
Peer evaluation of academic presentations by Japanese graduate
students: qualitative evaluation 17:40 18:00 SatE02-04
Jie Shi Univ. of Electro-Communications, Japan Study on cutting surface stripe and wettability of AZ91D
Cross John Univ. of Electro-Communications, Japan magnesium alloy by WEDM-HS
Jinkai Xu Changchun Univ. of Sci. and Tech., China
17:40 18:00 SatE01-04 Kui Xia Changchun Univ. of Sci. and Tech., China
Teaching poster by using genre approach to science and Rongxian Qiu Changchun Univ. of Sci. and Tech., China
technology students of a Japanese university Linshuai Zhang Changchun Univ. of Sci. and Tech., China
Toyo Univ., Univ. of Zhanjiang Yu Changchun Univ. of Sci. and Tech., China
Rajagopalan Maheswari Uma
Electro-Communications Huadong Yu Changchun Univ. of Sci. and Tech., China
Jie Shi Univ. of Electro-Communications, Japan

18:00 18:20 SatE02-05

18:00 18:20 SatE01-05 Influence of microstructure on surface wettability
Corpus-based analysis of academic RA genre: the Results Xinxin Zhang Changchun Univ. of Sci. and Tech., China
sub-genre Jinkai Xu Changchun Univ. of Sci. and Tech., China
Louvign Sbastien Univ. of Electro-Communications, Japan Zhongxu Lian Changchun Univ. of Sci. and Tech., China
Jie Shi Univ. of Electro-Communications, Japan Zhanjiang Yu Changchun Univ. of Sci. and Tech., China
Huadong Yu Changchun Univ. of Sci. and Tech., China
18:20 18:40 SatE01-06
Understanding the most essential types of words for Physics
undergraduate students
SatE03 Session Room No.03

Sonia Sharmin Univ. of Tsukuba, Japan

 Nonlinear systems modeling control and
applications
Organized Session
Organizer: Shiwen Tong Beijing Union Univ., China
Chair: Shiwen Tong Beijing Union Univ., China
Co-Chair: Dianwei Qian North China Electric Power Univ., China

16:40 17:00 SatE03-01

Influences from LCL filter on behaviors of three-phase
three-level PWM rectifier
Dongsheng Yu China Univ. of Mining and Tech., China
Hong Zhu China Univ. of Mining and Tech., China
Ciyan Zheng China Univ. of Mining and Tech., China

17:00 17:20 SatE03-02

Sliding mode output tracking control based on a fuzzy clustering
model
Shiwen Tong Beijing Union Univ., China
Dianwei Qian North China Electric Power Univ., China
Jianjun Fang Beijing Union Univ., China

17:20 17:40 SatE03-03

PID control of air tank temperature system with parameters
tuning through network
Shiwen Tong Beijing Union Univ., China
Yushan Li Beijing Union Univ., China
Junjie Ren Beijing Union Univ., China
Yinong Zhang Beijing Union Univ., China

17:40 18:00 SatE03-04

Sliding-mode-based disturbance rejection control of nonlinear
power systems with renewable sources
Hong Liu North China Electric Power Univ., China
Dianwei Qian North China Electric Power Univ., China

18:00 18:20 SatE03-05

Research and application of predictive control algorithm based
on fuzzy model
Yuan Li Beijing union Univ., China
Sen Cong Beijing union Univ., China
Shiwen Tong Beijing union Univ., China
Jianjun Fang Beijing union Univ., China

End of the 2nd Day

 Comprehensive functions of eco-industrial park in conserving
ICAMechS 2015 energy and improving ecology
Sunday, August 23, 2015 Xinjiang Univ. of Finance and
Litifu Zulati
Economics, China
08:50 - 09:50 Main Conference Hall Ken Nagasaka Tokyo Univ. of Agri. and Tech., Japan
Plenary Speech 3
Designed to Fail: Incorporating Kinematic 10:20 11:30 Poster-05

Redundancy into Robotic Systems Study on fuzzy-logic PID control system of synchronous jack-up
system of turbine runner static equilibrium experiment
Anthony A. Maciejewski
North China Univ. of Water Resources and
Colorado State University. USA Yongxing Hao
Electric Power, China
Chair: J. X. Zhang Swinburne University of Technology, Australia
North China Univ. of Water Resources and
Chaofeng Wang
Electric Power, China
09:50 - 10:20 Break
North China Univ. of Water Resources and
Lihui Song
Technical Poster Session Electric Power, China

10:20-11:30, Sunday, August 23, 2015

10:20 11:30 Poster-06
Poster Main Conference Room
The straight line mechanism base on peaucellier inversion
instrument

Chair: Changan Jiang Ritsumeikan Univ., Japan Beijing Univ. of Posts and
Kai He
Xiaoguang Zhou Beijing Univ. of Posts and Telecommunications, China

Telecommunications, China Beijing Univ. of Posts and

Lei Guo
Co-Chair: Shin Wakitani Tokyo Univ. of Agri. and Tech., Japan Telecommunications, China

Yixin Yin Univ. of Sci. and Tech. Beijing, China Beijing Univ. of Posts and
Yuan Song
Telecommunications, China
10:20 11:30 Poster-01
Bipolar neutrosophic sets and their application based on 10:20 11:30 Poster-07
multi-criteria decision making problems Mechanical design and realization of automatic docking
Irfan Deli Aralik University, Turkey mechanism
Mumtaz Ali Quaid-e-azam Univ., Pakistan Beijing Univ. of Posts and
Xinhu Mo
Florentin Smarandache Univ. of New Mexico, USA Telecommunications, China
Beijing Univ. of Posts and
Lei Guo
10:20 11:30 Poster-02 Telecommunications, China
Evaluation method of main production resource in society and Beijing Univ. of Posts and
Yuan Song
industrial park of extreme arid region (1) Telecommunications, China
Xinjiang Univ. of Finance and Beijing Univ. of Posts and
Litifu Zulati Ying Zhang
Economics, China Telecommunications, China

10:20 11:30 Poster-03 10:20 11:30 Poster-08

Research on control methods of permanent magnet synchronous Design and implementation of a mechanical calculator for
motor position servo system quadratic equation with one unknown
Fen Yang Xidian Univ., China
Beijing Univ. of Posts and
Bin Xing
Zhengfeng Ming Xidian Univ., China
Telecommunications, China
Tao Wen Xidian Univ., China
Beijing Univ. of Posts and
Lei Guo
Tong Zhang Xidian Univ., China
Telecommunications, China
Zhanxia Zhu Northwestern Polytechnical Univ., China
Beijing Univ. of Posts and
Shimin Wei
Telecommunications, China
10:20 11:30 Poster-04 Yuan Song Beijing Univ. of Posts and
 Telecommunications, China
Beijing Univ. of Posts and 10:20 11:30 Poster-15
Ying Zhang
Telecommunications, China Strongly stable GPC with suppression of steady state gain and
closed-loop poles
10:20 11:30 Poster-09 Akira Inoue Okayama Univ., Japan

A comparative study on inter-tern short circuit fault of PMSM National Institute of Tech., Kagawa
Tomohiro Henmi
using finite element analysis and experiment College, Japan
Yongcan Li Hebei Univ. of Sci. and Tech., China Mingcong Deng Tokyo Univ. of Agri. and Tech., Japan
Yongchun Liang Hebei Univ. of Sci. and Tech., China
10:20 11:30 Poster-16
10:20 11:30 Poster-10 Model free fault detection using one set of reference input and
A research on construction of 30 bus large-scale smart grid model output data
Sho Kainose Tokyo Univ. of Agri. and Tech., Japan Akira Inoue Okayama Univ., Japan
Ken Nagasaka Tokyo Univ. of Agri. and Tech., Japan National Institute of Tech., Kagawa
Tomohiro Henmi
College, Japan
10:20 11:30 Poster-11 Mingcong Deng Tokyo Univ. of Agri. and Tech., Japan
Modeling a lithium-ion battery based on a threshold model
Zhi Zhang Beijing Jiaotong Univ., China
10:20 11:30 Poster-17
10:20 11:30 Poster-12 Research on algorithm of nonlinear self-adaptive flight control
Modelling the relationship between constituents and anti-RSV system
effect of baikal skullcap root Hao Long Beijing Union Univ., China
Ke Li Shandong Acad. of Sci., China Shujie Song Aviation Industry Co. of China, China
Shandong Acad. of Chinese Medicine,
Bonian Zhao
China 10:20 11:30 Poster-18
Shandong Acad. of Chinese Medicine, The indoor real-time 3D localization algorithm using UWB
Yan Gao
China Xuehong Li Univ. of Sci. and Tech. Beijing, China
Shandong Acad. of Chinese Medicine, Shuhua Yang Univ. of Sci. and Tech. Beijing, China
Bianli Wang
China
Shandong Acad. of Chinese Medicine, 10:20 11:30 Poster-19
Zongyuan Yu
China Operator and sliding mode based nonlinear control for cooling
Qianfo Mountain Hospital of Shandong and heat-retention system actuated by Peltier devices
Hongpeng Zhao
Province, China Takafumi Hatano Tokyo Univ. of Agri. and Tech., Japan
Mingcong Deng Tokyo Univ. of Agri. and Tech., Japan
10:20 11:30 Poster-13 Shin Wakitani Tokyo Univ. of Agri. and Tech., Japan
Energy-saving improvement and simulation study on digital
hydraulic system 10:20 11:30 Poster-20
Fei Song Naval Univ. of Eng., China Control system design for ionic polymer metal composite using a
Jingjun Lou Naval Univ. of Eng., China single neuron based adaptive PID approach
Likun Peng Naval Univ. of Eng., China Yiping Chang Zhongyuan Univ. of Tech., China
Hui Wang Zhongyuan Univ. of Tech., China
10:20 11:30 Poster-14

Path planning of airfoil surface for robotic fibre placement 10:20 11:30 Poster-21
Evaluation method of main production resource of society and
Lina Li Chinese Acad. of Sci., China
industrial park of extreme arid region (2)
Xingang Wang Chinese Acad. of Sci., China
Xinjiang Univ. of Finance and
De Xu Chinese Acad. of Sci., China Litifu Zulati
Economics, China
Min Tan Chinese Acad. of Sci., China
 10:20 11:30 Poster-22 Qixian Zhou Henan Univ. of Sci. and Tech., China
Dynamic and energy analysis based on bumper system low-speed Tengteng Cui Henan Univ. of Sci. and Tech., China
collision test
Zhiming Wang Shanghai Univ., China 10:20 11:30 Poster-29
Li Chen Shanghai Univ., China Maximum efficiency control method of permanent magnet
synchronous motor based on three-dimensional table
10:20 11:30 Poster-23 Gensheng Li Henan Univ. of Sci. and Tech., China
Feasibility of neucube spiking neural network architecture for Jinfa Xie Henan Univ. of Sci. and Tech., China
EMG pattern recognition Liyou Xu Henan Univ. of Sci. and Tech., China
Long Peng Chinese Acad. of Sci., China
Zengguang Hou Chinese Acad. of Sci., China 10:20 11:30 Poster-30
Nikola Kasabov Auckland Univ. of Tech., New Zealand Modeling and simulation of hydro-mechanical continuously
Guibin Bian Chinese Acad. of Sci., China variable transmission system based on Simscape
Luige Vladareanu Romanian Acad., Romania Zhili Zhou Henan Univ. of Sci. and Tech., China
Hongnian Yu Bournemouth Univ., UK Jiazhen Zhang Henan Univ. of Sci. and Tech., China
Zhiqiang Guo YTO Group Corporation, China
10:20 11:30 Poster-24 Liyou Xu Henan Univ. of Sci. and Tech., China

Optimal fiber orientations and topology of compliant mechanisms

using lamination parameters 10:20 11:30 Poster-31
Xinxing Tong Northwestern Polytechnical Univ., China Performance comparison for savonius type wind turbines by
Wenjie Ge Northwestern Polytechnical Univ., China numerical analysis approaches
Yonghong Zhang Northwestern Polytechnical Univ., China Lei Song Henan Univ. of Sci. and Tech., China
Hongzhao Liu Xian Univ. of Tech., China
10:20 11:30 Poster-25 Zongxiao Yang Henan Univ. of Sci. and Tech., China
A software platform for interactive analysis on 3D medical data
Wei Cheng Shandong Acad. of Sci., China 10:20 11:30 Poster-32
Yongfei Xiao Shandong Acad. of Sci., China Fuzzy clustering analysis of batch production recipe
Taikoo (Shandong) Aircraft Eng. Shanzhong Liu Henan Univ. of Sci. and Tech., China
Youlei Zhao
Co.,Ltd., China Xiaonan Song Henan Univ. of Sci. and Tech., China

10:20 11:30 Poster-26 10:20 11:30 Poster-33

Master-slave force control based on Grey GM(1,1) model of robot Fully discrete subgrid stabilized finite element method for the
grippper Darcy-Brinkman equations in double-diffusion convection
Wei Cheng Shandong Acad. of Sci., China Yunzhang Zhang Henan Univ. of Sci. and Tech., China
Xuelin Wang Shandong Acad. of Sci., China Zhoufeng Wang Henan Univ. of Sci. and Tech., China
Haiyan Ma Shandong Acad. of Sci., China Qili Tang Henan Univ. of Sci. and Tech., China

10:20 11:30 Poster-27 10:20 11:30 Poster-34

Remote automatic test system based on MATLAB using VISA The simplified homogeneous balance method and its application
over LAN to Whitham-Broer-Kaup-Like equations
Wei Cheng Shandong Acad. of Sci., China Wei Li Henan Univ. of Sci. and Tech., China
Fenggui Wang Shandong Acad. of Sci., China Xinrui Wang Univ. of Chicago, USA
Haiyan Ma Shandong Acad. of Sci., China

10:20 11:30 Poster-28

11:30 - 13:00 Lunch Break
Cubic spline solution for a class of boundary value problems
using spectral collocation method
Tianjun Wang Henan Univ. of Sci. and Tech., China
 Early Afternoon Technical Sessions Mechatronic systems analysis and control I
13:00-14:40, Sunday, August 23, 2015 Regular Session
Chair: Xinkai Chen Shibaura Institute of Tech., Japan
SunM01 Session Room No.01
Co-Chair: Hongbo Wang Yanshan Univ., China

Advanced control systems design - theory and 13:00 13:20 SunM02-01

applications Size and weight reduction of semicircular duplex manipulator for
Organized Session rescue operation
Organizer: Shin Wakitani Tokyo Univ. of Agri. and Tech., Japan Soichiro Kimura Hosei Univ., Japan
Takao Sato Univ. of Hyogo Kazuyuki Ito Hosei Univ., Japan
Chair: Shiro Masuda Tokyo Metropolitan Univ., Japan
Co-Chair: Shin Wakitani Tokyo Univ. of Agri. and Tech., Japan 13:20 13:40 SunM02-02
Adaptive fault diagnosis for continuous time-delay repetitive
13:00 13:20 SunM01-01
system subject to sensor fault
Tracking control of quad-rotor helicopters suspended a power
Haigang Zhang Univ. of Sci. and Tech. Beijing, China
supply cable with on-line estimation of disturbances
Sen Zhang Univ. of Sci. and Tech. Beijing, China
Hayami Hayakawa Nagoya Institute of Tech., Japan
Yixin Yin Univ. of Sci. and Tech. Beijing, China
Shogo Obata Nagoya Institute of Tech., Japan
Manabu Yamada Nagoya Institute of Tech., Japan
13:40 14:00 SunM02-03
Model identification and adaptive control design for a 6 DOFs
13:20 13:40 SunM01-02
manipulator
Multivariable controller design evaluating closed-loop interaction
Univ. of Electronic Sci. and Tech. of
by iterative LMI optimization using frequency response data Weilaing Ge
China, China
Shogo Shinoda Mie Univ., Japan
Wei He Univ. of Sci. and Tech. Beijing, China
Kazuhiro Yubai Mie Univ., Japan
Yunchuan Li Univ. of Maryland College Park, USA
Daisuke Yashiro Mie Univ., Japan
Chenguang Yang Plymouth Univ., UK
Junji Hirai Mie Univ., Japan

14:00 14:20 SunM02-04

13:40 14:00 SunM01-03
Precise tracking control for Piezo-actuated stage using inverse
Flow disturbance suppression for a pneumatic vibration isolator
compensation and model predictive control
using a central pattern generator
Nguyen Manh Linh Shibaura Institute of Tech., Japan
Yukinori Nakamura Tokyo Univ. of Agri. and Tech., Japan
Tran Vu Minh Shibaura Institute of Tech., Japan
Daishi Funaki Tokyo Univ. of Agri. and Tech., Japan
Xinkai Chen Shibaura Institute of Tech., Japan
Mami Kimura Tokyo Univ. of Agri. and Tech., Japan
Shinji Wakui Tokyo Univ. of Agri. and Tech., Japan
14:20 14:40 SunM02-05
Oscillatory behavior based fault feature extraction for bearing
14:00 14:20 SunM01-04
fault diagnosis
Design of a fault tolerant control system using
Juanjuan Shi Univ. of Ottawa, Canada
Mahalanobis-Taguchi(MT) system
Ming Liang Univ. of Ottawa, Canada
Shin Wakitani Tokyo Univ. of Agri. and Tech., Japan
Mingcong Deng Tokyo Univ. of Agri. and Tech., Japan SunM03 Session Room No.03

14:20 14:40 SunM01-05 Complex systems control and applications I

Data-driven PID gain tuning for unknown impulse disturbance Regular Session
attenuation Chair: Ni Bu Qingdao Univ. of Sci. and Tech., China

Shiro Masuda Tokyo Metropolitan Univ., Japan Co-Chair: Zhengfeng Ming Xidan Univ., China

SunM02 Session Room No.02 13:00 13:20 SunM03-01

Switched reluctance motor/generator simulation research based
 on compressed air energy storage system Songming Cao Beijing Univ. of Posts and Telecommunications, China

Xiaoshu Zan China Univ. of Mining and Tech., China Yong Liu Beijing Univ. of Posts and Telecommunications, China

Hong Zhu China Univ. of Mining and Tech., China Ming Hu Beijing Univ. of Posts and Telecommunications, China

Xin Fu Beijing Univ. of Posts and Telecommunications, China

13:20 13:40 SunM03-02

Surface flaws detection algorithms for large aperture optical 15:10 15:30 SunP01-02
element Precision size measurement instrument for irregular objects
Zhengtao Zhang Chinese Acad. of Sci., China Wanji Liu SIPO , China
De Xu Chinese Acad. of Sci., China Xian Tao Chinese Acad. of Sci., China
Xian Tao Chinese Acad. of Sci., China

Feng Zhang Chinese Acad. of Sci., China 15:30 15:50 SunP01-03

A new method of periodicity estimation for mechanical acoustic
13:40 14:00 SunM03-03 data
An adaptive observer-based fault detection and diagnosis for Zhipan Hong Shandong Univ., China
nonlinear systems with sensor and actuator faults Guoliang Lu Shandong Univ., China
Xiaomo Yan Univ. of Manchester, UK
Bailing Tian Tianjing Univ., China 15:50 16:10 SunP01-04
Hong Wang Univ. of Manchester, UK Acoustic emission signal harr wavelet denoise research
Kenan Shen Southwest Forestry Univ., China
14:00 14:20 SunM03-04 Mingcong Deng Tokyo Univ. of Agri. and Tech., Japan
Minimum mutual information control for multi-variable Ming Li Southwest Forestry Univ., China
non-Gaussian stochastic systems
Qichun Zhang Univ. of Manchester, UK 16:10 16:30 SunP01-05
Aiping Wang Anhui Univ., China Operator-based nonlinear networked control system with packet
Hong Wang Univ. of Manchester, UK error
14:20 14:40 SunM03-05 Tomohito Hanawa Tokyo Univ. of Agri. and Tech., Japan
Design and simulation of dual fuzzy self-adjusting immune PID Mingcong Deng Tokyo Univ. of Agri. and Tech., Japan
controller for grain drying system
Aini Dai Beijing Univ. of Posts and Telecommunications, China 16:30 16:50 SunP01-06
Xiaoguang Zhou Beijing Univ. of Posts and Telecommunications, China Optimal control of non-prehensile manipulation control by two
Chi Zhang Beijing Univ. of Posts and Telecommunications, China cooperative arms
Xiangdong Liu Beijing Univ. of Posts and Telecommunications, China Changan Jiang Ritsumeikan Univ., Japan
Huiling Zhou Beijing Univ. of Posts and Telecommunications, China Satoshi Ueno Ritsumeikan Univ., Japan
Yoshikazu Hayakawa Nagoya Univ., Japan
14:40 - 14:50 Break

SunP02 Session Room No.02

Middle Afternoon Technical Sessions
14:50-16:50, Sunday, August 23, 2015
Mechatronic systems analysis and control II
SunP01 Session Room No.01 Regular Session
Chair: Shengjun Wen Zhongyuan Univ. of Tech., China

Control of mechatronic systems II Co-Chair: Dongyun Wang Zhongyuan Univ. of Tech., China
Regular Session
14:50 15:10 SunP02-01
Chair: Changan Jiang Ritsumeikan Univ., Japan
SVR-based input-output mapping of a micro-hand
Co-Chair: Junming Xiao Zhongyan Univ. of Tech., China
Kou Fujita Tokyo Univ. of Agri. and Tech., Japan
14:50 15:10 SunP01-01 Shuichi Wakimoto Okayama Univ., Japan
A brushless DC motor speed control system based on DSP Mingcong Deng Tokyo Univ. of Agri. and Tech., Japan

controller Shin Wakitani Tokyo Univ. of Agri. and Tech., Japan

 15:10 15:30 SunP02-02 Yong Liu Beijing Univ. of Posts and Telecommunications, China

Nonlinear remote temperature control of a spiral plate heat Ming Hu Beijing Univ. of Posts and Telecommunications, China

exchanger Songming Cao Beijing Univ. of Posts and Telecommunications, China

Ryohei Fujii Tokyo Univ. of Agri. and Tech., Japan

Mingcong Deng Tokyo Univ. of Agri. and Tech., Japan 15:10 15:30 SunP03-02
Shin Wakitani Tokyo Univ. of Agri. and Tech., Japan IPMC robust nonlinear tracking control design based on a
multi-objective particle swarm optimization-based RRCF
15:30 15:50 SunP02-03 approach
Operator based voltage control experiments of thermoelectric Wudai Liao Zhongyuan Univ. of Tech., China
generation system Tongbin Yan Zhongyuan Univ. of Tech., China
Ryuji Yoshida Tokyo Univ. of Agri. and Tech., Japan Aihui Wang Zhongyuan Univ. of Tech., China
Mingcong Deng Tokyo Univ. of Agri. and Tech., Japan Yiwen Fu Zhongyuan Univ. of Tech., China
Shin Wakitani Tokyo Univ. of Agri. and Tech., Japan
15:30 15:50 SunP03-03
15:50 16:10 SunP02-04 Robot arm with micro-hand robust control design using
Robust controller design for networked control system with operator-based robust right coprime factorization approach
packet dropout and quantization error Aihui Wang Zhongyuan Univ. of Tech., China
Haiquan Wang Zhongyuan Univ. of Tech., China Hongnian Yu Bournemouth Univ., UK
Rongxiang Gao Zhongyuan Univ. of Tech., China Dongyun Wang Zhongyuan Univ. of Tech., China
Qian Zhang Zhongyuan Univ. of Tech., China Mingcong Deng Tokyo Univ. of Agri. and Tech., Japan
Shengjun Wen Zhongyuan Univ. of Tech., China
16:10 16:30 SunP02-05 15:50 16:10 SunP03-04
Active power filter design for improving power quality Complex event processing on uncertain data streams in product
Junming Xiao Zhongyuan Univ. of Tech., China manufacturing process
Xiangming Zhang Zhongyuan Univ. of Tech., China Na Mao Chinese Acad. of Sci., China
Shengjun Wen Zhongyuan Univ. of Tech., China Jie Tan Chinese Acad. of Sci., China
Zhengbo Liu Zhongyuan Univ. of Tech., China
16:10 16:30 SunP03-05
16:30 16:50 SunP02-06 Detecting cracks on a concrete surface using histogram of
Fractional PID based stability control for a single link rotary oriented gradients
inverted pendulum Lin Meng Ritsumeikan Univ., Japan
Wudai Liao Zhongyuan Univ. of Tech., China Zhongkui Wang Ritsumeikan Univ., Japan
Zhengbo Liu Zhongyuan Univ. of Tech., China Yoshiyuki Fujikawa Ritsumeikan Univ., Japan
Shengjun Wen Zhongyuan Univ. of Tech., China Shigeru Oyanagi Ritsumeikan Univ., Japan
Shuhui Bi Univ. of Jinan, China
Dongyun Wang Zhongyuan Univ. of Tech., China 16:30 16:50 SunP03-06
Power balance control for switched reluctance generator
integrated in DC microgrid
SunP03 Session Room No.03
Lanzhou Institute of Physics and National Key Lab. of
T. Wang
Vacuum and Cryogenic, China
Complex systems control and applications II
Lanzhou Institute of Physics and National Key Lab. of
Regular Session D. He
Vacuum and Cryogenic, China
Chair: Aihui Wang Zhongyuan Univ. of Tech.., China
Q. Wang China Univ. of Mining and Tech., China
Co-Chair: Hong Wang Univ. of Manchester, UK
Lanzhou Institute of Physics and National Key Lab. of
Y. Huo
14:50 15:10 SunP03-01 Vacuum and Cryogenic, China

Design of the automatic sample preparation system L. Lu China Univ. of Mining and Tech., China
Xin Fu Beijing Univ. of Posts and Telecommunications, China J. Li China Univ. of Mining and Tech., China
Shimin Wei Beijing Univ. of Posts and Telecommunications, China
16:50 - 17:00 Break

17:00 - 18:00
Plenary Panel Discussion
New Developments in Advanced Mechatronic Systems

Speaker: Mingcong Deng Tokyo Univ. of Agri. and Tech., Japan

Hongnian Yu Bournemouth Univ., UK

18:30 - 20:30
Banquet

(Beijing Fujian Hotel, Beijing)

End of the 3rd Day

ICAMechS 2015
Monday, August 24, 2015

Technical Tour to The Great Wall of China

End of the Conference

 ABSTRACT
Organized Session
Renewable sources of energy
Time: 13:00-14:40, Saturday, August 22, 2015 Chair: Zongxiao Yang (Henan Univ. of Sci. and Tech., China)
SatM01, Room No. 01 Co-Chair: Aihui Wang (Zhongyuan Univ. of Tech., China)

SatM01-01 SatM01-02

Cluster analysis on the service oriented alliance enterprise Performance and exhaust emissions of a turbocharged common
manufacturing resource rail DI diesel engine fueled with alcohol
Yunxia Wang, Shenghai Qiu, Chengchong Gao, and Jun Wei Bin Xu, Yibin Liu, Jian Wu, Chuang Li, Weiwei Shang, and Zhihao Ma
(Nanjing Institute of Tech., China) (Henan Univ. of Sci. and Tech., China)

To alliance manufacturing enterprises, there are some problems in resources such Three kinds of alcohol fuels and diesel were blended and combusted in an
as different kinds, ununiform standards, high repeatability etc. Under cloud electronic control high-pressure common rail turbocharged DI diesel engine
manufacturing service platform, using cluster analysis method to classify respectively. Without any modifications in the original engine structure, the
intelligent resources of alliance enterprise was studied. The same or similar combustion characteristics and emissions of alcohol-diesel fuel were investigated.
resources are divided into one resources cluster. The uniqueness attribution of It shows that the peak heat release rate and the brake specific fuel consumption
resources is solved. According to the cloud user needs, resources in a cluster not in increase with the increase of mixing ratio. At the same blending ratio, the peak
all resources set are searched and matched. Using the evaluation method of heat release rate of methanol-diesel is the highest, but the brake specific fuel
manufacturing dynamic capacity, optimal resources are provided. It can improve consumption is lower; N-butanol/diesel has the lowest CO and HC emissions but
using ratio of the resources. Finally an instance was verified. has the highest NOX emissions. NOX emissions of methanol-diesel and
ethanol-diesel are approximated, but CO and HC emissions of ethanol-diesel is the
highest. In consideration of the fuel economy and emission performance,
methanol-diesel mixing fuel is better.

SatM01-03 SatM01-04

Design and simulation of passive filter based on fundamental Research on large supermarket fresh food supplier evaluation and
magnetic flux compensation selection based on SWOT-entropy weight fuzzy comprehensive
model
Qun Chen and Zongxiao Yang
(Henan Univ. of Sci. and Tech., China) Ying Yuan, Yi Zhang, Long Di, Ganggang Wu, and Zongxiao Yang
(Henan Univ. of Sci. and Tech., China)
The harmonic problem becomes more and more serious in power grid, passive Aiming at the subjectivity and random problems existing in traditional evaluation
filter is one of the effective way to filter harmonic. This paper presented the design and selection processes of supermarket fresh goods suppliers, and combined with
and simulation of passive filter based on the theory of fundamental magnetic flux the perishable and vulnerable characteristics and the current situation and
compensation (FMFC). The primary winding of transformer is in series between development trend of supermarket fresh food, the paper constructed the ladder
the passive filter and the power supply, in the primary winding of series type evaluation framework by using SWOT analysis method, which is used to
transformer, the fundamental current component of system was detected and study the external environment and capacity of the suppliers of supermarket fresh
tracked to produce a fundamental compensation current by inverter, which was goods. In view of the deficiencies of common comprehensive evaluation methods
injected into the secondary winding, when the injected fundamental current and on weight determination, therefore, this study proposed a structured methodology
the fundamental current component of the system satisfied the FMFC, the system for supplier selection and evaluation based on the Entropy Weight Fuzzy
fundamental current could pass transformer successfully, but most of the harmonic Comprehensive Evaluation. In the end, the method was applied on the course of
current was forced to flow into passive filter, thus achieved the purpose of the fresh goods suppliers optimization in one supermarket to validate the
harmonic isolation, and made most of the harmonic current flow into passive feasibility and correctnessexample.
filter. Compared with the using of passive filter singly, it improved the efficiency
of passive filter. The validity of the new method of filtering characteristics was
verified by simulation through Matlab. The simulation results show that this
method can filter out harmonic more efficiency in power systems when the
waveform distortion is serious.

SatM01-05
A novel outer rotor axial primary magnetic circuit permanent
magnet generator
Ximei Li and Zongxiao Yang, and Dongliang Song
(Henan Univ. of Sci. and Tech., China)

Wind energy has attracted persistent attention as a renewable energy, so it is very

significant to research the permanent magnet generator which is an essential part
of wind energy conversion system. This paper proposes a novel outer rotor axial
primary magnetic circuit permanent magnet generator. For the sake of small size
and light weight with extra low speed for direct coupling, the outer rotor structure
is used. The stator windings are adopted double fractional-slot concentrated to
reduce the cogging torque. The most highlight is a novel axial primary magnetic
circuit. Through the air gap magnetic flux density test experiments, there exist the
linear relationship of the negative slope between the magnetic flux density and the
air gap. This novel type of magnetic circuit structure makes greatly improved the
sensitivity gap, so that the key Tech. such as structure design process route
determined and manufacturing become easier to achieve, significantly reduce
generator costs, ease of construction practical application. Meanwhile, the novel
outer rotor axial primary magnetic circuit permanent magnet generator can
overcome the problems of the lager overall start reluctance generator torque and
the lower wind energy conversion rate under the condition of small wind.
 Organized Session
Advanced intelligent control in robotics and mechatronics
Time: 13:00-14:40, Saturday, August 22, 2015 Chair: Luige Vladareanu (Romanian Acad., Romania)
SatM02, Room No. 02 Co-Chair: Zengguang Hou (Chinese Acad. of Sci., China)

SatM02-01 SatM02-02

Active training research of a lower limb rehabilitation robot based Fuzzy controller design of the wire feeder of invasive vascular
on constrained trajectory interventional surgery robot
Hongbo Wang, Dong Zhang, Hao Lu, Yongfei Feng, and Peng Xu (Yanshan Univ., China) Hongbo Wang, Shuaishuai Wang, Wen Zhang, and Liyu Xie (Yanshan Univ., China)
Razvan-Viorel Mihai and Luige Vladareanu (Romanian Acad., Romania) Xue Yang (Univ. of Jinan, China)
Luige Vladareanu and Radu Munteanu (Romanian Acad., Romania)

In order to help patients achieve effective and coordinating multi-joints movement A fuzzy controller is analyzed for the wire feeder of the invasive vascular
easily, the paper presents an active training method of a lower limb rehabilitation interventional surgery robot in this paper. First of all, the mechanism of the wire
robot based on constrained trajectory. The realization process of the method is as feeder is introduced, and the axial motion is analyzed. Second, the dynamic model
follows. The patients movement intention is recognized by analyzing the signals of the wire feeder is established. Next, the fuzzy PID controller is analyzed to
from torque sensors installed in rehabilitation robot joints, and the moving improve the positioning accuracy. At last, the controller is simulated by
velocity of the mechanical arm terminal is calculated through the establishment of MATLAB. Comparing with normal PID controller, the fuzzy PID controller has
the damping function and the design of constrained trajectory. The velocity of better dynamic properties and anti-jamming ability. The result verifies the
joints can be calculated through kinematic Jacobi matrix, and is used to drive the effectiveness of the controller strategy.
joints. The moving velocity of the mechanical arm terminal on a constrained
trajectory can be controlled through adjusting the damping function, in order to
achieve the active rehabilitation training for different patients at different levels.
Finally, the feasibility of the method is validated by the experiment.

SatM02-03 SatM02-04

Advanced intelligent walking robot control through sliding motion Versatile intelligent portable robot platform for flexible robotic
control and bond graphs methods cells with AGV
Ionel Alexandru Gal and Luige Vladareanu (Romanian Acad., Romania) Luige Vladareanu, Cristian Spirleanu, and Mihaiela Iliescu
(Romanian Acad., Romania)
Hongbo Wang (Yanshan Univ., China)
Hongnian Yu (Bournemouth Univ., UK)
Hongnian Yu (Bournemouth Univ., UK) Mingcong Deng (Tokyo Univ. of Agri. and Tech., Japan)
Mingcong Deng (Tokyo Univ. of Agri. and Tech., Japan) Weizhong Guo and Feng Gao (Shanghai Jiao Tong Univ., China)

In this paper, we tested a dynamic control law on a 2DOF robot walking leg. To The paper studies the flexible robotic cells in cooperation with automated guided
achieve this, we modeled the system using bond graphs, and the Sliding Control vehicle (AGV), in the presence of obstacles, at constant or variable speed and
Method for real time control. To increase performance of the robot trajectory variable load, aiming to optimizing the interaction between AGV and flexible
tracking control a fuzzy gain adjustment inside the dynamic control method has robotic cell components. Overall system performance is analyzed by using
been used and the new results include the systems influence on the sliding modeling tools for discrete event systems like Generalized Stochastic Petri Net
parameter. The research results lead to improve of the walking robot movement (GSPN). The interaction between AGV and flexible robotic cell components is
control on unstructured and bumped surfaces. implemented through communication messages using serial data received from an
optical XY encoder, communication protocol receive function is modeled with
GSPN. Improving of the stability performances and real time motion control are
analyzed and the virtual projection method is adopted using the Versatile
Intelligent Portable Robot Platform VIPRO. The obtained results, validated on the
experimental RTOS robotic platform and DMQX language extension for robotic
applications, lead to higher performance in relation to interaction optimization,
decreased the flexible cells cycle time, increased mobility and stability of the
AGV and also the development of new technological capabilities of the control
systems.

SatM02-05
Haptic intelligent interfaces for NAO robot hand control
Octavian Melinte, Luige Vladareanu and Ligia Munteanu
(Romanian Acad., Romania)
Hongnian Yu and Shuang Cang (Bournemouth Univ., UK)
Zengguang Hou and Guibin Bian (Chinese Acad. of Sci., China)
Hongbo Wang (Yanshan Univ. , China)

The paper presents a new approach on haptic interface control for NAO robotic
hand. The haptic teleoperation of NAO robots hand raises some issues, the most
important being mapping the haptic device dynamic for NAO robot hand. The
mapping was achieved by determining a dynamic gain of robot haptic feedback
through Neural Network. The results lead to the achievement of the haptic
intelligent interfaces for the NAO robot hand control which can be integrated into
an innovative haptic robot control system.
 Regular Session
Optimization methods and application I
Time: 13:00-14:40, Saturday, August 22, 2015 Chair: Shuhui Bi (Univ. of Jinan, China)
SatM03, Room No. 03 Co-Chair: Huimin Xiao (Henan Univ. of Economics and Law, China)

SatM03-01 SatM03-02
Weak reachability of probabilistic boolean control networks Oscillation criteria for a class of nonlinear impulsive parabolic
system under dirichlet boundary condition
Zhiqiang Li and Huimin Xiao
(Henan Univ. of Economics and Law, China) Junhong Liu, Wenxue Liu, Lifeng Li, and Qi Jin
(Acad. of Armored Force Eng., China)

This paper concerns the weak reachability of probabilistic Boolean control In this paper, the authors discuss oscillation of non-zero solutions for a class of
networks (PBCNs). Using semi-tensor product, PBCNs are expressed as discrete impulsive parabolic system. Several oscillation criterias are obtained under
time bilinear systems with respect to states and controls. From the algebraic form Dirichlet boundary condition by using the Green formula and Jensen inequality.
of PBCNs, the maximum transfer probability matrix for PBCNs is constructed.
From the matrix, necessary and sufficient condition for weak reachability of
PBCNs is given. Also, a numerical example is presented to show the applicability
of our approach.

SatM03-03 SatM03-04
Operator-based left coprime factorization for nonlinear systems Passivity-based robust control for uncertain nonlinear feedback
systems
Fazhan Tao and Mingcong Deng
(Tokyo Univ. of Agri. and Tech., Japan) Ni Bu (Qingdao Univ. of Sci. and Tech., China)
Mingcong Deng (Tokyo Univ. of Agri. and Tech., Japan)

In this paper, a class of nonlinear systems are considered with the left factorization This paper is concerned with passivity-based robust control and tracking
approach. More specially, the stability of the nonlinear systems are guaranteed performance for nonlinear feedback system with unknown perturbations. By
using the proposed design scheme that links the stability to the Bezout identity. At designing the feedforward controller in smaller norm, not only the nonlinear
the same time, through the proposed scheme, the left coprime factorization of the system with nominal plant but also that with perturbed plant can be stabilized,
nonlinear systems is verified. Based on the proposed design scheme, a sufficient moreover, the plant output can be proved to be closer to the reference output. The
condtion is found to establish a relationship between the left coprime factorization effectiveness of the proposed is confirmed by the simulation results.
and the right coprime factorization. Finally, a numerical simulation is given to
confirm the effectiveness of the proposed method.

SatM03-05

Correlated neural activity in spiking networks with topographic

couplings
Jinli Xie, Qinjun Zhao, and Jianyu Zhao (Univ. of Jinan, China)

A neural field model with topographic feedforward and feedback is chosen to

confirm how correlated activity in spiking neural networks depends on spatial
couplings. Numerical simulations reveal that the intensity of the correlated firing
is suppressed as the feedback spatial scale decreases. Afterwards, as the feedback
becomes topographic, the correlation coefficient is almost unchanged with further
decreases in feedback spatial scale. Correspondingly, the relatively flat values of
correlation coefficient with decreasing feedforward spatial scale imply that the
effect of spatial spread of feedforward on correlated firing of the neurons keeps
almost invariable. Therefore, the correlation coefficient and the spatial scales are
uncorrelated in the network with topographic couplings. In brief, the global
feedback enables the system to modulate correlated neural activity with the spatial
scale, while the introduction of topography in couplings brings little effect on
network correlations.
 Regular Session
Optimization methods and applications II
Time: 14:50-16:30, Saturday, August 22, 2015 Chair: Guangdong Tian (Northeast Forestry Univ., China)
SatP01, Room No. 01 Co-Chair: Zongxiao Yang (Henan Univ. of Sci. and Tech., China)

SatP01-01 SatP01-02

Scheduling of rescue vehicles to forest fires via multi-objective An effective VNSSA algorithm for the blocking flowshop
particle swarm optimization scheduling problem with makespan minimization
Yaping Ren and Guangdong Tian (Northeast Forestry Univ., China) Chaoyong Zhang, Zhanpeng Xie, and Xinyu Shao
Mengchu Zhou (New Jersey Institute of Tech., USA) (Huazhong Univ. of Sci. and Tech., China)
Guangdong Tian (Northeast Forestry Univ., China)
It is complex and difficult to perform the rescue vehicle scheduling to handle This paper proposes a hybrid algorithm based on the variable neighborhood search
forest fires in order to reduce the operational cost and improve the efficiency of (VNS) and the simulated annealing approach (SA) for minimizing the makespan
fire-extinguishing services. A new research issue arises when a) decision-makers in a blocking flowshop scheduling problem. The proposed hybrid algorithm
want to minimize the number of rescue vehicles while minimizing the rescue time; (VNSSA) adopts simulated annealing approach (SA) as the local search method in
and b) decision-makers prefer to complete the fire-extinguishing task given the third stage of variable neighborhood search (VNS), and uses a perturbation
limited vehicle resources. To do so, this work presents a novel multi-objective mechanism consisting of various neighborhood operators to diversify the search.
scheduling model to handle forest fires subject to limited rescue vehicle The destruction and construction phases of iterated greedy algorithm are used to
constraints, in which a fire spread speed factor is introduced into this model to generate a trial solution during the perturbation process of SA to further promote
better describe a practical forest fire. Also, a Multi-objective Particle Swarm the search capability. To validate the performance of the proposed VNSSA
Optimization (MPSO) algorithm is proposed to yield a set of Pareto solutions for algorithm, computational experiments and comparisons were conducted on the
this problem. This approach is applied to a real-world emergency scheduling well-known benchmark problems of Taillard. The computational results of the
problem of the forest fire in Mt. Daxing'anling, China. Both theoretical and proposed algorithm when evaluated against the state-of-art algorithms from the
simulation results demonstrate that the proposed approach is able to quickly literature show good performance, some new best solutions for Taillards
produce Pareto solutions for decision makers. instances are reported for this problem.

SatP01-03 SatP01-04
Simulation modeling of shuttle racking system based on automod An expected value multi-objective optimization model to locate a
vehicle inspection station
Bingfeng Sang and Yukun Liu
(Beijing Univ. of Posts and Telecommunications, China) Guangdong Tian (Northeast Forestry Univ., China)
Mengchu Zhou (New Jersey Institute of Tech., USA)

The paper researches modeling method of shuttle racking system based on Optimally locating a transportation facility and automotive service enterprise is an interesting
Automod. By analyzing the real system, the system model of shuttle racking and important problem. In practice, many related factors, e.g., customer demands, allocations,
system was put forward. A method of simulation modeling of LIFO type and the and locations of customers and facilities, are changing, and thus the problem features with
modeling step for the shuttle racking system are provided. At last, an example of uncertainty. To account for this uncertainty, some researchers have addressed its stochastic time
some warehouse of refined grain verifies the feasibility, the effectiveness, and the and cost issues. A new research issue arises when a) decision-makers want to minimize the
reusability of the method. transportation time of customers while minimizing their transpiration cost when locating a
facility; and b) users prefer to arrive at the destination within the specific time and cost. By
taking a vehicle inspection station as a typical automotive service enterprise example, this work
proposes a novel stochastic multi-objective optimization approach to address it. Moreover, some
regional constraints can greatly influence its solution; while vehicle velocity is an uncertain
variable due to the influence of some unpredictable factors in a location process. This work
builds a practical stochastic expected value multi-objective programming model of its location
with regional constraints and varying velocity. A hybrid algorithm integrating stochastic
simulation and Genetic Algorithms (GA), namely a random weight based multi-objective GA, is
proposed to solve the proposed models. A numerical example is given to illustrate the proposed
models and the effectiveness of the proposed algorithm.

SatP01-05
Experimental study on nonlinear vibration control of an L-shape arm

Yanfeng Wu and Mingcong Deng

(Tokyo Univ. of Agri. and Tech., Japan)
Longguo Jin
(Qingdao Technical College, China)

In this paper, an experimental system is set up to control the vibration of an L-type

arm which is driven by linear motor. Firstly, the vibration dynamics the L-shape
arm is modelled based on Euler-Bernoulli beam theory, the relation between
motor motion and arm vibration is investigated. Secondly, based on nonlinear
operator control theory, two nonlinear control schemes are designed to control the
motion of the motor and suppress the vibration of arm by using piezoelectric
actuator, respectively. The vibration of arm is considered when controlling the
motor. A tracking compensator is designed to eliminate the hysteresis effect of the
piezoelectric actuator. Finally, by using two piezoelectric sensors to measure the
transverse displacement and one piezoelectric actuator to suppress the vibration of
arm, the experimental system is set up, and some experiments are conducted to
verify the effectiveness of the proposed control scheme.
 Regular Session
Advanced nonlinear control in robotics
Time: 14:50-16:30, Saturday, August 22, 2015 Chair: Hongnian Yu (Bournemouth Univ., UK)
SatP02, Room No. 02 Co-Chair: Yukinori Nakamura (Tokyo Univ. of Agri. and Tech., Japan)

SatP02-01 SatP02-02

The mechatronic system of eco-dolphin - a fleet of autonomous Extended development of LinuxCNC for control of a delta robot
underwater vehicles
Feng Huo, Geoksoon Hong, and Aunneow Poo
Hong Liu (Embry-Riddle Aeronautical Univ., USA), (National Univ. of Singapore, Singapore)
Xudong Shi (Civil Aviation Univ. of China, China)
Junzhen Shao, Qi Zhou, Stace y Joseph-Ellison, Johnathan Jaworski, and Ci Wen
(Embry-Riddle Aeronautical Univ., USA)

This paper presents the mechatronic system of a fleet of three autonomous In recent years, open architecture motion controllers, including those for CNC
underwater vehicles (AUVs) called Eco-Dolphin. Besides common features of machines and robots, have received much interest and support among the global
robots, the unique challenges for underwater robots such as water proof Tech. and control and automation community. This paper presents work done in extending a
underwater communication are addressed. The Eco-Dolphin fleet is built to serve well-known and supported open-source control software called LinuxCNC for the
both societal and the academic needs specified below. Coastal areas have the control of a Delta robot, a translational parallel mechanism. Key features in the
densest human populations and serve as critical economic and recreational areas. development process are covered and discussed and the final customized system
However, they are sensitive to both human-induced and natural alterations as well based on LinuxCNC described.
as disasters. While robots are widely used to replace humans for dangerous work
in many other fields, environmental monitoring missions in coastal waters still
mostly rely on human divers. To serve the societal need, the Eco-Dolphin fleet
presents a novel prototype of cost effective robots for monitoring coastal
environments. The fleet can collect data cooperatively with minimal supervision
from the users on a nearby boat. To contribute to academic research, the fleet is
used to test swarming algorithms for robot formation in 3D space. To contribute to
Sci. education, the Eco-Dolphin project offers a platform for students to gain
hands-on research experiences and an incubator to find first-hand scientific
research problems.

SatP02-03 SatP02-04

Minimizing energy consumption in a U-shaped robotic assembly Charged system search algorithm for robotic drill path
line optimization
J. Mukund Nilakantan and S.G. Ponnambalam G. Kanagaraj (Thiagarajar College of Eng., India)
(Monash Univ. Malaysia, Malaysia) S.G. Ponnambalam (Monash Univ. Malaysia, Malaysia)
George Q. Huang (The Univ. of Hong Kong, China) C.K. Loo (Univ. of Malaya, Malaysia)
Manufacturing industries these days are more concerned about the energy This paper presents a Charged System Search (CSS) algorithm to solve for robotic
consumption due to the increasing energy cost. It is observed that industries can drill path optimization involved in printed circuit board (PCB) manufacturing
achieve significant savings in the energy cost even when a small percentage of industries. Most of the operational time of a PCB Robotic Drill are spent on
energy consumption is reduced. In a manufacturing industry, assembly lines are moving the drill bit between the holes. An optimized path translates to a minimal
equipped with robots to perform the assembly tasks and there is a need for cost of operating the robot. The drill path consists of a number of potential
efficiently balancing the assembly line by reducing the energy consumption during locations where the holes are going to be drilled. As the number of holes required
the production phase. The authors could not find any research on optimizing increases so thus does the complexity to find the optimized path. Recently
energy consumption in U-shaped robotic assembly line systems to date. The developed CSS algorithm proposed to solve this complex problem with minimal
objective of this paper is to propose an optimization models with an objective of computational time. The performance of CSS algorithm is tested and verified with
minimizing the energy consumption in a U-shaped robotic assembly line when the four case studies from the literature. The computational experience conducted in
number of workstations are fixed. Since the assembly line balancing problems this research indicates that the proposed algorithm is capable to efficiently find the
falls under the category of NP-hard, an efficient bio inspired algorithm is optimal path for PCB holes drilling process with reasonable computational time.
implemented to solve the problem. The model proposed is very well applicable for
automobile assembly line where robots are employed. The energy consumption is
evaluated on the benchmark data and energy consumption obtained by the
U-shaped robotic assembly line is compared with the energy consumption
obtained with straight robotic assembly line. The cycle time of the layouts are also
compared. The results obtained using this proposed model can be used for the
decision making process in the industry. The paper also proposes an optimization
model for minimizing the cycle time. The energy consumption of the assembly
line when the cycle time is minimized is evaluated and presented.
SatP02-05

Implementation of remotely controllable mobile robot system over

mobile Ad-hoc network
Mohammad Shahidul Hasan (Staffordshire Univ., UK)
Hongnian Yu (Bournemouth Univ., UK)

Remotely controllable wireless mobile robots are being used in many applications
now days. Such systems can be restricted by the range of wireless signal. Mobile
Ad-hoc Networks offer a number of advantages e.g. multi hop communication i.e.
longer range; can be deployed without any infrastructure etc. This paper presents
the implementation details of a wireless mobile robot system over Mobile Ad-hoc
Network. It consists of a target mobile robot and a relay / observer robot. The
target robot is equipped with a robotic arm to execute various tasks. The relay /
observer robot can route network packets between the controller and the target
robot. It also produces visual feedback of the target robot to the user at the
controlling end.
 Regular Session
Control of mechatronic systems I
Time: 14:50-16:30, Saturday, August 22, 2015 Chair: Shinji Wakui (Tokyo Univ. of Agri. and Tech., Japan)
SatP03, Room No. 03 Co-Chair: Kazuyuki Ito (Hosei Univ., Japan)

SatP03-01 SatP03-02

An interpretation of unbalance vibration compensator for Vibration suppression of galvano mirror considering whirling of
five-axes active magnetic bearing systems based on internal model shaft
principle
Keisuke Nakade, Taiki Nakamura, Shinji Wakui, and Yukinori Nakamura
Taiki Nakamura, Shinji Wakui, and Yukinori Nakamura (Tokyo Univ. of Agri. and Tech., Japan)
(Tokyo Univ. of Agri. and Tech., Japan)
Five-axes active magnetic bearing (AMB) systems are widely utilized for the field Main movement of a galvano mirror is the rolling motion. However, since the
of precision such as semiconductor exposure machines. This paper considers an structure of galvano mirror is cantilever, this mirror is vertically vibrated to mirror
interpretation of an unbalance vibration compensator PGC for five-axes AMB. As reflective surface. This paper considers the suppression of the above vibration,
methods for the rejection of periodic disturbance, repetitive control and PIS which is called pitching vibration. Although the galvano mirror is equipped with
control, which are based on internal model principle, have been known. The PIS an encoder to detect the roll angle of mirror, the pitching vibration of mirror
control can be considered as a special case of the repetitive control. This paper cannot be detected. To overcome this problem, the pitching vibration is verified by
investigates the relationship between the PGC and the PIS control from the the frequency response and the time response using two dimensional position
similarity of the pole assignment and shows that the PGC suppresses the sensitive detector. Moreover, three rubbers with different hardness are inserted on
unbalance vibration from the perspective of the internal model principle. the basis of a structure of the motor shaft in order to suppress the vibration.

SatP03-03 SatP03-04

A consideration of parameter identification of a linear stage using Development of a torque-sensorless viscometer for
particle swarm optimization food processing applications
Marino Watanabe, Yukinori Nakamura, and Shinji Wakui Seiji Hashimoto, Yuqi Jiang, and Amat Basari (Gunma Univ., Japan)
(Tokyo Univ. of Agri. and Tech., Japan) Wei Jiang (Yangzhou Univ., China)

This paper considers the model-based controller in order to achieve the high speed Demands for automatic food processors are growing rapidly year by year. In this
and high precision positioning of a linear stage. In the controller design, particle paper, a novel viscosity measurement method which can provide real-time
swarm optimization (PSO) is employed for the parameter identification of the monitoring in food processing application has been proposed. Instead of using a
linear stage. Particles of the PSO are updated so that the difference between torque sensor for the viscosity measurement, the proposed method applies
simulation and experimental results becomes small. After the parameter disturbance observer-based estimation technique to the viscometer. The
identification, the feedforward controller whose parameters correspond to the effectiveness of the proposed method is verified through simulations and
identified model of the linear stage is designed. Effectiveness of the proposed experiments.
method is shown by simulation and experiment.

SatP03-05

Study on detection of earthquake using DI value and its application

to switching control for air type anti-vibration apparatus
Masayuki Kubota, Hirokazu Mineo, Shinji Wakui, and Yukinori Nakamura
(Tokyo Univ. of Agri. and Tech., Japan)

In semiconductor manufacturing, air type-vibration apparatuses (AVAs) are

widely employed.Sheng
WhenLi (Qingyang Chemical
an earthquake Industry
occurs, Co., have
the AVAs China)to be stopped in
order to avoid serious damage to semiconductor exposure machines. Spectrum
intensity (SI) value is the metric effective for the detection of earthquake though it
takes time to calculate this value. To avoid long computation time, this paper uses
destructive intensity (DI) value instead of SI value. Simulation demonstrates that
DI value, whose computation time is short, can be applied to switching control for
AVAs.
 Invited Session
New developments on ESP
Time: 16:40-18:40, Saturday, August 22, 2015 Chair: Jie Shi (Univ. of Electro-Communications, Japan)
SatE01, Room No. 01 Co-Chair: Yan Yu (Univ. of Electro-Communications, Japan)

SatE01-01 SatE01-02

An overview of supplementary teaching materials for advanced Profiling Japanese EST students vocabulary ability using the new
EST students in Japan general service list test (NGSLT) and the new academic word list
test (NAWLT)
Yan Yu and Jie Shi (Univ. of Electro-Communications, Japan)
Shinichi Hashimoto, Yan Yu, Jie Shi, and Haiyan Zhan
(Univ. of Electro-Communications, Japan)
EST (English for Science and Technology) is an important area of ESP in Japan. Understanding the vocabulary profiles of students is an important point for improving classroom-based pedagogy as
The success of EST courses for science and technology/engineering disciplines is well as materials development based on students abilities and needs in ESP (English for Specific Purposes). As the
partially dependent on the choices of teaching materials. As shown in previous second phase of a longitudinal research to assess Japanese university students vocabulary competency, the
research done by the same presenters of this study in ICAMechs 2014, the researchers collected data on the performance of second and third-year students majoring in science and
quantity of teaching materials for lower and intermediate levels of EST courses in engineering on the New General Service List Test (NGSLT) and the New Academic Word List Test (NAWLT)
Japan is considered to be satisfactory though the concerns over the quality and developed by Phil Bennett and Tim Stoeckel in 2015 using words from the New General Service List 1.01 (NGSL
variations of technical disciplines still exist. This research is a continuation of the 1.01) and the New Academic Word List (NAWL) developed by Charles Browne, Brent Culligan and Joseph
investigation of teaching materials for EST courses in Japan with a focus on the Phillips in 2014. The NGSL 1.01 consists of 2,801 of the most frequently-appearing words in general use, and is a
supplementary or reference materials for advanced students of EST at Japanese revision of the original General Service List (about 2,000 words) which was compiled by Michael West in 1953.
universities. This study aims to conduct an overview of the market available The NWAL consists of 963 frequently-appearing words in academic text, and is a revision of the original Academic
reference books for advanced EST students at Japanese universities. The Word List (570 words) which was compiled by Averil Coxhead in 2000. Both the NGSLT and NAWLT are
categories of the overview include languages used, publication information, main diagnostic tests of written receptive knowledge of the New General Service List and the New Academic Word List
content themes/topics and/or skills, target audience/level of English, and length. respectively. The NGSLT and NAWLT use the same specifications as the Vocabulary Size Test (VST) which was
developed by Paul Nation and David Beglar in 2007, and have relatively good test reliability. This paper introduces
the background about the tests, the participants in the study, and the data collection method. It also interprets the
results of the collected scores, and discusses the implications of the results between departments and years. In
future studies, the researchers hope to find ways to perceive the vocabulary abilities of students in their particular
ESP fields. By doing so, they can be better informed providers of instruction and create materials tailored to
students particular needs.

SatE01-03 SatE01-04

Peer evaluation of academic presentations by Japanese graduate Teaching poster by using genre approach to science and
students: qualitative evaluation technology students of a Japanese university
Jie Shi and Cross John Rajagopalan Maheswari Uma
(Univ. of Electro-Communications, Japan) (Toyo Univ., Univ. of Electro-Communications, Japan)
Jie Shi (Univ. of Electro-Communications, Japan)
This study reports an investigation on the qualitative data of Japanese graduate Science and technology students require the specialized skill of making poster and the explicit
students evaluation on peers academic presentation in English which is designed teaching of which is lacking in the curricula of general Japanese English tertiary education. This
to be part of a collaborative learning project. A total of 772 peer comments was paper considers posters as an expanded form of abstracts and reports teaching of posters to 20
extracted from 265 peer evaluation on 10 group presentations as the dataset of this third-year students of an undergraduate ESP course at a Japanese university of science and
investigation. Results indicate that students focuses are on the quality of slides engineering. First, students were taught to write an abstract using genre theory which required
including figures, tables and texts, the understanding of the technical content of teaching of the generic moves of an abstract of a research article (RA). Teaching of the abstract
the research presented by peers though students come from various laboratories, involved the explicit teaching of the basic moves of background, purpose, methods, results and
and peers delivery abilities such as eye contact, reading from prepared notes and summary that are needed to construct RA abstract. Following that, students deconstructed
rate/speed of speech. The data also shows that students gave more positive abstracts taken from authentic research articles selected by both the teacher and the student
comments than negative ones and that the suggestion type of comment is the least themselves into moves. Having understood the moves, a democratic choice of several topics
in terms of the quantity. from the experiments done as part of the subject curriculum was chosen for constructing
abstracts. The abstract drafts were peer evaluated with checklists, followed by teachers
comments based on which the students submitted a final draft. Next, with the final draft, the
students expanded the abstract to construct a poster. This task involved writing the contents of
each of the moves in the abstract to detailed content expanding each of the sections,
Introduction, Methods, Results and Summary. Contents of all the sections except for the Results
were prepared in groups of three or four with each member being responsible for each section.
Although group members constructed the contents through discussion, poster making and
presentation were done individually. Though there were some overlaps in the contents, the
current project was found to empower the students with the required poster making and
presentation skills which was also shown from the post survey.

SatE01-05 SatE01-06

Corpus-based analysis of academic RA genre: the Results Understanding the most essential types of words for Physics
sub-genre undergraduate students
Louvign Sbastien and Jie Shi Sonia Sharmin (Univ. of Tsukuba, Japan)
(Univ. of Electro-Communications, Japan)

Research Article (RA) is the kind of academic publication that all researchers The field of English for Specific Purposes (ESP) concerns not only English
must produce constantly but not necessarily have learned systematically in their practitioners or EFL (English as a Foreign Language) but also disciplinary
university education. RA consists of multiple parts as typically summarized to be specification and the teaching faculty of disciplines of English learners, commonly
IMRAD (Introduction, Method, Result and Discussion) which is a considered to known as subject-matter specialists because those specialists could also instruct or
be one of the academic genres. Each of the multiple parts of the RA genre is assist the EFL students in terms of proper usage of science and technological
comprised of one or more sub-genres, which makes the RA genre complex and English. The participation of the specialists in ESP education is a growing area of
challenging, especially for novice EST (English for Science and Technology) both ESP and disciplinary research fields and has become a fast developing area of
researchers. As an effort to decode the RA genre in order to help the novice EST ESP research. Identifying content-area English academic language is essential for
researchers, the researchers of this study analyze the sub-genre of Results using proper instruction of students. This study analyzed a well-known and widely used
corpus-based approach, after they analyzed the Abstract, Introduction and Method physics textbook for undergraduate students of first year, namely, Fundamentals
sub-genres in previous studies. The purpose of this study is threefold: to determine of Physics by David Halliday, Robert Resnick, and Jearl Walker (8th edition), to
the linguistic features of the common moves (structure) in the Results sub-genre understand the words of highest frequency used in this book. The high frequency
of RA using corpus analysis, to compare the linguistic features of Results with words as well as a low frequency words in the textbook were assessed. In this
those of the whole RA, Introduction, and Method. study, investigation of the language functions based on the analysis of the corpus
and identification of the words that construction the functions are the ultimate aim
of this study. The free corpus software, AntWordProfiler, was employed for the
analysis.
 Regular Session
Information processing
Time: 16:40-18:20, Saturday, August 22, 2015 Chair: S.G. Ponnambalam (Monash Univ. Malaysia, Malaysia)
SatE02, Room No. 02 Co-Chair: Ken Nagasaka (Tokyo Univ. of Agri. and Tech., Japan)

SatE02-01 SatE02-02

Modelling a 20MW scale solar farm in an unused angled area Temporal and spatial projection of wind speed based on modular
near Fukushima nuclear power plant netrowk SOM for installation of off-shore wind generation
Amin Mohammadirad and Ken Nagasaka Mitsuharu Hayashi and Ken Nagasaka
(Tokyo Univ. of Agri. and Tech., Japan) (Tokyo Univ. of Agri. and Tech., Japan)

Mega-Solar generation becomes one of the important parts in the global energy supply with current situation of the Wind generation is one of the fast growing and introduced resources among renewable energies
fossil fuels exhaust and replacement of nuclear power energy after the nuclear power plant acceded caused by Great through worldwide including Japan. As Japan, on the other hand, is an island country
East Japan Earthquake and tsunami on March 11,2011. Also, the CO2 emission must be reduced based on the surrounded by ocean, the topography suitable for wind generation is limited for the on-shore.
Kyoto protocol. One of the effective ways to reduce the CO2 emission is to use renewable energy such as wind and Therefore, based on the wind map of up to 2030, it is expected that new introduction suitable for
solar energies. For this reason, how to make optimal allocation of large-scale PV and wind farms with maximum wind power generation will be more on off-shore. For this reason, it is very important to
power become necessary both for the world as well as Japan. In this study we are focusing on Mega-Solar issue for determine the wind characteristics of the candidate area for installing wind generation, however
the flat and angled lands. As different slop directions will affect the Maximum Point Power Tracking (MPPT) of in most cases of off-shore installation, existence of weather condition data is poor and it may
each PV systems. We also consider the MPPT for these angled lands. So far, Mega-Solar farms are installed in flat need lots of time and cost for collecting such weather condition data as a new pin-point
lands. However, recently, there are great needs (almost more than 1000 projects are waiting to run) for introducing information. In this study, the goal of this research is to project a wind speed of an unseen area
Mega-Solar farms not only in flat but also in hills or angled lands. These lands with slope are no longer omitted (where its weather condition data are not available) by mapping the modularized Artificial
from the planning of a Mega-Solar farm. In this paper, we consider on Mega-Solar farm which combines flat and Neural Network (SOM: Self-Organization Map) of seen areas (where their weather condition
angled lands and solve the Maximum Power Point Tracking (MPPT) of PV , cost and environment issues. In this data are available) around the target area. By learning the correlation between modularized
study, for analysis the slope directions and site selection, the Geographical Information System (GIS) technique is ANNs of seen and unseen areas, the result of this temporal and spatial projection will be the
used. Maximum Power Point Tracking (MPPT) using Perturb and Observe (PandO) Mega-Solar components are all prediction of wind speed of target place. It is believed, by the help of the proposed technique, a
designed on MATLAB/SIMULINK environment. One of the unique points of this study is to use the lands located huge amount of time and cost will be saved in the selection of installation point of off-shore
near the area where affected by the radiation. These lands cannot be used for living and agricultural activity, wind power generation. Moreover, it is intended to contribute to the introduction increase in the
however, as for Mega-Solar farm, it is suitable. The targeted site was a golf-course which we are planning to install amount of wind power.
about 20[MW] solar farm.

SatE02-03 SatE02-04

A household load priority scheduling model based on real-time Study on cutting surface stripe and wettability of AZ91D
electricity price magnesium alloy by WEDM-HS
Bo Hu (State Grid Energy Research Institute, China), Jinkai Xu, Kui Xia, Rongxian Qiu, Linshuai Zhang,
Xiao Zhong, and Song Liu Zhanjiang Yu, and Huadong Yu
(North China Electric Power Univ., China) (Changchun Univ. of Sci. and Tech., China)
In this paper, we present a household load priority scheduling model based on Low discharge energy was used to machine AZ91D magnesium alloy using
real-time electricity price, the optimization objective is to satisfy users comfort reciprocating type high speed wire electrical discharge machining (WEDM-HS).
demand and reduce the electricity cost. The constraints are electric power Co.'s Different microstructures were found on bright and dark stripes. Using water
response to electrical and the working time of household appliances that user contact angle of the droplet at the magnesium alloy surface evaluated the
expect. The model divides all appliances into three categories: important wettability. Results show that contact angle value of magnesium alloy surface
appliances which cannot be dispatched, non-interruptible appliances which can be after WEDM-HS was higher than magnesium alloy matrix. There were differences
dispatched and interruptible appliances which can be dispatched, and the third surface microstructures between bright and dark stripes, which is the key reason of
type can be subdivided into demandable appliances and rechargeable appliances. hydrophoby. Compared with the bright stripe, the value of contact angle of dark
The proposed model not only takes various electrical models of appliances into stripe was higher and more stable. On submillimeter groove surface, the value of
account, but also includs users psychological satisfaction. Electricity bill and contact angle was increased to 154.71. This paper provided the facile and
users satisfaction curve will generated by the system, through adjusting the cost-effective process method for support the industrialized fabrications of
proportion of satisfaction, different users can get different power saving and super-hydrophobic AZ91D surface.
comfort effect according to their different demands, it has important practical
significance. Finally, the article verify the effectiveness of the proposed
optimization model by simulations.

SatE02-05

Influence of microstructure on surface wettability

Xinxin Zhang, Jinkai Xu, Zhongxu Lian, Zhanjiang Yu and Huadong Yu
(Changchun Univ. of Sci. and Tech., China)

Wettability is one of the important performance measuring the hydrophobicity of

metal surface. There has important relationship between microstructures and
surface wettability. In this paper, square pillar microarrays with different
geometrical parameters were processed on hydrophilic metal substrates by
high-speed micro-milling Tech., observed superficial morphologies and apparent
contact angles of microstructures, and analyzed the influencing laws of
geometrical parameters on surface wettability. Experimental results showed that
the wetting transition from hydrophilicity state to hydrophobicity state was
successfully achieved by fabricating micron-grade regular array on aluminum
alloy surface, and the hydrophobic property on solid surface is a common coupling
results of micronano-binary structures both microscale pillars and nanoscale pits
and convexities. In this experiment, a good hydrophobic surface with larger
apparent contact angle can be obtained when pillar width is 20 m, pillar spacing
is 100 m, pillar height is 50 m and the droplet volume is 3 l. In addition, the
water droplets on aluminum alloy surface with square pillar microarray is in a
mixed state between Wenzel and Cassie-Baxter, and the surface wettability can be
control by changing the geometry parameters of microstructures.
 Regular Session
Nonlinear systems modeling, control and applications
Time: 16:40-18:20, Saturday, August 22, 2015 Chair: Shiwen Tong (Beijing Union Univ., China)
SatE03, Room No. 03 Co-Chair: Dianwei Qian (North China Electric Power Univ., China)

SatE03-01 SatE03-02

Influences from LCL filter on behaviors of three-phase three-level Sliding mode output tracking control based on a fuzzy clustering
PWM rectifier model
Dongsheng Yu, Hong Zhu, and Ciyan Zheng Shiwen Tong (Beijing Union Univ., China),
(China Univ. of Mining and Tech., China) Dianwei Qian (North China Electric Power Univ., China)
Jianjun Fang (Beijing Union Univ., China)
In this paper, system description of a three-phase Neutral-Point-Clamped (NPC) A sliding mode output tracking control method based on a fuzzy clustering model
rectifier is briefly given by depicting its power circuit and controller. In has been proposed in this paper. This technique describes the controlled system by
consideration of tracking target current and minimizing current ripples, the rulebased fuzzy models. Parameters (antecedent membership functions,
allowable parameters of inductorcapacitorinductor (LCL) filter are calculated. consequent parameters, etc.) in this structure can be fine-tuned with input-output
The filter inductor is then adopted to investigate the influences from LCL filter on data. By means of this technique, the complicated system can be decomposed into
dynamic behaviors of this three-phase three-level rectifier. Based on a number of simpler local models with a characteristic of smooth transitions
proportional-integral control method, the simulation analysis is carried out to between these sub-models by means of overlap membership functions. Then the
explore the operational behavior differences caused by the variation of filter implicit membership functions have been transformed into explicit functions of
inductor. The curves of output variables and their phase trajectories are the antecedent variables with neural networks. Thus the control actions of a
comparatively addressed in terms of different parameters of the LCL filter. The discrete sliding mode controller can be designed. Simulative results demonstrate
simulated results indicate that the filter parameters has significant impacts on the the effectiveness of this method.
operational performance of rectifier system, which implies that suitable control
method is required to eliminate unexpected influences.

SatE03-03 SatE03-04

PID control of air tank temperature system with parameters Sliding-mode-based disturbance rejection control of nonlinear
tuning through network power systems with renewable sources
Shiwen Tong, Yushan Li, Junjie Ren, and Yinong Zhang Hong Liu and Dianwei Qian
(Beijing Union Univ., China) (North China Electric Power Univ., China)

This paper presents an application of Networked Control in the air tank In this paper, an integral sliding mode control (I-SMC) method for load frequency
temperature system. The system is controlled by control (LFC) is investigated in interconnected power systems in the presence of
Proportion-Integration-Differential (PID) algorithm running in the Siemens wind turbines. Generation rate constraint (GRC) has been considered. For
S7-200 PLC. Through an OPC server component, controller parameters implementation, neural networks are adopted to approximate the entire
(proportion, integration, differential) can be remote tuning by Matlab. Thus, uncertainties. The integral sliding mode controllers and the neural networks work
complex control algorithm such as fuzzy inference, expert system and genetic in parallel to achieve disturbance rejection control. It has been observed from
optimization can be utilized. The process is supervised by configuration simulation results that the presented method was able to maintain the feasibility,
software, for example King Views, located in different geographical areas at the validity and robustness.
same time.

SatE03-05

Research and application of predictive control algorithm based on

fuzzy model
Yuan Li, Sen Cong, Shiwen Tong, and JianJun Fang
(Beijing Union Univ., China)

In this paper, predictive control algorithm is described in detail and the ANFIS
(Adaptive neuro fuzzy inference system) is introduced. The fuzzy model is built
by ANFIS. Fuzzy model and predictive control are combined to form a predictive
control based on fuzzy model which is used to complex system with time delay.
For example, as system of three water tanks. First, using matlab software, we
established the model of three-water tank with ANFIS, and verified it by the prior
data. Simulation shows that the model is effective. Then we design the actual
control system of three-water tank in which AC800F is controller. Predictive
control based on fuzzy model is programmed with Function Block language. It is
also demonstrated that, by comparing the operation results of the PID with
predictive control based on fuzzy model, the control method introduced in the
paper is practical and effective, has great value in practical application.
 Poster Session
Time: 10:20-11:30, Sunday, August 23, 2015 Chair: Changan Jiang (Ritsumeikan Univ., Japan)
Poster, Main Conference Hall Xiaoguang Zhou (Beijing Univ. of Posts and Telecommunications, China)
Co-Chair: Shin Wakitani (Tokyo Univ. of Agri. and Tech., Japan)
Co-Chair: Yixin Yin (Univ. of Sci. and Tech. Beijing, China)

Poster-01 Poster-02

Bipolar neutrosophic sets and their application based on Evaluation method of main production resource in society and
multi-criteria decision making problems industrial park of extreme arid region (1)
Irfan Deli (Aralik University, Turkey), Litifu Zulati
Mumtaz Ali (Quaid-e-azam Univ., Pakistan) (Xinjiang Univ. of Finance and Economics, China)
Florentin Smarandache (Univ. of New Mexico, USA)

In this paper, we introduce concept of bipolar neutrosophic set and its some Three kinds of alcohol fuels and diesel were blended and combusted in an
operations. Also, we propose score, certainty and accuracy functions to compare electronic control high-pressure common rail turbocharged DI diesel engine
the bipolar neutrosophic sets. Then, we develop the bipolar neutrosophic weighted respectively. Without any modifications in the original engine structure, the
average operator (Aw) and bipolar neutrosophic weighted geometric operator combustion characteristics and emissions of alcohol-diesel fuel were investigated.
(Gw) to aggregate the bipolar neutrosophic information. Furthermore, based on It shows that the peak heat release rate and the brake specific fuel consumption
the (Aw) and (Gw) operators and the score, certainty and accuracy functions, we increases with the increase of mixing ratio. At the same blending ratio, the peak
develop a bipolar neutrosophic multiple criteria decision-making approach, in heat release rate of methanol-diesel is the highest, but the brake specific fuel
which the evaluation values of alternatives on the attributes take the form of consumption is lower; N-butanol/diesel has the lowest CO and HC emissions but
bipolar neutrosophic numbers to select the most desirable one(s). Finally, a has the highest NOX emissions. NOX emissions of methanol-diesel and
numerical example of the method was given to demonstrate the application and ethanol-diesel are approximated, but CO and HC emissions of ethanol-diesel is the
effectiveness of the developed method. highest. In consideration of the fuel economy and emission performance,
methanol-diesel mixing fuel is better.

Poster-03 Poster-04

Research on control methods of permanent magnet synchronous Comprehensive functions of eco-industrial park in conserving
motor position servo system energy and improving ecology
Fen Yang, Zhengfeng Ming, Tao Wen, and Tong Zhang Litifu Zulati (Xinjiang Univ. of Finance and Economics, China)
(Xidian Univ., China) Ken Nagasaka (Tokyo Univ. of Agri. and Tech., Japan)
Zhanxia Zhu (Northwestern Polytechnical Univ., China)
This paper compares common PI controller with proximate time-optimal This research explained the comprehensive function of eco-industrial park in
servomechanism (PTOS), composite nonlinear feedback (CNF) control law, and a the aspects of intensive utilization of energy and other main resources used in
mode switching control (MSC) scheme .The structure of PI controller is simple production in extreme arid areas, such as Turpan Prefecture of Xinjiang, China, in
and easy to implement .The PTOS control law devotes to fast acceleration and which water resource is seriously lack of. On the other hand, this research
deceleration ,while the CNF control law consists of a linear feedback part for analyzed environmental efficiency formed by eco-industrial. Extended service of
achieving fast response and a nonlinear feedback part for suppressing the eco-industrial parks can be realized by greening area around the parks by using
overshoot caused by the linear part, so as to improve the transient performance in waste water. Three basic functions of eco-industrial park are considered, namely
set-point tracking. The MSC scheme is proposed to achieve fast and precise expanding carrying capacity of limit energy and other main resource of
set-point tracking in servo systems .The control scheme incorporates a CNF production, formation of concentrated district based on industrial park, and the
control law into PTOS control. These control laws were then applied to the function of expanded service of eco-industrial park for ecology and agriculture.
position-velocity control loop in a permanent magnet synchronous motor (PMSM) Mentioned three functions are important for the extreme arid area.
servo system in simulation. The simulation results verify that the PTOS control is
capable of tracking the target position fast, the CNF control can achieve the target
position with a small overshoot and under the MSC scheme the servo system is
able to track a wide range of target positions fast and accurately.

Poster-05 Poster-06

Study on fuzzy-logic PID control system of synchronous jack-up The straight line mechanism base on peaucellier inversion
system of turbine runner static equilibrium experiment instrument
Yongxing Hao, Chaofeng Wang, and Lihui Song Kai He, Lei Guo, and Yuan Song
(North China Univ. of Water Resources and Electric Power, China) (Beijing Univ. of Posts and Telecommunications, China)

Basing on the LabVIEW platform, a Fuzzy-logic PID controller is developed. This paper gives the preliminary analysis based on the Peaucellier inversion
And the controller is applied to a control system of synchronous jack-up system of instrument, and calculates the inverse apparatus according to the theory of
turbine runner static equilibrium experiment. Inputting with a unit step signal, the inversion, a straight mechanism based on the Peaucellier inversion is designed and
control system was simulated. The result shows that, with only a small amount of calculate the maximum circle radius of inversion instrument. Then instrument
overshoot, the control system using fuzzy-logic PID controller responses and could switch between the straight line and circle by improving the straight line
stabilizes faster than the one using normal PID controller. Fuzzy-logic PID organization of inversion instrument. An assembly of straight-line mechanism is
controller shows great value in the practical implementation of the hydraulic drew. The prototype is built, and the validity of the mechanism is verified by
system control. It also can be extended to other areas such as bridge reinforcing experiments.
and migration of cultural relic and so on.
 Poster Session
Time: 10:20-11:30, Sunday, August 23, 2015 Chair: Changan Jiang (Ritsumeikan Univ., Japan)
Poster, Main Conference Hall Xiaoguang Zhou (Beijing Univ. of Posts and Telecommunications, China)
Co-Chair: Shin Wakitani (Tokyo Univ. of Agri. and Tech., Japan)
Co-Chair: Yixin Yin (Univ. of Sci. and Tech. Beijing, China)

Poster-07 Poster-08

Mechanical design and realization of automatic docking Design and implementation of a mechanical calculator for
mechanism quadratic equation with one unknown
Xinhu Mo, Lei Guo, Yuan Song, and Ying Zhang Bin Xing, Lei Guo, Shimin Wei, Yuan Song, and Ying Zhang
(Beijing Univ. of Posts and Telecommunications, China) (Beijing Univ. of Posts and Telecommunications, China)

Automatic docking technology is broadly used in the fields of aerospace and The mechanism structure and the mechanical measure method of the calculator are
mechanical manufacturing. However, the technology is rather complicated to designed, which are the key technology of calculator for the quadratic equation
understand. In this paper, an effective automatic docking model based on the with one unknown. Firstly, the basic principle of a kind of mechanical calculator
mechanical properties of the four-bar parallelogram linkage is proposed for for quadratic equation with one unknown is proposed. Secondly, the function and
teaching demostration. To make the docking mechanism keep equilibrium, the performance requirements of the system is analyzed and the linkage mechanism is
auxiliary link and leveling mechanism are added to the original four-bar selected as the core computing mechanism. Then the overall structure is designed.
parallelogram mechanism. In order to keep the two docking objects in the correct Finally, the control system is designed, after designing the system operation
attitude during docking, movement status of the model is analyzed. Then the process, the GUI is realized. The ability of solving quadratic equation with one
modeling, virtual assembling and motion simulation are realized in the software unknown is verified through the physical prototype debugging.
Solidworks. At last, a physical prototype is built, and the feasibility of the docking
mechanism is verified by experiments.

Poster-09 Poster-10

A comparative study on inter-tern short circuit fault of PMSM A research on construction of 30 bus large-scale smart grid model
using finite element analysis and experiment
Sho Kainose and Ken Nagasaka
Yongcan Li and Yongchun Liang (Tokyo Univ. of Agri. and Tech., Japan)
(Hebei Univ. of Sci. and Tech., China)

Permanent magnet synchronous machines (PMSM) are used in many This study aims at constructing a large-scale smart grid model on the
applicationsinter-turn short fault is the most probably fault in permanent motor. MATLAB/Simulink. Smart grid is the next-generation power network which
The motor we use in this reaearch is a 3-phase, 10-pole, 12-slot PMSM machine. introduces information and communication technology into a power system. In our
Model the motor by using MAXWELL 14, the waveform and the frequency laboratory, a power system simulation model based on the IEEE 30 Bus Test
spectrum of stator current under normal and inter-turn short-circuit fault condition System was constructed on the MATLAB/Simulink in past. In this study, we
is given out. In order to verify the result, we conducted experiments on normal and concentrate on modeling a suitable communication system model combined with
fault motors by using LabVIEW. Obtain the waveform and the frequency the IEEE 30 Bus power system model on the MATLAB/Simulink. This
spectrum of stator current under normal and inter-turn short-circuit fault condition. communication model can reproduce influence which usually occurs in
The research was under without load and with load condition.The comparison communication, such as noise, delay, fluctuation of arrival time of packets, and
result shows that the simulation result is correct and FEA method is an effective packet loss. In model construction, effective design is given to prevent a drop of
method to study the inter-tern short circuit fault of PMSM. simulation speed. Finally, it is tried to incorporate the communication blocks into
the 30 bus power system model.

Poster-11 Poster-12
Modeling a lithium-ion battery based on a threshold model Modelling the relationship between constituents and anti-RSV
effect of baikal skullcap root
Zhi Zhang
(Beijing Jiaotong Univ., China) Ke Li (Shandong Acad. of Sci., China),
Bonian Zhao, Yan Gao, Bianli Wang, and Zongyuan Yu
(Shandong Acad. of Chinese Medicine, China)
Hongpeng Zhao (Qianfo Mountain Hospital of Shandong Province, China)

Modeling a lithium-ion (Li-ion) battery is the core issue for electric vehicles In this work, based on least squares support vector machine regression, a model
applications. The valid battery model and accurate model's parameters can that characterizes the relationship between constituents of Baikal skullcap root and
improve accuracy of the state of charge (SOC) estimation and thus promote the therapeutic index of anti-respiratory syncytial virus was established. The
commercialization of electric vehicles. Considering an electrical battery model computational simulation showed that this model fits well with the experimental
with two resistance-capacitance (RC) parallel networks, the series resistance and data, and validation experimental results also supported the theoretical predictions.
the open-circuit voltage (OCV)-SOC function, the accurate estimate of battery
parameters through a threshold model, namely threshold autoregressive and
moving average with exogenous inputs (TARMAX) model is proposed, in this
paper, which only needs online terminal voltage and current data. The approach is
built by dividing voltage data into several regimes according to the thresholds. For
each regime, a linear ARMA model is established which can be represented by the
state-space equation for the linear battery model. Thus, the battery's parameters
can be identified. To facilitate obtaining the time-variant model parameters, the
terminal voltage and current data are collected by the sliding window. Finally, the
effectiveness of the proposed modeling approach is verified by the simulation.
 Poster Session
Time: 10:20-11:30, Sunday, August 23, 2015 Chair: Changan Jiang (Ritsumeikan Univ., Japan)
Poster, Main Conference Hall Xiaoguang Zhou (Beijing Univ. of Posts and Telecommunications, China)
Co-Chair: Shin Wakitani (Tokyo Univ. of Agri. and Tech., Japan)
Co-Chair: Yixin Yin (Univ. of Sci. and Tech. Beijing, China)

Poster-13 Poster-14

Energy-saving improvement and simulation study on digital Path planning of airfoil surface for robotic fibre placement
hydraulic system
Lina Li, Xingang Wang, De Xu, and Min Tan
Fei Song, Jingjun Lou, and Likun Peng (Chinese Acad. of Sci., China)
(Naval Univ. of Eng., China)

In order to solve the inefficiency problem of the constant pressure digital In this paper, a new path planning algorithm based on triangular meshes of the
hydraulic system, this paper built the model of the system and found primary airfoil open surface for four commonly used laying angles in robotic fibre
reasons leading to power losses through model simulation. With the pressure placement is proposed. In the proposed algorithm, the 0o paths are easily
compensation technology and the variable frequency hydraulic technology, a new computed by the slicing method, the 45o paths are obtained by general rotation
load-sensing digital hydraulic system was designed to instead of the original transformation, and the method for equally divide arc is adopted to get the 90o
system and the energy-saving controller of the new system was also designed. paths which can ensure the complete fibre tows connecting the root and tip part
Through the simulation comparison study of two systems, it was found that the distributed evenly on the surface. Then the redundant path points are deleted, the
load-sensing system can effectively reduce the overflow loss and the throttling final path points are obtained to generate the poses of the fibre placement head
loss and that the average efficiency of the new system is 30% more than the which is mounted at the robot manipulators end-effector. Finally the simulation is
original system. carried out in software RobotStudio and the effect of laying is displayed using
OpenGL. The results of the simulation verify the effectiveness of the proposed
method.

Poster-15 Poster-16

Strongly stable GPC with suppression of steady state gain and Model free fault detection using one set of reference input and
closed-loop poles output data
Akira Inoue (Okayama Univ., Japan), Akira Inoue (Okayama Univ., Japan),
Tomohiro Henmi (National Institute of Tech., Kagawa College, Japan) Tomohiro Henmi (National Institute of Tech., Kagawa College, Japan)
Mingcong Deng (Tokyo Univ. of Agri. and Tech., Japan) Mingcong Deng (Tokyo Univ. of Agri. and Tech., Japan)

Strongly stable control system is defined as a system having both of stable poles This paper proposes a method for fault detection of control plants without using
of closed loop systems and stable poles of controllers. The strong stability is plant models. Safety is the most important issue in control and for safety, fault
important for safety, that is, the system is stable even when feedback loop get detection is inevitable. So far, mainly, there exist two methods to detect faults. The
breakdown by an accident. The authors have already derived a strongly stable first is a statistical method, which uses statistical characteristics of input signals
controller for GPC systems. In designing the strongly stable GPC, three and fault signals. To extract the statistical characteristics, this method requires
characteristics should be specified. (1) The poles of the closed loop system. (2) sufficiently fluctuated signals, which is not satisfied in practical applications. The
The poles of the controller. (3) The steady state gain. The existing strong stable second method is to use a model, such as to estimate an output in normal state of
GPC by authors is extended from the standard GPC to two degree of freedom the plant using an observer and compares the estimated output to a measured
compensators by using coprime factorization approach and introducing new output of the plant to be detected. But usually plant models are not available in
design parameters. In the extended GPC, there exist hidden poles brought by industry. Hence, a method for fault detection without using signals of sufficient
pole-zero cancellation. And in initial condition mismatch or sudden change of richness and also without using plant models is required. The method in this paper
state by impulsive disturbances, the effect of the hidden poles appears. This means does not require sufficient richness of signals and plant models, hence it is
for safety, all the poles including the hidden poles should be stable. To design safe practically useful. The method uses a pair of input and output data of a normal
GPC, it requires to satisfy these three characteristics and needs complicated state as reference data instead of using a model. From the data, the method derives
trade-off between design parameters. The paper proposes to calculate the an impulse response sequence and compares the output estimated from the
mathematical expressions by using symbolic computation software to make the sequence with the measured output and detects a fault from the difference of the
trade-off simple. In an example of a water level control plant, the mathematical outputs. A simulation example of a water level control experimental plant is given
expressions are obtained explicitly and using the expressions, a controller is to show the effectiveness of the method.
designed to satisfy the three specifications.
Poster-17 Poster-18

Research on algorithm of nonlinear self-adaptive flight control The indoor real-time 3D localization algorithm using UWB
system
Xuehong Li and Shuhua Yang
Hao Long (Beijing Union Univ., China) (Univ. of Sci. and Tech. Beijing, China)
Shujie Song (Aviation Industry Co. of China, China)

For the nonlinear control problem of Super-maneuver aircraft, a nonlinear flight The actual location application always demands rapid initial positioning, high real
control law was presented based on the nonlinear structure adaptive model time performance and high accuracy. In order to meet the demands above, this
inversion system. The sufficient and necessary condition of the control law was paper mainly focus on the UWB indoor real-time 3D TDOA localization
analyzed. Finally the results of the height-angle maneuver simulation shows that algorithms based on the kalman filter. Firstly, the LKF, EKF and UKF algorithms
the designed control system has good performance. are applied to 3D localization, and the model parameters are given. And then, this
paper proposes a collaborative algorithm which is the Chan algorithm provides the
initial value and the LKF, EKF, UKF real time positioning algorithms do the
subsequent operations, and proposes a method to switch the algorithms
dynamically. Finally, the result of the simulation shows the indoor real-time 3D
dynamic positioning algorithms achieve a good performance in both speed and
precision.
 Poster Session
Time: 10:20-11:30, Sunday, August 23, 2015 Chair: Changan Jiang (Ritsumeikan Univ., Japan)
Poster, Main Conference Hall Xiaoguang Zhou (Beijing Univ. of Posts and Telecommunications, China)
Co-Chair: Shin Wakitani (Tokyo Univ. of Agri. and Tech., Japan)
Co-Chair: Yixin Yin (Univ. of Sci. and Tech. Beijing, China)

Poster-19 Poster-20

Operator and sliding mode based nonlinear control for cooling Control system design for ionic polymer metal composite using a
and heat-retention system actuated by Peltier devices single neuron based adaptive PID approach
Takafumi Hatano, Mingcong Deng, and Shin Wakitani Yiping Chang and Hui Wang
(Tokyo Univ. of Agri. and Tech., Japan) (Zhongyuan Univ. of Tech., China)

In this paper, a nonlinear control system is designed for a Cooling and Owing to simple structure and strong robustness characteristics, the traditional
Heat-retention system using peltier devices. First, stability of this system is (proportional-integral-derivative) PID controllers have been widely used in
guaranteed by the operator theory. Second, tracking performance is compensated designing control systems of industrial process. However, due to complex
by Sliding mode control with reducing the chattering using sigmoid function. nonlinear properties in some dynamics plants, it is difficult to obtain accurate
Third, tracking perforance is proven. Finally, the effectiveness of the designed mathematical models, and vulnerable to the object and environment. So, it is
controller is showned by simulation results. difficult to set controller parameters, especially no online self-tuning. Addressing
an ionic polymer metal composite (IPMC) artificial muscle with high nonlinear
properties and model uncertainties, an IPMC position tracking control system
based on single neuron adaptive-PID control approach is proposed by using neural
network self-learning and nonlinear mapping ability. The designed system can not
only achieve position tracking and online self-tuning, but also guarantee robust
stability in the presence of effect of uncertainties the effectiveness of the
proposed method is confirmed by simulation results.

Poster-21 Poster-22

Evaluation method of main production resource of society and Dynamic and energy analysis based on bumper system low-speed
industrial park of extreme arid region (2) collision test
Litifu Zulati Zhiming Wang and Li Chen
(Xinjiang Univ. of Finance and Economics, China) (Shanghai Univ., China)

This research presented three kinds of evaluation methods that can be feasibly The dynamic load strength test of bumper system is an important parameter for
applied for the short energy and water sources areas, such as the evaluation vehicle passive safety, but different test standards, different test requirements,
methods of regional potential energy including the estimation of general potential which caused much trouble to the test authority. So dynamic and energy analysis
energy, and the energy used in the industrial production process, and also the based on bumper system low-speed collision test is very impotant. Build
available renewable energy that may be developed. The Potential Energy Index parameterized model of bumper system, analyse the dynamic characteristics and
that can describe the effective support period for society economy is presented in energy conversion characteristics when the bumper system low-speed collision
this paper. The Input and Output Theory is used for converting the renewable occurs, which is based on the ADAMS virtual prototype technology. At the end of
energy into the regular energy resources that can be regarded as a steady energy the paper we will use the test machine for the dynamic load strength trials of
element in the society economy system. This paper also considered the evaluation bumper system on the basis of three typical standards, the result is compared with
method of the degree of energy intensive utilization in the process of industrial the below result that is analysed by ADAMS. According to the comparisonwe
production and society utilization, the evaluation of industrial process mainly can find the correctness of the simulation result is verified.
deals with the Industrial Parks that widely exist in short energy area.

Poster-23 Poster-24

Feasibility of neucube spiking neural network architecture for Optimal fiber orientations and topology of compliant mechanisms
EMG pattern recognition using lamination parameters
Long Peng and Zengguang Hou (Chinese Acad. of Sci., China), Xinxing Tong, Wenjie Ge, and Yonghong Zhang
Nikola Kasabov (Auckland Univ. of Tech., New Zealand), Guibin Bian (Chinese Acad. of Sci., China), (Northwestern Polytechnical Univ., China)
Luige Vladareanu (Romanian Acad., Romania), and Hongnian Yu (Bournemouth Univ., UK)

Multichannel electromyography (EMG) signals have been used as human-machine A design approach for compliant mechanisms with straight fiber laminate is
interface (HMI) for the control of pattern-recognition based prosthetic system in presented to obtain the optimal fiber orientations and topology structure
recent years. This paper is a feasibility analysis of using recently proposed simultaneously in this paper. Firstly, an equivalent constitutive relation of straight
NeuCube spiking neural network (SNN) architecture for a 6-class recognition fiber laminate is parameterized in terms of lamination parameters. Taking
problem of hand motions. NeuCube is an integrated environment, which uses SNN lamination parameters and relative density as design variables, minimizing the
reservoir and dynamic evolving SNN classifier. NeuCbube has the advantage of weighted linear combination of the mutual strain energy (MSE) and the strain
processing complex spatio-temporal data. The preliminary experiments show that energy (SE) is considered as objective function to achieve the desired deformation
Neucube is more efficient for EMG classification than commonly used machine and enough load carrying capacity with the volume constraint. The nonlinear
learning techniques since it achieves better accuracy as well as consistent optimization problem is solved via the well known method of moving asymptotes
classification outcomes. The performance of NeuCube combined with TD features (MMA). The outcomes of first step are the optimal topology structure and
reaches up to 95.33% accuracy after a careful selection of the features. This paper lamination parameters. Secondly, the optimal fiber orientation for each layer is
demonstrates that NeuCube has the potential to be employed in practical found by matching the optimal lamination parameters. Finally, the numerical
applications of myoelectric control. examples of designing compliant inverters are investigated to demonstrate the
effectiveness of the proposed method.
 Poster Session
Time: 10:20-11:30, Sunday, August 23, 2015 Chair: Changan Jiang (Ritsumeikan Univ., Japan)
Poster, Main Conference Hall Xiaoguang Zhou (Beijing Univ. of Posts and Telecommunications, China)
Co-Chair: Shin Wakitani (Tokyo Univ. of Agri. and Tech., Japan)
Co-Chair: Yixin Yin (Univ. of Sci. and Tech. Beijing, China)

Poster-25 Poster-26
A software platform for interactive analysis on 3D medical data Master-slave force control based on Grey GM(1,1) model of robot
grippper
Wei Cheng and Yongfei Xiao (Shandong Acad. of Sci., China)
Youlei Zhao (Taikoo (Shandong) Aircraft Eng. Co.,Ltd., China) Wei Cheng, Xuelin Wang, and Haiyan Ma
(Shandong Acad. of Sci., China)

Analysis on the anatomical relationship between the conventional 2D medical This paper describes the position and force control of a novel two-finger gripper
images and correlative tissues is vital to both the physicians and the interns in using force control technology, a master-slave force control strategy is deveoped
surgery planning. However, regular measures are not intuitive and greatly depend for making the force track the desired force quickly with lower overshoot. PID
on the physicians experience and their medical knowledge. In this paper, a controller is adapted to the position system for the master finger, and force control
software platform is developed for a more intuitive and accurate analysis on the is used to the slave finger with grey GM(1,1) model, the prediction models are
anatomical relationship in the diagnosis and surgery planning. Both the oblique built by the signal of force acquired from sensors, the force controller enforces a
clipping image and a 3D tissue model with an interactive transfer function module relationship between the position of each finger and the force on the manipulated
are showed in a stereoscopic scene to illustrate the 3D human pathology and objects, and the weights of the current and the future force errors are automatic
anatomy in real time. For convenience, a software interface is also designed so adjusted in integrated error according to the model precision. The force controllers
that different 6DOF interactive devices can be applied in the project. This article can employ the information of past, present and future force to calculate an
offers a novel resolution for the improvement in the assessment of anatomical appropriate control correction to pre-compensate the force errors, thus it can
structures and understanding of medical imaging technology in a more realistic obtain lower overshoot and faster response. Simulation results are presented to
manner. demonstrate the efficacy of the proposed master-slave force controllers.

Poster-27 Poster-28

Remote automatic test system based on MATLAB using VISA Cubic spline solution for a class of boundary value problems using
over LAN spectral collocation method
Wei Cheng, Fenggui Wang, and Haiyan Ma Tianjun Wang, Qixian Zhou, and Tengteng Cui
(Shandong Acad. of Sci., China) (Henan Univ. of Sci. and Tech., China)

VISA over LAN was used to design remote automatic test system, in which In this paper, we investigate the numerical solution of a class of boundary value
MATLAB was selected as the application development environment to enhance problems of ordinary differential equations using parametric cubic spline function
the capability of data processing and analysis. Native GPIB commands and and spectral collocation method. Efficient algorithms are implemented. Numerical
queries were funneled to/from VISA instrument object using Instrument Control results demonstrate the high efficiency and accuracy of the proposed method.
Toolbox functions, and then used to control electrical measuring instruments
through VISA I/O Library API and VXI-11 protocol. An example system for
microwave imaging test with digital oscilloscopes integrated was developed to
perform the test of radio frequency and intermediate frequency signals.

Poster-29 Poster-30

Maximum efficiency control method of permanent magnet Modeling and simulation of hydro-mechanical continuously
synchronous motor based on three-dimensional table variable transmission system based on Simscape
Gensheng Li, Jinfa Xie, and Liyou Xu Zhili Zhou and Jiazhen Zhang (Henan Univ. of Sci. and Tech., China)
(Henan Univ. of Sci. and Tech., China) Zhiqiang Guo (YTO Group Corporation, China)
Liyou Xu (Henan Univ. of Sci. and Tech., China)
Increase the operating efficiency of the motor can reduce energy consumption and The traditional methods of dynamic modeling for the simulation of
the requirements for the radiator, for some applications, the operating efficiency of hydro-mechanical continuously variable transmission (HMCVT) are cumbersome
the motor even become one of the most important factors. In order to improve the and complex. The physical simulation models of engine, hydraulic speed
efficiency of the motor, in this paper, the maximum efficiency permanent magnet regulating mechanism, mechanical speed regulating mechanism and the load were
synchronous motor control method is proposed, the method takes into account the established based on Simscape, and the dynamic characteristics of the components
copper and iron losses based on the motor model, and can make a permanent were analyzed in the paper. The simulation results show that with the change of
magnet synchronous motor to achieve optimal efficiency at different speeds by the displacement ratio e, the output speed of the system is continuous and stepless,
way of look-up table. The proposed method reduces the amount of computation and the output torque of the system is constant when the torque of load is constant.
greatly, and it is easy to implement in practical control. The simulation results From the analysis results, it is concluded that the physical models are feasible. The
showed that the method can improve the efficiency of permanent magnet research provides a theoretical foundation and a verification platform for further
synchronous motor effectively. development and improvement of HMCVT shifting strategy.
 Poster Session
Time: 10:20-11:30, Sunday, August 23, 2015 Chair: Changan Jiang (Ritsumeikan Univ., Japan)
Poster, Main Conference Hall Xiaoguang Zhou (Beijing Univ. of Posts and Telecommunications, China)
Co-Chair: Shin Wakitani (Tokyo Univ. of Agri. and Tech., Japan)
Co-Chair: Yixin Yin (Univ. of Sci. and Tech. Beijing, China)

Poster-31 Poster-32

Performance comparison for savonius type wind turbines by Fuzzy clustering analysis of batch production recipe
numerical analysis approaches
Shanzhong Liu and Xiaonan Song
Lei Song (Henan Univ. of Sci. and Tech., China) (Henan Univ. of Sci. and Tech., China)
Hongzhao Liu (Xian Univ. of Tech., China)
Zongxiao Yang (Henan Univ. of Sci. and Tech., China)
Wind energy is a clean and renewable energy resource and has been paid more and According to the recipe information model specified in batch standard (ISA S88),
more attention by scholars around the world. As a drag-driven type Vertical Axis via the modified recipe similarity measurement method, an improved fuzzy
Wind Turbine (VAWT), traditional Savonius wind rotor has disadvantage of low clustering algorithm (IFCA) is proposed based on fuzzy c-means (FCM) algorithm
wind energy utilization but has a lot of the advantages compared with Horizontal in the PNN network. IFCA can effectively improve the choice of clustering center
Axis Wind Turbines (HAWTs) and lift-driven VAWTs. A fish-ridged wind rotor and has a good clustering effect for the recipe of batch production. Then is used
is proposed based Savonius rotor to improve the flow field characteristics. In this IFCA to cluster for production recipe of an equipment unit wort evaporator in beer
paper, the two wind rotors are analyzed and compared by computational fluid production. Finally, the recipe classification result is compared with the
dynamics approach. The simulation results show that the flow field of the classification of FCM algorithm, and the effectiveness of IFCA is proved.
fish-ridged rotor is smoother than the Savonius rotor and the maximum value of
wind energy utilization of the former rotor is bigger than that of the latter. It
indicates that the new type of the rotor has good performance.

Poster-33 Poster-34

Fully discrete subgrid stabilized finite element method for the The simplified homogeneous balance method and its application to
Darcy-Brinkman equations in double-diffusion convection Whitham-Broer-Kaup-Like equations
Yunzhang Zhang, Zhoufeng Wang, and Qili Tang Wei Li (Henan Univ. of Sci. and Tech., China)
(Henan Univ. of Sci. and Tech., China) Xinrui Wang (Univ. of Chicago, USA)

We present a fully discrete subgrid stabilized finite element method to solve the In this paper, we have simplified the homogeneous balance method (HB). Using
Darcy-Brinkman equations in Double-Diffusion convection. The time is advanced this simplified approach, the Whitham-Broer-Kaup-Like (WBKL) equations are
by one order Backward Euler scheme. With the proper choosing of stabilized investigated. Various exact solutions of the WBKL equations are obtained via a
parameters, the optimal error estimates in space can be obtained for velocity, nonlinear transformation with the aid of solutions for the linear equation.
temperature and concentration in H1 semi-norm. The derived theoretical results are
supported by numerical experiments. One example is to verify the convergence
results and the other example is a pure thermal convection in a porous medium to
verify the stability of ours method.
 Organized Session
Advanced control systems design - theory and applications
Time: 13:00-14:40, Sunday, August 23, 2015 Chair: Shiro Masuda (Tokyo Metropolitan Univ., Japan)
SunM01, Room No. 01 Co-Chair: Shin Wakitani (Tokyo Univ. of Agri. and Tech., Japan)

SunM01-01 SunM01-02

Tracking control of quad-rotor helicopters suspended a power Multivariable controller design evaluating closed-loop interaction
supply cable with on-line estimation of disturbances by iterative LMI optimization using frequency response data
Hayami Hayakawa, Shogo Obata, and Manabu Yamada Shogo Shinoda, Kazuhiro Yubai, Daisuke Yashiro, and Junji Hirai
(Nagoya Institute of Tech., Japan) (Mie Univ., Japan)

In this paper, we propose a tracking controller for quad-rotor suspended a flexible This paper proposes a multivariable controller design method using frequency
cable. The first contribution is to present a new disturbance observer to estimate response data. The proposed method evaluates a diagonal dominance in a
online both the tension of the cable and unknown disturbances applied to closed-loop manner by using a singular value and designs a controller by solving
dynamics. The second one is to propose a new exact linearization method which an optimization problem that involves nonconvex quadratic matrix inequalities.
reduces the nonlinear closed-loop system to a linear controllable state equation by This optimization problem is rewritten as linear matrix inequalities (LMI) by
changing coordinates in the state-space. By using the linearization, a new tracking introducing iterative LMI constrains. The controller achieving highly diagonal
controller based on disturbance observer is presented for obtaining asymptotically dominance is designed by repeating an update of the optimization problem and the
rejection against unknown constant disturbance. convex optimization. This optimization algorithm guarantees monotonically
convergence of an evaluation value in terms of the iteration. The experimental
results show the effectiveness of the proposed method in comparison with the
conventional method.

SunM01-03 SunM01-04

Flow disturbance suppression for a pneumatic vibration isolator Design of a fault tolerant control system using
using a central pattern generator Mahalanobis-Taguchi(MT) system
Yukinori Nakamura, Daishi Funaki, Mami Kimura, and Shinji Wakui Shin Wakitani and Mingcong Deng
(Tokyo Univ. of Agri. and Tech., Japan) (Tokyo Univ. of Agri. and Tech., Japan)

This paper presents the suppression method of flow disturbance, which is the In this research, a failure tolerance control system design scheme using
variation of compressed air supplied to a pneumatic vibration isolator. A central Mahalanobis-Taguchi (MT) system is proposed. In this method, we consider a
pattern generator (CPG) is utilized to compensate effect of flow disturbance. In situation that an actuator such as a heater gradually deteriorates and eventually
order to implement a CPG-based controller, feedback type control scheme, which breaks down. The failure detection component in a control system detects whether
does not require a pressure sensor, is adopted. Moreover, CPG-based and the system is broken, and switches to a backup actuator if a fault is detected in the
displacement controllers are connected in parallel so as to avoid the vibration of actuator. In the proposed algorithm, an augmented system including an actuator is
isolation table during the start-up of isolator. The proposed approach is verified by described as a first-order plus time-delay system, and their system parameters are
simulation and experiment. estimated by the recursive least squares method. Unit space is designed using the
estimated system parameters that are obtained during normal operation; based on
the Mahalanobis distance, the failure detection component determines whether a
newly obtained system parameter in operation belongs to the unit space. The
control system switches to a backup actuator if the newly estimated parameters do
not belong to the unit space. The effectiveness of this method is evaluated by
simulation experiments.

SunM01-05
55
Data-driven PID gain tuning for unknown impulse disturbance
attenuation
Shiro Masuda
(Tokyo Metropolitan Univ., Japan)

Data-driven design approaches based on input-output measurements with no need

for help from a plant model have attracted attention from several researchers. We
have proposed such a disturbance attenuation Fictitious Reference Iterative Tuning
(FRIT) using input-output data generated by a step or impulse disturbance, and
showed the effectiveness of theapproach through experiments of DC motor control
and helicopter attitude control. The present work applies the approach to the
experimental data for temperature control. The PID gains are tuned so that
temperature of a piece of metal would follow a disturbance reference model output
signal. From the experimental result, the disturbance attenuation FRIT is effective
for PID gain tuning from closed-loop input and output experimental data
generated by a impulse disturbance with unknown magnitude.
 Regular Session
Mechatronic systems analysis and control I
Time: 13:00-14:40, Sunday, August 23, 2015 Chair: Xinkai Chen (Shibaura Institute of Tech., Japan)
SunM02, Room No. 02 Co-Chair: Hongbo Wang (Yanshan Univ., China)

SunM02-01 SunM02-02

Size and weight reduction of semicircular duplex manipulator for Adaptive fault diagnosis for continuous time-delay repetitive
rescue operation system subject to sensor fault
Soichiro Kimura and Kazuyuki Ito Haigang Zhang, Sen Zhang, and Yixin Yin
(Hosei Univ., Japan) (Univ. of Sci. and Tech. Beijing, China)

Applications of robots for rescue operations have attracted considerable attention. This note proposes an adaptive fault diagnosis scheme for the continuous
However, at real disaster sites, some of the spaces in the rubble are so narrow that time-delay repetitive system subject to the sensor fault. The theory of iterative
conventional rescue mobile robots cannot pass through them. Hence, the learning control (ILC) is employed in the fault reconstructed method. With the
development of a robotized endoscope is expected. In this study, we redesigned increase of iterative step, the sensor fault estimation signal can track the actual
our previous rescue manipulator to downsize its body without losing robustness. fault tightly including slow- and fast- varying fault. In order to verify the
We developed a new small-sized duplex manipulator and successfully reduced its effectiveness of the proposed fault diagnosis method, three kinds of sensor faults
size by 70%. are taken into consideration in the simulation part. The results show that the
reconstructed fault signal can track the actual fault well despite of the kinds of
sensor faults.

SunM02-03 SunM02-04

Model identification and adaptive control design for a 6 DOFs Precise tracking control for Piezo-actuated stage using inverse
manipulator compensation and model predictive control
Weiliang Ge (Univ. of Electronic Sci. and Tech. of China, China), Nguyen Manh Linh, Tran Vu Minh, and Xinkai Chen
Wei He (Univ. of Sci. and Tech. Beijing, China)
Yunchuan Li (Univ. of Maryland College Park, USA) (Shibaura Institute of Tech., Japan)
Chenguang Yang (Plymouth Univ., UK)

In this article, kinematic modeling design of a humanoid robot is presented by In this paper, we consider tracking control problem of piezo-actuated stage which
using Devanit-Hartenberg (D-H) model. Based on the coordinate frame, the model is composed by a piezoelectric actuator (PEA) and a positioning mechanism (PM).
identification of a 6 degree-of-freedom (DOF) upper limb is investigated with This plant can be modeled by cascading a hysteresis with a linear dynamical
Newton-Euler (NE) formula. Particle Swarm Optimization (PSO) is used to system. The tracking performance of this system is significantly affected by
optimize the trajectory of each joint, the adequate excitation of the robot is hysteresis phenomenon of PEA. In order to improve the performance, a modified
provided and the estimated result is improved. The estimated inertia parameters Bouc-Wen model is proposed to describe the hysteresis more accurately. Then,
are taken as the initial values of the Recursive Newton-Euler (RNE) adaptive inverse Bouc-Wen is used to compensate hysteresis nonlinearity. Finally, model
control algorithm. Simulations are provided to verify the result of the predictive control (MPC) with integral of error state variable is employed for
identification algorithm. control design. Experimental results show that the proposed method has excellent
tracking performance in comparison with conventional proportionalintegral (PI)
controller.

SunM02-05

Oscillatory behavior based fault feature extraction for bearing

fault diagnosis
Juanjuan Shi and Ming Liang
(Univ. of Ottawa, Canada)

An intelligent fault signature extraction scheme based on oscillatory behaviors is

reported in this paper for bearing fault diagnosis. The proposed method is based
on the joint application of morphological component analysis (MCA) and tunable
Q-factor wavelet transform (TQWT) to decompose a signal into two signal
components (i.e., low- and high-oscillation components) according to whether
they having sustained oscillations. As bearing fault-induced transients
(low-oscillation component) oscillate differently from periodic interferences and
noise (high-oscillation component and residual), they can be separated via the
MCA with the aid of TQWT which is parameterized by Q-factor and plays a role
of distinguishing signal components presenting different oscillatory behaviors.
The low- and high-oscillation components can be obtained by solving the
objective function formulated based on MCA and TQWT. The determination of
Q-factor for each signal component representation is also explored in this paper.
The effectiveness of the proposed method is examined by experimental data.
 Regular Session
Complex systems control and applications I
Time: 13:00-14:40, Sunday, August 23, 2015 Chair: Ni Bu (Qingdao Univ. of Sci. and Tech., China)
SunM03, Room No. 03 Co-Chair: Zhengfeng Ming (Xidan Univ., China)

SunM03-01 SunM03-02

Switched reluctance motor/generator simulation research based Surface flaws detection algorithms for large aperture optical
on compressed air energy storage system element
Xiaoshu Zan and Hong Zhu Zhengtao Zhang, De Xu, Xian Tao, and Feng Zhang
(China Univ. of Mining and Tech., China) (Chinese Acad. of Sci., China)

Compressed air energy storage (CAES) system has become a popular energy In this paper we consider the inspection of surface flaws in large aperture optical
storage device for micro grid because of many advantages. The principle, element. A high efficiency and precision instrument is proposed which contains
structure, simulation model and the simulation results of the switched reluctance two kinds of imaging systems. One is dark-field imaging system(DFIS)
motor/generator system based on the CAES system have been studied in the paper. constructed by line scan camera with 10m resolution. The other is bright-field
The nonlinear simulation model by the flux linkage data calculated by imaging system (BFIS) constructed by microscope with 1m resolution.
two-dimensional finite element is built firstly. Then the sliding+PI control method Considering the small depth of field of DFIS, an adaptive scanning method based
is put forward in the motor process and the fuzzy+PI control method is put on collimation laser and several focusing points is proposed to keep the clarity of
forward in the generation process. The simulation results have validated the DFIS in large scope scanning. After the scanning, an image mosaic algorithm for
effectiveness of the control strategy and control parameters. The switched the DFIS is presented based on SIFT features and clustering algorithm. Then, the
reluctance motor/generator system has well rotor speed tracking and voltage feature extraction algorithm of flaws in DFIS and BFIS is designed. In order to
output performance and it has reference value and application prospect in the check the flaws more precisely, a location algorithm that can guide the BFIS to
CAES system. inspect the same flaw in DFIS is introduced. Finally, the calibration method of two
imaging systems is studied. Experiments show that the device can scan an optical
element with size of 810mm460mm in less than 6min without complex manual
operation and the detection precision can reach 3m satisfying the requirement of
practical need.

SunM03-03 SunM03-04

An adaptive observer-based fault detection and diagnosis for Minimum mutual information control for multi-variable
nonlinear systems with sensor and actuator faults non-Gaussian stochastic systems
Xiaomo Yan (Univ. of Manchester, UK), Qichun Zhang (Univ. of Manchester, UK)
Bailing Tian (Tianjing Univ., China), Aiping Wang (Anhui Univ., China)
Hong Wang (Univ. of Manchester, UK) Hong Wang (Univ. of Manchester, UK)
A novel fault diagnose method is presented for the nonlinear systems with coupled In this paper, a novel control algorithm is presented to minimize output statistical
fault and disturbance viaadaptive observer technique. Residual matrix is used in couplings for a class of dynamic nonlinear multi-variable non-Gaussian stochastic
this paper to describe the values of unknown inputs. Instead of adding robustness systems. Using Cauchy-Schwartz mutual information, a new performance
on the estimate states, this novel observer structure focuses on eliminating the criterion is given and this associated control algorithm is presented to minimized
disturbance signal on the residual matrix by using vector unitization so that the the mutual information of the system outputs so as to minimize the statistical
residual matrix can be approached precisely. The adaptive law is obtained by couplings among system outputs. The convergence of this presented control
using Lyapunov stability theory. Finally, an application on second order system is algorithm is analyzed and the effectiveness of the proposed algorithm is shown by
included in the end of this paper to show the effectiveness of the proposed simulation.
strategy.

SunM03-05

Design and simulation of dual fuzzy self-adjusting immune PID

controller for grain drying system
Aini Dai, Xiaoguang Zhou, Chi Zhang, Xiangdong Liu, and Huiling Zhou
(Beijing Univ. of Posts and Telecommunications, China)

The grain drying process has the characteristics of large time delay, nonlinearity,
multi disturbance and strong coupling. In order to reduce the grain moisture
content during drying process, to ensure the grain quality of the grain, and enable
the uniform moisture content of outlet grain, a new improved controller model for
grain drying which called the dual fuzzy self-adjusting immune PID controller in
this paper is proposed based on immune feedback mechanism, and the problem of
fixed change rate of the parameters of general fuzzy immune PID controller is
solved. Matlab simulation tool is used to simulate the control models of grain
drying system, and the effectiveness of this new improved control method for the
grain drying system is shown by simulation on the performance comparison with
the other control methods, such as: general PID, fuzzy self-adjusting PID and
general fuzzy immune PID. The simulation results show that the new fuzzy
immune PID controller has better dynamic performance in the aspects of overshot,
stability and anti-jamming.
 Regular Session
Control of mechatronic systems II
Time: 14:50-16:50, Sunday, August 23, 2015 Chair: Changan Jiang (Ritsumeikan Univ., Japan)
SunP01, Room No. 01 Co-Chair: Junming Xiao (Zhongyan Univ. of Tech., China)

SunP01-01 SunP01-02

A brushless DC motor speed control system based on DSP Precision size measurement instrument for irregular objects
controller
Wanji Liu (SIPO, China)
Songming Cao, Yong Liu, Ming Hu, Xin Fu Xian Tao (Chinese Acad. of Sci., China)
(Beijing Univ. of Posts and Telecommunications, China)

In this paper, the control system of brushless DC motor, using the TI Co.'s digital Precision size measurements are widely used in industrial production such as
signal processor TMS320F2812 as the main control chip, design for hardware of candies and cookies in the food industry. However, it is a challenge to measure
system are as follows :a rectifier circuit, inverter circuit, drive circuit, control irregular objects in a universal device because of poor adaptability and low
circuit, voltage and current detection circuit; software program design in DSP precision. In this paper, a high-precision measurement device for irregular objects
software development environment CCS, using C language and assembly based on machine vision is designed. It consists of a rotational adjustment
language a hybrid programming, the realization of the PWM wave generation and platform, two imaging units, a weighing device and a host computer. Armed with
brushless DC motor speed control and other functions. this device, a novel algorithm with good robustness is proposed to rapidly measure
irregular objects. The sub-pixel location algorithm based on spatial moment is
used for the accuracy size measurement. Experiments show that the measurement
accuracy is less than 25m.

SunP01-03 SunP01-04

A new method of periodicity estimation for mechanical acoustic Acoustic emission signal harr wavelet denoise research
data
Kenan Shen (Southwest Forestry Univ., China),
Zhipan Hong and Guoliang Lu Mingcong Deng (Tokyo Univ. of Agri. and Tech., Japan)
(Shandong Univ., China) Ming Li (Southwest Forestry Univ., China)

Periodicity estimation in mechanical acoustic time-series data is a well-established In order to obtain wood acoustic emission signal waveform. A multi channel
problem in data mining as it can be applicable in variety of disciplines either for acoustic emission signal acquisition system based on NI high-speed data
anomaly detection or for prediction purposes in industry. In this paper, we develop acquisition equipment was constructed. At the same time, a signal
a new approach for capturing and characterizing periodic patterns in time-series wavelet extraction and processing platform based on LabVIEW software
data by virtue of the dynamic time warping (DTW). We have conducted extensive was designed, which can be used to collect separate extract the acoustic
experiments to evaluate the proposed approach with synthetic data and our emission signal, spectral analysis and other basic functions. The system was
collected data in practice. Experimental results demonstrated its effectiveness and proved to be effective through the acoustic emission experiment. Through the
robustness on periodicity detection in highly noised data. comparison of three different signal extract ways, further verify the advantages of
harr wavelet analysis. As a platform for wood acoustic emission signal acquisition
and processing, this system provides the necessary foundation for the processing
of wood acoustic emission signal.

SunP01-05 SunP01-06

Operator-based nonlinear networked control system with packet Optimal control of non-prehensile manipulation control by two
error cooperative arms
Tomohito Hanawa and Mingcong Deng Changan Jiang, Satoshi Ueno (Ritsumeikan Univ., Japan)
(Tokyo Univ. of Agri. and Tech., Japan) Yoshikazu Hayakawa (Nagoya Univ., Japan)

This paper is concerned with the operator-based nonlinear networked control In this paper, an optimal control method to non-prehensile manipulation control by
systems with packet errors. In communication network, packet may include errors two cooperative arms is proposed. In details, based on the dynamic model of the
caused by noise in network, and in conventional control method, these packets are two-rigid-link object, an approximate model is derived. According to the obtained
regarded as lost packets. This means information of control signals is also lost model, an optimal regulator is designed to realize the holding motion of two
completly, even if the disturbance caused by errors acceptable for control systems. cooperative arms which is used to keep the two-rigid-link object stable on the
The proposed system treats packets in communication networks with errors as arms. Finally, simulation results are shown to verify the effectiveness of the
control signals with perturbation. This control scheme reduces the numbers of proposed method.
dropped packets caused by noise in communication network. Firstly, the
operator-based plant model and packet communication model with packet error
are proposed. Secondly, the operator-based nonlinear control system is proposed.
Finally, the effectiveness of the proposed control scheme is confirmed by
simulation.
 Regular Session
Mechatronic systems analysis and control II
Time: 14:50-16:50, Sunday, August 23, 2015 Chair: Shengjun Wen (Zhongyuan Univ. of Tech., China)
SunP02, Room No. 02 Co-Chair: Dongyun Wang (Zhongyuan Univ. of Tech., China)

SunP02-01 SunP02-02
SVR-based input-output mapping of a micro-hand Nonlinear remote temperature control of a spiral plate heat
exchanger
Kou Fujita (Tokyo Univ. of Agri. and Tech., Japan)
Shuichi Wakimoto (Okayama Univ., Japan), Ryohei Fujii, Mingcong Deng, and Shin Wakitani
Mingcong Deng and Shin Wakitani (Tokyo Univ. of Agri. and Tech., Japan)
(Tokyo Univ. of Agri. and Tech., Japan)

Micro-hand is one of soft actuators. It has many merits, however, it is of In this paper, the time-delay is considered when process equipment is controlled
nonlinearity. Moreover, it needs sensorless control. Sensorless control with by a wireless network in a large-scale plant. To prevent a decline of a control
SVR-based mapping of a micro-hand is proposed in this paper. Effectiveness of performance by the time-delay, the time-delay compensation controller is
the proposed method is verified by the experiment of the control system. designed. What's more, the experimental environment which simulates the remote
control is made, and the effectiveness of the designed control system is confirmed
by simulation and experiment results.

SunP02-03 SunP02-04

Operator based voltage control experiments of thermoelectric Robust controller design for networked control system with
generation system packet dropout and quantization error
Ryuji Yoshida, Mingcong Deng, and Shin Wakitani Haiquan Wang, Rongxiang Gao, Qian Zhang, and Shengjun Wen
(Tokyo Univ. of Agri. and Tech., Japan) (Zhongyuan Univ. of Tech., China)

Boost converters have nonlinear characteristics by semiconductor switching The stability of networked control systems with packet dropout and quantization
elements in a circuit.Generally, controllers are designed by a linearized model of error is researched in this paper. As a typical distributed closed-loop control
the boost converter based on state space averaging method.This study proposes system connected with network, quantization errors and packet dropouts
design scheme considering discontinuous conduction mode without linearization. frequently occur that the previous model of networked control system tend to
In detail, firstly, modeling of boost converter in discontinuous conduction mode is ignore them. So in order to describe exactly how the system will behave during
derived.Secondly, control system designed based on operator theory is control, the nonlinear quantizer to quantify status signals is introduced and
shown.Finally, the effectiveness of the proposed control system is verified by Markov jump system model is established to describe the packet dropout existing
simulation and experimental results to the change of the load and the input in networked control system. Subsequently, based on the established model,
voltage. state feedback controller is designed to ensure the stability of system and the
Linear Matrix Inequality (LMI) is used to solve the controller. The simulation
results show that state feedback controller can ensure the stability of controlled
system with packet dropout and quantization error.

SunP02-05 SunP02-06

Active power filter design for improving power quality Fractional PID based stability control for a single link rotary
inverted pendulum
Junming Xiao, Xiangming Zhang, Shengjun Wen, and Zhengbo Liu
(Zhongyuan Univ. of Tech., China) Wudai Liao, Zhengbo Liu, and Shengjun Wen
(Zhongyuan Univ. of Tech., China),
Shuhui Bi (Univ. of Jinan, China)
Dongyun Wang (Zhongyuan Univ. of Tech., China)

The paper design a current control based active power filter (APF) for improving Fractional order based stability control for the system of the single link rotary
power quality by reactive power compensation and harmonic filtering. The inverted pendulum is presented. The mathematical model is derived using
designed APF is based on a voltage source inverter (VSI), where the VSI is Lagrange Equation and the G-L fractional calculus. Then the integer order PID
controlled by two loops, the voltage control loop and the current control loop. The controller and fractional order PID controller are designed respectively. Under the
voltage control loop is used to regulate the DC link capacitor voltage of the VSI. condition of the same parameters, and, the integrator of order and a differentiator
Via triangular wave control, the current control loop is used to produce a tracking of order of the fractional order PID controller are adjusted adaptively. Finally,
PWM signal to shape the harmonic current such that it is in-phase with the same comparing with the integer order PID controller, the results show the system of
shape as the input voltage. The major advantage of the proposed APF is to search fractional order PID controller is more quickly swigged, and the stability control
the reference current for improving power quality by using dq0 detecting method. performance is more effective.
Simulation results are given to validate the proposed scheme. From the results, it
shows that the reduction of THD is better in power system networks.
 Regular Session
Complex systems control and applications II
Time: 14:50-16:50, Sunday, August 23, 2015 Chair: Aihui Wang (Zhongyuan Univ. of Tech., China)
SunP03, Room No. 03 Co-Chair: Hong Wang (Univ. of Manchester, UK)

SunP03-01 SunP03-02
Design of the automatic sample preparation system IPMC robust nonlinear tracking control design based on a
multi-objective particle swarm optimization-based RRCF approach
Xin Fu, Shimin Wei, Yong Liu, Ming Hu, and Songming Cao
(Beijing Univ. of Posts and Telecommunications, China) Wudai Liao, Tongbin Yan, Aihui Wang, and Yiwen Fu
(Zhongyuan Univ. of Tech., China)

This paper introduces a kind of automatic device(called ASPS), wihch implements In this paper, a robust nonlinear tracking control design for an ionic polymer metal
the function of automatic sample preparation. The components and design composite (IPMC) is proposed by using a multi-objective particle swarm
principles of the device are described in detail. The device consists of the optimization-based robust right coprime factorization approach. Addressing the
membrane feeding system, the x-y-z manipulator and the boxes feeding and difficult in obtaining the PI control parameters Kp, Ki of the former proposed
retriving mechanism. The design principles mainly involve tension control and the nonlinear robust tracking control system based on PI-based robust right coprime
avoidance of slippage. factorization approach, how to obtain the optimal control parameters Kp, Ki is
investigated by using MATLAB system identification toolbox and multi-objective
particle swarm optimization algorithm. That is, firstly, a new equivalent transfer
function model of the robust stable control system of IPMC is identified. For the
obtained transfer model, a multi-objective particle swarm algorithm optimization
is used obtain the control parameters Kp, Ki of PI controller. Finally, the
effectiveness of the proposed method system is confirmed by simulation results.

SunP03-03 SunP03-04

Robot arm with micro-hand robust control design using Complex event processing on uncertain data streams in product
operator-based robust right coprime factorization approach manufacturing process
Aihui Wang (Zhongyuan Univ. of Tech., China) Na Mao and Jie Tan
Hongnian Yu (Bournemouth Univ., UK) (Chinese Acad. of Sci., China)
Dongyun Wang (Zhongyuan Univ. of Tech., China)
Mingcong Deng (Tokyo Univ. of Agri. and Tech., Japan)

This work focuses on robust nonlinear control design of a robot arm with With the development of automatic production, manufacturing factories record
micro-hand by using operator-based robust right coprime factorization (RRCF) tremendous amounts of data with sensor devices deployed in a factory. Because of
approach. In detail, to control the precise endpoint position of robot arm and the inherent inaccuracy of sensor readings, these data are of high level of
obtain the desired force using micro-hand according to the external environment uncertainty. How to use Complex Event Processing (CEP) to get useful
or task involved, a connected feedback control system based on operator-based information for quality monitoring of products from a lot of uncertain raw data
RRCF approach is proposed. In first sub-system, to control the angular position of continually generated from the production lines is becoming a challenging
the robot arm, the operator controllers and the tracking controller are designed, research. Therefore, in this paper, we propose a model of uncertain complex event
and the robust stability and tracking conditions are derived using RRCF approach. processing system for real-time monitoring in product manufacturing process. And
The second sub-system is designed to control the micro-hand force, and the robust then we define the probabilistic event model and propose a probabilistic event
tracking conditions are also discussed based operator-based RRCF approach. detection algorithm based on rNFA and its optimization plan by event filtering. At
Finally, the effectiveness of the proposed control system is verified by simulation the same time, we introduce Conditional Probability Matrix (CPM) and describe
results. the calculation of probability of complex events with the multiplication theorem of
probability. The experimental results show that our proposed method is efficient to
detect complex events over probabilistic event streams with better event
throughput capabilities and lower time consumption.

SunP03-05 SunP03-06

Detecting cracks on a concrete surface using histogram of Power balance control for switched reluctance generator
oriented gradients integrated in DC microgrid
T Wang., D He. (Lanzhou Institute of Physics and National Key Lab. of Vacuum and CryogenicLanzhou Hongrui Aerospace
Lin Meng, Zhongkui Wang, Yoshiyuki Fujikawa, and Shigeru Oyanagi
.
Electromechanical Equipment Co. Ltd Wuxi Hongrui Aerospace Electromechanical Equipment Co. Ltd., China) Q Wang . (China Univ. of
(Ritsumeikan Univ., Japan) Mining and Tech., China) Y Huo . (Lanzhou Institute of Physics and National Key Lab. of Vacuum and Cryogenic Lanzhou Hongrui
Aerospace Electromechanical Equipment Co. Ltd., China) L Lu. and J Li. (China Univ. of Mining and Tech., China)

With the aging of structures such as bridges, buildings, and tunnels, investigating This paper presents a power balance control (PBC) scheme for switched
the cracking of concrete surfaces have become very important. The standard reluctance generator (SRG) wind power system integrated with DC microgrid.
method for investigating cracks has usually been to perform a visual inspection First, a switched reluctance generator wind power system is, which acts as
and acoustic diagnosis. However, the inspection results when using these distributed energy source, is integrated with a 24V utility DC microgrid. With
techniques vary depending on the experience of the inspector, causing there are proper convertor design and control characteristic for SRG system, output voltage
not objectivity and quantifiability enough. Therefore, digital image analysis of ripple is minimized, which makes it possible for directly connection. Then
concrete cracks is being developed with the aim of providing efficient, automatic, proposed PBC is achieved on DSP controller for a 3 phase 12/8 SRG system. With
and quantitative inspection. However, in the process of binarization, some of the proposed control scheme and control system, SRG can change control target
cracks become fractured or invisible, which have a negative effect on the dynamically with signals on DC bus. In the end, both simulations and experiments
investigation accuracy. To improve the accuracy when detecting cracks, we demonstrated proposed PBC well.
propose a method that uses the histogram of oriented gradients (HOG) to detect
cracks on the concrete surface. We are also working on the use of unmanned
helicopters to take photos of the concrete surfaces to enable automatic
investigation.
AUTHORS INDEX
K = Keynote speaker, C = Chair, CC = Co-Chair, O = Organizer
Name Session Page SatP01-05 589 Hong, Zhipan SunP01-03 519

A Di, Long SatM01-04 15 Hou, Zengguang SatM02 CC

Ali, Mumtaz Poster_session-01 249 F SatM02-05 50

B Fang, Jianjun SatE03-02 228 Poster_session-23 365

Basari, Amat SatP03-04 154 SatE03-05 244 Hu, Bo SatE02-03 206

Bi, Shuhui SatM03 C Feng, Yongfei SatM02-01 24 Hu, Ming SunP01-01 508

SunP02-06 562 Fu, Xin SunP01-01 508 SunP03-01 567

Bian, Guibin SatM02-05 50 SunP03-01 567 Huo, Feng SatP02-02 114

Poster_session-23 365 Fu, Yiwen SunP03-02 571 Huo, Y. SunP03-06 594

Bu, Ni SatM03-04 70 Fujii, Ryohei SunP02-02 542 I

SunM03 CC Fujikawa, Yoshiyuki SunP03-05 103 Iliescu, Mihaiela SatM02-04 42

C Fujita, Kou SunP02-01 538 Inoue, Akira Poster_session-15 322

Cang, Shuang SatM02-05 50 Funaki, Daishi SunM01-03 435 Poster_session-16 328

Cao, Songming SunP01-01 508 G Ito, Kazuyuki SatP03 CC

SunP03-01 567 Gal, Ionel Alexandru SatM02-03 36 SunM02-01 450

Chai, Tianyou PS01 K Gao, Chengchong SatM01-01 1 J

Chang, Yiping Poster_session-20 349 Gao, Feng SatM02-04 42 J., Mukund Nilakantan SatP02-03 119

Chen, Li Poster_session-22 359 Gao, Rongxiang SunP02-04 552 Jaworski, Johnathan SatP02-01 108

Chen, Qun SatM01-03 10 Gao, Yan Poster_session-12 306 Jiang, Changan SunP01 C

Chen, Xinkai PS02 C Ge, Weiliang SunM02-03 461 Poster_session C

SunM02 C Ge, Wenjie Poster_session-24 370 SunP01-06 533

SunM02-04 467 George, Q. Huang SatP02-03 119 Jiang, Wei SatP03-04 154

Cheng, Wei Poster_session-25 375 Guo, Lei Poster_session-06 277 Jiang, Yuqi SatP03-04 154

Poster_session-26 380 Poster_session-07 282 Jin, Longguo SatP01-05 589

Poster_session-27 384 Poster_session-08 286 Jin, Qi SatM03-02 61

Cong, Sen SatE03-05 244 Guo, Weizhong SatM02-04 42 John, Cross SatE01-03 174

Cui, Tengteng Poster_session-28 388 Guo, Zhiqiang Poster_session-30 397 Joseph-Ellison, Stacey SatP02-01 108

D H K

Dai, Aini SunM03-05 503 Hanawa, Tomohito SunP01-05 528 Kainose, Sho Poster_session-10 295

Deli, Irfan Poster_session-01 249 Hao, Yongxing Poster_session-05 273 Kanagaraj, G. SatP02-04 125

Deng, Mingcong SatM02-03 36 Hasan, Mohammad Shahidul SatP02-05 131 Kasabov, Nikola Poster_session-23 365

SatM02-04 42 Hashimoto, Seiji SatP03-04 154 Kimura, Mami SunM01-03 435

SatM03-03 66 Hashimoto, Shin'ichi SatE01-02 170 Kimura, Soichiro SunM02-01 450

SatM03-04 70 Hatano, Takafumi Poster_session-19 343 Kubota, Masayuki SatP03-05 158

Poster_session-15 322 Hayakawa, Hayami SunM01-01 423 L

Poster_session-16 328 Hayakawa, Yoshikazu SunP01-06 533 Li, Chuang SatM01-02 5

Poster_session-19 343 Hayashi, Mitsuharu SatE02-02 200 Li, Gensheng Poster_session-29 392

SunM01-04 441 He, D. SunP03-06 594 Li, J. SunP03-06 594

SunP01-04 523 He, Kai Poster_session-06 277 Li, Ke Poster_session-12 306

SunP01-05 528 He, Wei SunM02-03 461 Li, Lifeng SatM03-02 61

SunP02-01 538 Henmi, Tomohiro Poster_session-15 322 Li, Lina Poster_session-14 316

SunP02-02 542 Poster_session-16 328 Li, Ming SunP01-04 523

SunP02-03 548 Hirai, Junji SunM01-02 429 Li, Wei Poster_session-34 418

SunP03-03 577 Hong, Geoksoon SatP02-02 114 Li, Ximei SatM01-05 20

Li, Xuehong Poster_session-18 337 Mo, Xinhu Poster_session-07 282 Shang, Weiwei SatM01-02 5

Li, Yongcan Poster_session-09 290 Mohammadirad, Amin SatE02-01 194 Shao, Junzhen SatP02-01 108

Li, Yuan SatE03-05 244 Munteanu, Ligia SatM02-05 50 Shao, Xinyu SatP01-02 86

Li, Yunchuan SunM02-03 461 Munteanu, Radu SatM02-02 30 Sharmin, Sonia SatE01-06 190

Li, Yushan SatE03-03 233 N Shen, Kenan SunP01-04 523

Li, Zhiqiang SatM03-01 56 Nagasaka, Ken SatE02 CC Shi, Juanjuan SunM02-05 473

Lian, Zhongxu SatE02-05 217 SatE02-01 194 Shi, Xudong SatP02-01 108

Liang, Ming SunM02-05 473 SatE02-02 200 Shinoda, Shogo SunM01-02 429

Liang, Yongchun Poster_session-09 290 Poster_session-04 267 Smarandache, Florentin Poster_session-01 249

Liao, Wudai SunP02-06 562 Poster_session-10 295 Song, Dongliang SatM01-05 20

SunP03-02 571 Nakade, Keisuke SatP03-02 143 Song, Fei Poster_session-13 310

Linh, Nguyen Manh SunM02-04 467 Nakamura, Taiki SatP03-01 137 Song, Lei Poster_session-31 402

Liu, Hong SatP02-01 108 SatP03-02 143 Song, Lihui Poster_session-05 273

SatE03-04 238 Nakamura, Yukinori SatP02 CC Song, Shujie Poster_session-17 332

Liu, Hongzhao Poster_session-31 402 SatP03-01 137 Song, Xiaonan Poster_session-32 408

Liu, Junhong SatM03-02 61 SatP03-02 143 Song, Yuan Poster_session-06 277

Liu, Shanzhong Poster_session-32 408 SatP03-03 149 Poster_session-07 282

Liu, Song SatE02-03 206 SatP03-05 158 Poster_session-08 286

Liu, Wanji SunP01-02 513 SunM01-03 435 Spirleanu, Cristian SatM02-04 42

Liu, Wenxue SatM03-02 61 O Su, Chun-Yi PS02 K

Liu, Xiangdong SunM03-05 503 Obata, Shogo SunM01-01 423 T

Liu, Yibin SatM01-02 5 Oyanagi, Shigeru SunP03-05 103 Tan, Jie SunP03-04 583

Liu, Yong SunP01-01 508 P Tan, Min Poster_session-14 316

SunP03-01 567 Peng, Likun Poster_session-13 310 Tang, Qili Poster_session-33 413

Liu, Yukun SatP01-03 90 Peng, Long Poster_session-23 365 Tao, Fazhan SatM03-03 66

Liu, Zhengbo SunP02-05 557 Ponnambalam, S.G. SatP02-03 119 Tao, Xian SunM03-02 485

SunP02-06 562 SatP02-04 125 SunP01-02 513

Long, Hao Poster_session-17 332 SatE02 C Tian, Bailing SunM03-03 491

Loo, C.K. SatP02-04 125 Poo, Aunneow SatP02-02 114 Tian, Guangdong SatP01 C

Lou, Jingjun Poster_session-13 310 Q SatP01-01 79

Lu, Guoliang SunP01-03 519 Qian, Dianwei SatE03 CC SatP01-02 86

Lu, Hao SatM02-01 24 SatE03-02 228 SatP01-04 96

Lu, L. SunP03-06 594 SatE03-04 238 Tong, Shiwen SatE03 O, C

M Qiu, Rongxian SatE02-04 212 SatE03-02 228

Ma, Haiyan Poster_session-26 380 Qiu, Shenghai SatM01-01 1 SatE03-03 233

Poster_session-27 384 R SatE03-05 244

Ma, Zhihao SatM01-02 5 Ren, Junjie SatE03-03 233 Tong, Xinxing Poster_session-24 370

Maciejewski, Anthony A. PS03 K Ren, Yaping SatP01-01 79 U

Mao, Na SunP03-04 583 S Ueno, Satoshi SunP01-06 533

Masuda, Shiro SunM01 C Sbastien, Louvign SatE01-05 185 Uma, Rajagopalan Maheswari SatE01-04 179

SunM01-05 445 Shi, Jie SatE01 O, C V

Melinte, Octavian SatM02-05 50 SatE01-01 164 Vladareanu, Luige SatM02 O, C

Meng, Lin SunP03-05 103 SatE01-02 170 SatM02-01 24

Mihai, Razvan-Viorel SatM02-01 24 SatE01-03 174 SatM02-02 30

Mineo, Hirokazu SatP03-05 158 SatE01-04 179 SatM02-03 36

Ming, Zhengfeng Poster_session-03 261 SatE01-05 185 SatM02-04 42

Minh, Tran Vu SunM02-04 467 Sang, Bingfeng SatP01-03 90 SatM02-05 50

 Poster_session-23 365 Watanabe, Marino SatP03-03 149 Yashiro, Daisuke SunM01-02 429

W Wei, Jun SatM01-01 1 Yin, Yixin Poster_session CC

Wakimoto, Shuichi SunP02-01 538 Wei, Shimin Poster_session-08 286 SunM02-02 456

Wakitani, Shin Poster_session CC SunP03-01 567 Yoshida, Ryuji SunP02-03 548

Poster_session-19 343 Wen, Ci SatP02-01 108 Yu, Dongsheng SatE03-01 223

SunM01 O, CC Wen, Shengjun SunP02 C Yu, Hongnian PS01 C

SunM01-04 441 SunP02-04 552 SatM02-03 36

SunP02-01 538 SunP02-05 557 SatM02-04 42

SunP02-02 542 SunP02-06 562 SatM02-05 50

SunP02-03 548 Wen, Tao Poster_session-03 261 SatP02 C

Wakui, Shinji SatP03 C Wu, Ganggang SatM01-04 15 SatP02-05 131

SatP03-01 137 Wu, Jian SatM01-02 5 Poster_session-23 365

SatP03-02 143 Wu, Yanfeng SatP01-05 589 SunP03-03 577

SatP03-03 149 X Yu, Huadong SatE02-04 212

SatP03-05 158 Xia, Kui SatE02-04 212 SatE02-05 217

SunM01-03 435 Xiao, Huimin SatM03 CC Yu, Yan SatE01 CC

Wang, Aihui SunP03 C SatM03-01 56 SatE01-01 164

SunP03-02 571 Xiao, Junming SunP01 CC SatE01-02 170

SunP03-03 577 SunP02-05 557 Yu, Zhanjiang SatE02-04 212

Wang, Aiping SunM03-04 497 Xiao, Yongfei Poster_session-25 375 SatE02-05 217

Wang, Bianli Poster_session-12 306 Xie, Jinfa Poster_session-29 392 Yu, Zongyuan Poster_session-12 306

Wang, Chaofeng Poster_session-05 273 Xie, Jinli SatM03-05 75 Yuan, Ying SatM01-04 15

Wang, Dongyun SunP02 CC Xie, Liyu SatM02-02 30 Yubai, Kazuhiro SunM01-02 429

SunP02-06 562 Xie, Zhanpeng SatP01-02 86 Z

SunP03-03 577 Xing, Bin Poster_session-08 286 Zan, Xiaoshu SunM03-01 479

Wang, Fenggui Poster_session-27 384 Xu, Bin SatM01-02 5 Zhan, Haiyan SatE01-02 170

Wang, Haiquan SunP02-04 552 Xu, De Poster_session-14 316 Zhang, Chaoyong SatP01-02 86

Wang, Hong SunM03-03 491 SunM03-02 485 Zhang, Chi SunM03-05 503

SunM03-04 497 Xu, Jinkai SatE02-04 212 Zhang, Dong SatM02-01 24

SunP03 CC SatE02-05 217 Zhang, Feng SunM03-02 485

Wang, Hongbo SatM02-01 24 Xu, Liyou Poster_session-29 392 Zhang, Haigang SunM02-02 456

SatM02-02 30 Poster_session-30 397 Zhang, J. X. PS03 C

SatM02-03 36 Xu, Peng SatM02-01 24 Zhang, Jiazhen Poster_session-30 397

SatM02-05 50 Y Zhang, Linshuai SatE02-04 212

SunM02 CC Yamada, Manabu SunM01-01 423 Zhang, Qian SunP02-04 552

Wang, Hui Poster_session-20 349 Yan, Tongbin SunP03-02 571 Zhang, Qichun SunM03-04 497

Wang, Q. SunP03-06 594 Yan, Xiaomo SunM03-03 491 Zhang, Sen SunM02-02 456

Wang, Shuaishuai SatM02-02 30 Yang, Chenguang SunM02-03 461 Zhang, Tong Poster_session-03 261

Wang, T. SunP03-06 594 Yang, Fen Poster_session-03 261 Zhang, Wen SatM02-02 30

Wang, Tianjun Poster_session-28 388 Yang, Shuhua Poster_session-18 337 Zhang, Xiangming SunP02-05 557

Wang, Xingang Poster_session-14 316 Yang, Xue SatM02-02 30 Zhang, Xinxin SatE02-05 217

Wang, Xinrui Poster_session-34 418 Yang, Zongxiao SatM01 O, C Zhang, Yi SatM01-04 15

Wang, Xuelin Poster_session-26 380 SatM01-03 10 Zhang, Ying Poster_session-07 282

Wang, Yunxia SatM01-01 1 SatM01-04 15 Poster_session-08 286

Wang, Zhiming Poster_session-22 359 SatM01-05 20 Zhang, Yinong SatE03-03 233

Wang, Zhongkui SunP03-05 103 SatP01 CC Zhang, Yonghong Poster_session-24 370

Wang, Zhoufeng Poster_session-33 413 Poster_session-31 402 Zhang, Yunzhang Poster_session-33 413
Zhang, Zhengtao SunM03-02 485

Zhang, Zhi Poster_session-11 301

Zhao, Bonian Poster_session-12 306

Zhao, Hongpeng Poster_session-12 306

Zhao, Jianyu SatM03-05 75

Zhao, Qinjun SatM03-05 75

Zhao, Youlei Poster_session-25 375

Zheng, Ciyan SatE03-01 223

Zhong, Xiao SatE02-03 206

Zhou, Huiling SunM03-05 503

Zhou, Mengchu SatP01-01 79

SatP01-04 96

Zhou, Qi SatP02-01 108

Zhou, Qixian Poster_session-28 388

Zhou, Xiaoguang Poster_session C

SunM03-05 503

Zhou, Zhili Poster_session-30 397

Zhu, Hong SatE03-01 223

SunM03-01 479

Zhu, Zhanxia Poster_session-03 261

Zulati, Litifu Poster_session-02 255

Poster_session-04 267

Poster_session-21 353
Proceedings of the 2015 International Conference on
Advanced Mechatronic Systems, Beijing, China, August, 22-24, 2015

Cluster Analysis on the Service Oriented Alliance

Enterprise Manufacturing Resource

Yunxia Wang , Shenghai Qiu, Chengchong Gao , Jun Wei
Department of Mechanical Engineering
Nanjing Institute of Technology
Nanjing ,China
[email protected], [email protected],[email protected], [email protected]

AbstractTo alliance manufacturing enterprises, there are Generalized intelligent resources refer to all resources
some problems in resources such as different kinds, ununiform within the enterprise product life cycle, which is the
standards, high repeatability etc. Under cloud manufacturing application of artificial intelligence technology to resource
service platform, using cluster analysis method to classify aggregation, searching, matching and scheduling. According
intelligent resources of alliance enterprise was studied. The same to the different needs of users, resources can be fast searched
or similar resources are divided into one resources cluster. The and matched services efficiently; the manufacturing resources
uniqueness attribution of resources is solved. According to the have the characteristics of intelligent under the help of
cloud user needs, resources in a cluster not in all resources set are information technology, computer technology and other
searched and matched. Using the evaluation method of supports. Each enterprise classify resources according to use.
manufacturing dynamic capacity, optimal resources is provided. Resources can be divided into eight kinds, including service,
It can improve use ratio of the resources. Finally an instance was human, material, equipment, software, cooperation, logistics
verified.
and data[1]. Part classification is as shown in figure 1.
KeywordsCloud manufacturing; intelligent resources;
cluster analysis; resource cluster; manufacturing capacity
. INTRODUCTION
With the rapid development of information technology and
network technology, cloud manufacturing mode is gradually
moving towards the wisdom cloud manufacturing.
Manufacturing resources transform into intelligent
manufacturing resources (referred to as intelligent resources).
Intelligent resources not only inherit characteristics of the
traditional manufacturing resources, but also reflect the
characteristics of intelligence and service etc. Manufacturing
resources are numerous kinds, rich varieties and no standard in
classifying and describing to all enterprises. It leads to
Fig. 1 Classification of enterprise resources
resource semantic ambiguity, repetition; the same resource
Different resources have different characteristics and
representation is not consistent. Using cluster analysis method,
intelligent resource of alliance enterprise were aggregated and attributes, such as processing equipment attributes including
categorized. And the uniqueness attribution of resources is processing size, processing price, processing type and
processing precision. In the fierce market competition
judged. According to the different needs of cloud users,
dynamic manufacturing capability evaluation method was environment, the same industry or related enterprises will
form the enterprise alliance in the strategic objectives, through
used to achieve the optimal allocation of resources and
complementary and exchange, to obtain long-term competitive
services.
advantage in the market. There are different forms such as
. INTELLIGENT RESOURCES CLUSTER price alliance, logistics alliance, promotion alliance, and
product or service alliance. One of the most deep-seated
ANALYSIS METHOD alliances is to realize the service alliance, including resources
sharing, matching and restructuring. All manufacturing
A. Intelligent resource classification resources of alliance enterprise through layers of classification,
the same or similar resources are aggregated. The clustering
This research is sponsored by Ministry of Education, Humanities and Social process is shown in figure 2.
Sciences Project of China (No. 12YJCZH209, 12YJAZH151)

978-1-4673-7995-3/15/$31.00 2015
c IEEE 1
 a) The clustering samples transform into feature
metrics matrix of Vk. Matrix description for:
V11k V12k V1km
% k k
%V21 V22 V2km
Vk V (O k , A k )
%    
% k k
&%Vi1 Vi 2 Vimk #
b) Normalization
Due to the nature of different metrics, the different index
dimension cannot be compared, so it is necessary to normalize
the matrix. By using the average value function and the
Fig. 2 the cluster and service processing of alliance standard function was normalized to obtain matrix, which
enterprises resources consists of vij :
Because there is no uniform standard of enterprise m
resources, resource semantics may exist not uniqueness. So 1
using cluster analysis method to classify resources, the same
Vjm Vijm (1!10
c i 1
or similar resources are classified as a group and storage in
cloud manufacturing virtual resource pool. According to the Vijm
different needs of cloud users to search for resources in the vij m
(1!20
pool, and then manufacturing capacity of the selected V j
resources are evaluated. It will meet the requirements and the c) Determine the similarity matrix
strongest resources available to users in the form of services,
to achieve efficient use of resources and dynamic matching. There are a variety of methods to determine the similarity
coefficient, such as Euclidean distance method, integral
B. THE APPLICATION OF CLUSTER ANALYSIS
method, maximum minimum method, and minimum
METHOD IN INTELLIGENT RESOURCES
geometric average method. The similarity coefficient rij is
Cluster analysis is to classify things according to certain obtained by maximum minimum method.
rules, the method used in resources mainly include three
m m
steps: classification, determine the metrics and clustering (vik v jk ) min(vik , v jk )
analysis[2]. According to use classification, all kinds of
rij k 1 k 1 (1!30
resources are divided in detail hierarchically. Such as m m
equipment resources according to the processing methods (vik ) v jk ) max(vik , v jk )
can be divided into turning, milling, and planning, drilling, k 1 k 1
boring and other resources groups. Common attributes of After the above calculation, the similarity matrix is
group are the metrics index, the same or similar resources composed of n resource collection objects, as shown below:
of the alliance enterprises were classified by clustering r11 r12 r1n
algorithm. The basic definition is as follows:
1) The classification of resources r21 r22 r2 n
R (rij ) n+n
Select a kind of resource as clustering samples, and    
defines the resource set Rs= {Rs1, Rs2,,Rsn}, where n is
rn1 rn 2 rnn
the number of resources object. According to the
characteristics and attributes of resources, the elements of The matrix is reflexive, symmetric and transitive etc...
Rs are divided into groups O= {O1, O2,, Ok}; the number d) Clustering analysis
of resources is qi, then According to the similarity matrix and cluster size
k
parameters
(
is real, and
. ,0,1 ), resources are clustered
qi n (2!/0 into several groups. The result has a great relationship with the
i 1 value of
, too large or too small will make the results
2) Determine the index meaningless.
Different objects have different metrics index. According e) Attribution of uniqueness
to the specific characteristics of resources determine the After the clustering, the same resource may exist in two or
common features for metrics. The index should follow certain more than two groups, the result are not unique. Therefore,
criteria, such as key features, quantitative characteristics. The attributions of uniqueness should be determined. The average
index feature vector Ak corresponds with the group Ok . A value method and cluster analysis method are combined to
mathematical description model is: determine samples uniqueness of each cluster resources, until
Ak { A1k , A2k ,..., A pk } there are no public resources.
3) Clustering analysis

2
 manufacturing resources, which generate group O:
. EVALUATION OF INTELLIGENT
Rs={Rs1,Rs2,,Rs8}={MK215,MK2110,MK2620,MK211
MANUFACTURING RESOURCES CAPABILITY 5,MK2120B,MK2150,MK2110,MK2860}
The manufacturing capability of cloud manufacturing Select 7 static capabilities such as the maximum and
mode is the ability to manufacture configuration and minimum grinding diameter, grinding depth, machining
integration of resources, reflects the level of the precision and total power as clustering metric index, the
manufacturing enterprises to complete a task and target[3]. formation of an ordered set:
Such as design capability, manufacturing capability, A= {A1,A2,,A7}
simulation capability etc... According to the characteristics of The measurement index of each sample is mapping Eigen
manufacturing resources, manufacturing capabilities are value matrix elements in the Vkij, the specific value as shown
divided into static capabilities and dynamic capabilities. Static in table 1.
capability is the inherent attribute of the manufacturing Using the formula (2-1) and (2-2), the matrix has been
resources, such as equipment processing size, machining normalized:
accuracy; dynamic capability of manufacturing resources is 0.113 0.211 0.281 0.301 0.410 0.8 1.013
influenced by environmental factors and generated by some %0.282 0.421 0.439 0.635 0.615 0.8 1.266
dynamic information. Such as equipment operators, state %
%0.845 0.842 2.632 2.540 2.051 1.6 0.405
information, processing price and other factors. For the same %
manufacturing resource alliance, static manufacturing T % 0.282 0.632 0.351 0.952 0.821 0.8 1.266
capability as metrics of clustering analysis, cluster resource v
% 0.563 0.842 0.526 0.635 0.821 0.8 1.266
into groups. Due to various aspects of enterprises scale, %
management level and other factors, the dynamic %2.817 2.105 1.579 0.952 1.026 1.6 1.013
manufacturing capability of the same resources is not the same. %0.282 0.421 0.439 0.635 0.615 0.8 1.266
Therefore, in the face of cloud users demand, in resource %
search and matching, to fully consider the dynamic %&2.817 2.526 1.754 1.270 1.641 0.8 0.506#
manufacturing capacity and evaluate, select the optimal
allocation of resources. The application of AHP and fuzzy The maximum and minimum value method determines the
mathematics to evaluate the dynamic manufacturing resources similarity coefficient of symmetric matrix:
and manufacturing capabilities, the specific steps are as 1 0.702 0.219 0.613 0.574 0.282 0.697 0.222
follows. % 1 0.305 0.842 0.818 0.371 1 0.306
%
1) Determine the dynamic manufacturing capability % 1 0.36 0.39 0.491 0.305 0.515
measurement B %
% 1 0.83 0.428 0.842 0.36
The index feature vector Bk corresponds with the group Ok, R
% 1 0.458 0.818 0.389
mathematical description: %
% 1 0.371 0.775
Bk {B1k , B k2 ,..., B kp } % 1 0.306
The same resource capability index of different enterprises %
&% 1 #
is normalized, as mentioned before. The index such as
Boolean type, state type can not be described quantitatively,
use {0, 1} to deal with. For example, if the equipment is idle According to symmetric coefficient matrix and cluster size
for 1, processing is for 0. parameters
, resources were classified. Because the
2) The application of AHP to determine the weight of clustering parameter is larger, clustering is fine, and vice versa.
index value.
p
So the average value of similarity R was selected as
, that
 {1
 2

 p }, f 1 (2!/0 is:
f 1 1 n n

3) Calculation of similar resources dynamic

R ( rij ) (3!/0
n2 i 1 j 1
manufacturing ability level

Was calculated to be 0.581, get the similar triangle
 (B p )B p (2!10 matrix:
According to the dynamic manufacturing capability level, 1 1 0 1 0 0 1 0
% 1 0 1 1 0 1 0
the optimal resources are provided to cloud users. %
IV. CASE ANALYSIS % 1 0 0 0 0 0
%
% 1 1 0 1 0
Manufacturing resources of three alliance enterprises of a r
certain industry were selected. CNC internal grinder % 1 0 1 0
%
equipment will be analyzed, grinding machine information as % 1 0 1
shown in table 1: % 1 0
%
An ordered set of Rs is composed of a sample of &% 1#

3
 Table . CNC internal grinder information of alliance enterprises
Enterprise Type Min Max Grinding Maching precision(m) Total
grinding grinding depth power
diamater diamater
mm Roundness Cylindricity Roughness
kw

mm
mm

Alliance MK215 6 50 80 1.2 2 0.4 20

Enterprise 1 MK2110 15 100 125 2 3 0.4 25
MK2620 45 200 750 8 10 0.8 8
Alliance MK2115 15 150 100 3 4 0.4 25
Enterprise 2 MK2120B 30 200 150 2 4 0.4 25
MK2150 150 500 450 3 5 0.8 20
Alliance MK2110 15 100 125 2 3 0.4 25
Enterprise 3 MK2860 150 600 500 4 8 0.4 10

Therefore the sample space can be divided into 4 clusters: Through the evaluation of resource dynamic manufacturing
C1= {Rs1, Rs2, Rs4,Rs7}, C2= {Rs2, Rs5, Rs7}, C3= {Rs3}, C4= capability in the cluster resources, optimal allocation of
{Rs6,Rs8}. There are public elements Rs2 and Rs7 in cluster C1, resources was realized.
C2, which need to redetermine the uniqueness of ownership.
V Conclusions
C1, C2 are made as the sample, and the cores C1, C2 index
value are determined. The specified metric average value was The intelligent of resources reflected has high integration
defined as the core value: and clustering ability, can fast searching and matching
C1= {12.75, 100,107.5, 2.05, 3, 0.4, 23.75} according to user needs. The correct and effective resource
C2= {20,133.33, 133.33, 2, 3.33, 0.4, 25} classification can improve the use ratio of resources. The
Common elements Rs2 and Rs7 with C1 and C2 form a clustering analysis method is applied to the classification of
cluster of new samples, the samples using the clustering intelligent resources. The ownership of resources and
method analysis again, obtain the similarity matrix: uniqueness is judged. Dynamic manufacturing ability will be
1 1 0.948 0.8 evaluated in the same resource. And the most matching
% 1 0.948 0.8 service resources are selected. The method can change the
% original search strategy, to reduce the search space, greatly
s improve the efficiency for better personalized needs service.
% 1 0.761
%
& 1 #
ACKNOWLEDGMENT
According to the similarity matrix calculate
, so
is This research is sponsored by Ministry of Education,
0.907. The new clustering samples can be divided into two Humanities and Social Sciences Project of China (No.
clusters: {Rs2,Rs7, C1} and {C2}, then the uniquessness was 12YJCZH209, 12YJAZH151)
judged. There are no public resources in a cluster of elements. REFERENCES
Finally the resources sample classification was get: [1] Yunxia Wang, Shenghai Qiu, Zhiliang Wang. A new manufacturing
C1={Rs1, Rs2, Rs4, Rs7}={MK215,MK2110, MK2115, model for servitization-cloud manufacturing overview[J].Modern
MK2110}, C2={Rs5}={MK2120B}, C3={Rs3}={MK2620}, Manufacturing Engineering,2013(3):124-128.
[2] Huanmin Xu, Minghai Yuan. The research and application of reverse
C4={Rs6,Rs8}={MK2150, MK2860}. When the cloud user chain decision logic method for process planning[M]. Beijing: Beijing
request resources, according to the needs of the user Institute of Technology Press,2011.
information, the search space is reduced to a cluster of [3] Yongliang Luo, Lin Zhang, Fei Tao,et al. Key Technologies of
resources, can greatly improve the search efficiency. When Manufacturing Capability M Deling in cloud Manufacturing
there are multiple identical or similar resources in a cluster, Mode[J].Computer Integrated Manufacturing
System,2012,18(7):1357-1367.

4
Proceedings of the 2015 International Conference on
Advanced Mechatronic Systems, Beijing, China, August, 22-24, 2015

Performance and Exhaust Emissions of a

Turbocharged Common Rail
DI Diesel Engine Fueled with Alcohol

Bin Xu, Yibin Liu 1, Jian Wu, Chuang Li, Weiwei Shang, Zhihao Ma
College of Vehicle and Transportation Engineering
Henan University of Science and Technology
Luoyang, China
1
[email protected]

AbstractThree kinds of alcohol fuels and diesel were 3) High evaporative cooling, which results in a cooler
blended and combusted in an electronic control high-pressure intake process and compression stroke. This raises the
common rail turbocharged DI diesel engine respectively. Without volumetric efficiency of the engine and reduces the required
any modifications in the original engine structure, the work input in the compression stroke.
combustion characteristics and emissions of alcohol-diesel fuel
were investigated. It shows that the peak heat release rate and the 4) High laminar flame propagation speed, which may make
brake specific fuel consumption increases with the increase of combustion process finish earlier, thus improve engine thermal
mixing ratio. At the same blending ratio, the peak heat release efficiency [7-9].
rate of methanol-diesel is the highest, but the brake specific fuel
consumption is lower; N-butanol/diesel has the lowest CO and Significant amount research has been carried out on the
HC emissions but has the highest NOX emissions. NOX emissions performance and emission characteristics of diesel engines
of methanol-diesel and ethanol-diesel are approximated, but CO fuelled with alcoholdiesel blends. Najafi and Yusaf.[10]
and HC emissions of ethanol-diesel is the highest. In examined the consequences of using methanoldiesel fuel
consideration of the fuel economy and emission performance, blends in diesel engine at mixing ratios of 10:90, 20:80 and
methanol-diesel mixing fuel is better. 30:70.The output power and torque of neat diesel fuel was
found to be lower than those of methanoldiesel blends. The
Keywordsalcohol fuel, combustion characteristics, common brake specific fuel consumption (BSFC) didnt vary
rail engine, emissions. significantly, but the lowest was associated with the 30%
methanol blend; Huang et al. [11] studied the combustion
I. INTRODUCTION behavior of a CI engine fueled with dieselmethanol blends
under various fuel delivery advance angles. The heat release
Diesel engines are the most common internal combustion rate in the premixed burning phase increased and the
engines that are widely used in various sectors such as combustion duration of the diffusive burning phase both
agriculture, transportation and industry, due to their excellent decreased as the methanol mass fraction in the dieselmethanol
drivability and fuel economy. They are expected to become blends increased.
more widespread in the foreseeable future. The increasing
concern about the energy demand, depleting oil reserves and Koganti et al. [12] studied the effect of a 5% ethanol diesel
the instability of the prices of fossil fuels, together with fuel blend in a three cylinder, four-stroke DI tractor engine.CO
environmental and health problems related to atmospheric and PM emissions were reduced, but NOx emissions were
pollution, are impelling the search of alternative cleaner fuels increased. Further, slight drops in engine power and torque
that can alleviate such problems. Among the different were noted with a corresponding loss of fuel economy.
oxygenated compounds that have been studied in order to be Rakopoulos et al. [13] studied the effects of ethanol blends (5%
used as substitutes and additives for conventional fuels, and 10% by volume) compared with conventional diesel fuel,
alcohols such as methanol, ethanol and butanol are competitive on the performance and emissions of a turbocharged, DI diesel
alternative fuels. [1-6] engine. CO and NOx emissions decreased, while THC
emissions increased with the use of ethanol. BSFC increased
The advantages of alcohol as alternative fuel include: and BTE decreased as the amount of ethanol increased.
1) Low viscosity compared to diesel fuel, therefore it can Rakopoulos et al. [14] studied the effects of n-butanol (8%,
easily be injected, atomized and mixed with air. 16% and24%)diesel fuel blends on the performance and
2) Less emission because of its high stoichiometric fuelair emissions of a high speed DI diesel engine operating at a speed
ratio, high oxygen content, high H/C ratio and low sulfur of 2000 rpm under three different loads (low, medium and
content. high). Smoke density was significantly reduced while NOx and

978-1-4673-7995-3/15/$31.00 2015
c IEEE 5
CO emissions were slightly reduced with the use of the Tab. II. The properties of the test fuels of diesel, methanol,
butanoldiesel fuel blends compared to the neat diesel fuel. ethanol and n-butanol used in this experiment are presented in
However, the HC emissions increased with the use of the Table . The fuel M10, M20, E10, E20, N10, N20 and D100
butanoldiesel fuel blends. As the percentage of butanol in the are used in the experiment respectively. Load characteristic
blends increased, marginally higher specific fuel consumption test of the seven kinds of fuels are taken at the 2000r/min and
was observed with a correspondingly slight increase in BTE. 3000r/min without changing the original engine structure,
Dogan [15] investigated the performance and exhaust emissions
comparatively analyze the economy and emission
of a single cylinder, four-stroke, naturally aspirated DI diesel
performance data under different operating conditions,
engine at a constant engine speed (2,600 rpm) under four
different engine loads using n-butanol (5%, 10%, 15% and comparison the effect of different fuel blending ratio and
20%) blended diesel fuel, and found that BSFC and BTE were alcohol on combustion and emissions of diesel engine.
slightly increased with increasing n-butanol content in fuel
blends while exhaust gas temperature, NOx, smoke opacity and
TABLE III. COMPARISON OF VARIOUS PROPERTIES OF METHANOL,
CO emissions were reduced. However, HC emissions were ETHANOL AND BUTANOL WITH THOSE OF DIESEL FUELS. [1]
increased.
Property Methanol Ethanol Butanol Diesel
The most common methods for achieving dual fuel Density at20 C (g/ml) 0.7866 0.7851 0.8098 0.829
operation are alcoholdiesel fuel blends, alcoholdiesel fuel Boiling point,(C) 65 78 117-118 187-343
emulsions, dual injection, and alcohol fumigation [1]. This study Cetane no. 3.8 5-8 25 40-55
use alcohol in diesel engine is in the form of blends, Lower heating 20.1 26.9 33.1 42.8
value(MJ/kg)
investigates and compares the combustion and emission
Higher heating 22.7 29.7 36.6 44.8
performance of diesel engine when mixing methanol, ethanol value(MJ/kg)
and butanol respectively. Freezing point,(C) -97 -114 -89.3 -40-34
Flash point, (C) 12 13 35 74
Auto ignition temp (C) 463 423 397 235
TABLE I. MAIN TECHNICAL SPECIFICATIONS OF THE TEST ENGINE
Heat of 1.2 0.92 0.43 0.23-0.6
Description Specification vaporization(MJ/kg)
Engine model GW4D20
Cylinder 4
Bore 83.1mm
Stroke 92mm III. RESULTS AND DISCUSSION
Displacement 1.996L
Compression ratio 16.71 A. Combustion characteristics
Rated output & speed 110kW&4000r/min Fig.1 shows the variations of combustion heat release rate
Injection system Common rail injection system
for different alcohol ratios at 2000 rmin-1/270 Nm and 3000
Maximum torque & speed 310Nxm&1800~2800r/min
Maximum injection pressure 1600bar rmin-1/270 Nm respectively. It can be observed from Fig. 1
that there are two wave crests in heat release rate curves, both
of the peak values decrease with the increase of revolving
speed. This is because with the increase of revolving speed,
II. TEST FUEL AND LABORATORY EQUIPMENT time through the same crank shaft angle decreases, reduce the
Fuels used in test are conventional 0# diesel, analysis-level amount of combustible mixed gas.
anhydrous methanol, analysis-level anhydrous ethanol and n- Under the same operating condition and mixing ratio, heat
butyl alcohol. On the basis of diesel oil, configure the volume release rate peak value of methanol-diesel fuel increases
content of 10% and 20% mixed fuel respectively. M10 and significantly, ethanol-diesel fuel is the next, n-butanol/diesel
M20 mean methanol-diesel, E10 and E20 mean ethanol-diesel, mixed fuel is closed to pure diesel relatively. This can be
N10 and N20 mean n-butanol/diesel, D100 means pure diesel. attributed to the influence of evaporation latent heat, reduce the
temperature in cylinder and extend the ignition delay period, at
the same time, the low cetane number of alcohol fuel extends
TABLE II. THE MAIN MEASURE EQUIPMENT AND INSTRUMENTS the ignition delay period further, the amount of combustible
Exhaust Fuel Combustion mixed gas increases, leading to the increasing of heat release
Device Dynamometer analysis consumption analyzer rate peak value. Moreover, the combustion heat release rate
name system meter peak value of methanol-diesel is the highest.
FC2000 of Horiba FC2210Z of
Model Hunan MEXA- Hunan Kistler When in the same operating condition, the peak value of
instrument 7100D instrument heat release rate increases with the increase of the mixing ratio
when fueled with diesel - alcohol mixed fuel. This result can be
Experiments for pure diesel and alcohol-diesel fuel blends explained because the combined action of low cetane number,
were conducted in a 4 cylinder, 4 stroke, water-cooled, high evaporation latent heat value and oxygen.
turbocharged, common-rail, direct injection diesel engine.
Main technical specifications of the engine are given in Table
I. The main measure equipment and instruments are shown in

6
 The brake specific fuel consumption of mixed fuel
increases with the increase of the mixing ratio of alcohol fuel in
the same operating condition. When the speed and mixing ratio
are constant, brake specific fuel consumption of ethanol-diesel
fuel is higher, which of methanol-diesel is lower.

(a) 2000 rmin-1/270 Nm

(a) 2000r/min

(b) 3000 rmin-1/270 Nm

Fig. 1. The comparison of the heat release rate for various alcohol-diesel

B. Brake specific fuel consumption

The characteristic curves of brake specific fuel
consumption changing with the load at 2000rpm and 3000rpm, (b) 3000r/min
as shown in Fig.2. When at the same speed, the brake specific
fuel consumption decreases with the increase of load and gets Fig. 2. Brake specific fuel consumption as a function of load for various
to be the same. Firstly, because of the low temperature in alcohol-diesel at 2000rpm and 3000rpm
cylinder when operates in medium or small load, and the high
evaporation latent heat value reduces the temperature further,
C. CO Emission.
so the combustion efficiency reduces. Secondly, calorific value
of alcohol fuel is lower than diesel, which rises the fuel Fig.3. shows the trends of CO emissions as a function of
consumption of diesel-alcohol fuel [7]. When in high load, the load for the various alcohol-diesel fuel blends at 2000rpm and
temperature in cylinder rises, the influence of evaporation 3000rpm. CO emission increases significantly with the increase
latent heat is subdued, at the same time, high oxygen content of of the mixing ratio of alcohol in medium or small load, but
alcohol fuel promotes the combustion, thus reduce the brake decreases with the increase of the load and gets to be the same.
specific fuel consumption of the mixed fuel. When operating condition and mixing ratio are constant, CO

7
emission of ethanol-diesel is the highest, CO emission of n- D. HC Emission.
butanol/diesel is the lowest. As with CO emissions, unburned HC emissions are caused
This is, perhaps, because in medium or small load, the because of incomplete combustion. As combustion improves,
lower cylinder temperature and evaporation latent heat of less unburned HC emissions are expected. Fig.4 shows the HC
alcohol fuel, inhibiting the oxidation of CO, together with emission curves of revolving speed at 2000r/min and
evaporation latent heat reduces the cylinder temperature, 3000r/min. When engine operates in medium or small load,
leading to the quenching layer in cylinder, CO emission HC emission increases significantly with the increase of the
increases. Besides, the incomplete combustion of alcohol fuel mixing ratio of alcohol, but decreases with the increase of the
can also increase CO emission. With the load increases, the load and gets to be the same. When operating condition and
cylinder temperature also increases, the effect of evaporation ratio are constant, HC emission of ethanol-diesel is the
latent heat of alcohol fuel is mitigated, and the oxygen of highest, which of n-butanol/diesel is the lowest. Because of in
alcohol fuel promotes combustion process, leads to reducing medium or small load, the effect of evaporation latent heat on
the CO emission. cylinder temperature is obvious relatively, this makes more
quenching layer, leading to more HC. As the load ascends,
better combustion occurs and the cylinder temperature also
ascends, the effect of evaporation latent heat of alcohol fuel is
mitigated, besides, the oxygen of alcohol fuel reduce the thick
area of mixture in cylinder, leads to reducing HC emission.

(a) 2000r/min

(a) 2000r/min

(b) 3000r/min

Fig. 3. CO emission as a function of load for various alcohol-diesel at (b) 3000r/min

2000rpm and 3000rpm
Fig. 4. HC emission as a function of load for various alcohol-diesel at
2000rpm and 3000rpm

8
E. NOx Emission. IV. CONCLUSIONS
NOX emission increased as the load increased for all the In this work, alcohols (methanol, ethanol and n-butanol) as
cases at 2000rpm and 3000rpm, as expected, shown in Fig.5. additives to diesel were investigated in terms of performance
In medium or small load, NOX emission descends with ascend and emission characteristics of a DI diesel engine running on
of the mixing ratio of alcohol. When working condition and several fuel blends. Under the same operating condition and
ratio are constant, NOX emission of n-butanol/diesel mixed alcohol fuel, the heat release rate peak value increases with the
fuel is the highest and gets to be the same with pure diesel, the increase of the mixing ratio. At the constant mixing ratio, the
NOX emission difference between methanol-diesel and combustion heat release rate peak value of methanol-diesel is
ethanol-diesel is small, but the NOx emission of all kinds of the highest. The brake specific fuel consumption decreases
blended fuels are lower than n-butanol/diesel. Because of the with the increase of load but increases with the increase of the
cetane number and evaporation latent heat, extending the mixing ratio of alcohol fuel. At the same speed and mixing
ratio, brake specific fuel consumption of ethanol-diesel fuel is
ignition delay period, premixed combustion increases, the
higher, which of methanol-diesel is lower; When the operating
combustion temperature rises, leading to increase the NOX
condition and fuel ratio are constant, CO and HC emission of
emission. On the other hand, the viscosity, density and surface ethanol-diesel is the highest, n-butanol/diesel is the lowest.
tension of alcohol fuel are small relatively, this accelerates NOX emission of n-butanol/diesel is the highest, NOX emission
fuel combustion and diffusion speed and shorten the difference between methanol-diesel and ethanol-diesel is small,
combustion duration, which shorten the high temperature but both of them are lower than n-butanol/diesel.
duration, leads to decreasing NOX emission.
ACKNOWLEDGMENT
The work was supported by Henan province science and
technology projects (122102210055).
REFERENCES
[1] Satish Kumar, Jae Hyun Cho. Advances in dieselalcohol blends and
their effects on the performance and emissions of diesel engines.
Renewable and Sustainable Energy Reviews 22(2013)4672
[2] Agarwal AK. Biofuels applications as fuels for internal combustion
engines. Progr Energy Combust Sci 2007; 33:233-71.
[3] Sayin C. Engine performance and exhaust gas emissions of methanol
and ethanol-diesel blends. Fuel 2010; 89:3410-5.
[4] Chunde Yao, Qi Xia and Xuping Chen: submitted to Journal of
Combustion Science and Technology (2011).
[5] Deming Jiang and Zuohua Huang: Alternative Fuel Combustion in
Internal Combustion Engine (Xi'an Jiaotong University Press,Xian
2007).
[6] Xingcai Lv, Zhen Huang and Wugao Zhang: submitted to Journal of
Transactions of CSICE (2003).
[7] Wagner TO, Gray DS, Zarah BY, Kozinski AA. Practicality of alcohols
(a) 2000r/min as motor fuel. SAE paper no. 790429; 1979.
[8] Bin Xu, Yongfang Pan and Jian Wu: submitted to Journal of
Agricultural Mechanization Research (2012).
[9] Adelman H. Alcohols in diesel engine. SAE paper no. 790956; 1979.
[10] Nagafi G., Yusaf T.F. Experimental investigation of using methanol
diesel blended fuels in diesel engine. In: Proceedings of the fourth
international conference on thermal engineering: theory and
applications. Abu Dhabi, UAE; 2009.p.15.
[11] Huang ZH, Lu HB. Combustion behaviors of a compression ignition
engine fuelled with diesel/methanol blends under various fuel delivery
advance angles. Bioresource Technology 2004; 95: 33141.
[12] Koganti RB, Maheshwari M, et al. Performance evaluation of ethanol
diesel blend in tractors. SAE paper no.2004-28-0085; 2004.
[13] Rakopoulos DC, Rakopoulos CD. Effect of ethanoldiesel fuel blends
on the engine performance and emissions of heavy duty DI diesel
engine. Energy Conversion and Management 2008; 49 (11): 315562.
[14] Rakopoulos DC, Rakopoulos CD, et al. Effects of butanoldiesel fuel
blends on the performance and emissions of a high speed DI diesel
engine. Energy Conversion and Management 2010; 51(10):198997.
(b) 3000r/min
[15] Dogan O. The influence of n-butanol/diesel fuel blends utilization on a
Fig. 5. NOx emission as a function of load for various alcohol-diesel at small diesel engine performance and emissions. Fuel2011; 90(7): 2467-
2000rpm and 3000rpm 72

9
Proceedings of the 2015 International Conference on
Advanced Mechatronic Systems, Beijing, China, August, 22-24, 2015

Design and Simulation of Passive Filter Based on

Fundamental Magnetic Flux Compensation
Qun Chen*,1,2 and Zong-Xiao Yang1
1
Information Engineering School, Henan University of Science and Technology, Luoyang, China
2
College of Physics & Electronic Information, Luoyang Normal University, Luoyang, China
*Corresponding author Email: [email protected]

AbstractThe harmonic problem becomes more and more transformer is chosen and its primary winding is connected in
serious in power grid, passive filter is one of the effective way to series between the power source and harmonic-producing
filter harmonic. This paper presented the design and simulation loads. The fundamental current instead of the harmonic
of passive filter based on the theory of fundamental magnetic flux currents of the transformers primary winding is detected and
compensation (FMFC). The primary winding of transformer is in followed by applying a voltage-source PWM converter so as
series between the passive filter and the power supply, in the
to produce a compensating fundamental current. The
primary winding of series transformer, the fundamental current
component of system was detected and tracked to produce a compensating fundamental current is injected to the secondary
fundamental compensation current by inverter, which was winding of the series transformer. When the injected
injected into the secondary winding, when the injected fundamental current and power source fundamental current
fundamental current and the fundamental current component of component satisfy the FMFC condition, the main magnetic
the system satisfied the FMFC, the system fundamental current flux of the series transformer to fundamental can be
could pass transformer successfully, but most of the harmonic compensated to zero. Then the series transformer exhibits very
current was forced to flow into passive filter, thus achieved the low primary leakage impedance at the fundamental frequency
purpose of harmonic isolation, and made most of the harmonic and naturally magnetizing impedance to harmonics so that the
current flow into passive filter. Compared with the using of
harmonic currents are forced to flow into the passive filter
passive filter singly, it improved the efficiency of passive filter.
The validity of the new method of filtering characteristics was branches. As a result, the series transformer improves the
verified by simulation through Matlab. The simulation results source impedance to harmonics rather than fundamental so as
show that this method can filter out harmonic more efficiency in to really act as harmonic isolator, [5, 6] and makes most of the
power systems when the waveform distortion is serious. harmonic current of system flow into passive filter, thus
improving the efficiency of passive filter further. Using
Keywordsfundamental magnetic flux compensation; passive Matlab to simulate passive filter singly and based on the
filter; series transformer; harmonic; inverter FMFC passive filter respectively, the simulation results show
that this design has a better filtering characteristics when
I. INTRODUCTION harmonic is serious.
The wide application of power electronic devices makes
II. SYSTEM CONFIGURATION AND FUNDAMENTAL
the grid harmonic problem become more and more serious.
PRINCIPLE
Although active filter overcomes the inherent shortcomings of
passive filter, has realized the dynamic compensation of The system configuration is shown in Fig.1. Ls, Rs and Us
harmonic and reactive power in the power grid, which is an represent the source impedance and the source voltage,
effective method for comprehensive treatment of electrical respectively. The passive filters consisting of third-, fifth-and
energy [1]. However, installation of active filter need to seventh-tuned LC filters, which play the main role of filtering,
increase the equipment investment and operation cost, and the and a high-pass filter branch which is added to filer high times
harmonics, are shunted with a harmonic-producing load. The
higher harmonic amplitude need to suppress, the higher
primary winding of a transformer is inserted in series between
investment and running costs bring. Passive filter can only the power utility and the harmonic-producing load, it will
suppress a specific one or some frequency band harmonic, greatly improve the impedance of the source side to harmonic
although can't realize dynamic compensation, its simple current when the transformer satisfy the principle of the
structure, mature run maintenance experience, low cost of fundamental magnetic flux compensation, while the secondary
one-time investment and operation maintenance, large winding is connected with a voltage-source PWM converter,
capacity and easy to implement, make most of practical which is used to produce fundamental current.
engineering application still use it[2-4]. But it can't meet the
national standard requirements only using passive filter when III. LOAD MODEL AND DESIGN OF PASSIVE FILTER
the waveform distortion is very serious. In this paper, it
presents one method of passive filter combined fundamental The high-power of power electronic equipment is given
magnetic flux compensation (FMFC) of transformer to priority to three-phase circuit at present, but there is no lack of
improve the filtering effect of passive filter. A series single-phase rectifier equipment while is rich in harmonic

978-1-4673-7995-3/15/$31.00 2015
c IEEE 10
content. For the sake of simplicity, in this paper it designs a Table. Harmonic content percentage of system current%
model of single phase bridge rectifier with resistance load as
fundamental
an example for system modeling analysis. wave 3 5 7 9 11 13
ES 100 28.99 16.47 8.17 7.61 6.49 4.59
LS RS
Harmonic
source Known from the analysis of the above, the 3th, 5th, 7th
T1 K=1:1 C3 C5 C7 CH ih harmonic content is large, but the 9th or more times harmonic
content is low. According to the characteristics of tuned filter
L7
APF fundamental L3 L5
LH
and high-pass filter, harmonic which is high content can be
wave flux RH filtered by a single tuned filter, while higher harmonic which
compensation R3 R5 R7
is low content can be filtered by a group of high-pass filter.
Because the capacitor of passive filter has the ability of
Fig.1. The principle diagram of system reactive power compensation, it can compensate the reactive
power of the system through reasonable design, it also can
The single-phase model designed for this system is shown improve the power factor of the system. In this paper, passive
in Fig.2. The sources of system is 6000/ 3 V (Us), the filter parameter design adopts a more economic method of
equivalent resistance and equivalent reactance of system is reactive power capacity compensation [8, 9]. According to this
0.1 (Rs), 0.05mH (Ls) respectively. The source of harmonic kind of design method to calculate the 3th, 5th, 7th for each
is single-phase bridge rectifier with the inductive load when single tuned filter parameters is as follows: C3=43.68F,
the control angle is 60e L3=0.02579H, R3=0.30378; C5=47.75F, L5=0.008496H,
R5=0.2223; C7=8.8F, L7=0.02352H, R7=0.86166. Each
The current waveform of system and its FFT spectrum
parameter value of second-order high-pass filter is: CH=160F,
analysis diagram are shown in Fig.3 through simulation.
LH=0.3913mH, RH=2.2116.
VT1 VT3
us
The system current waveform and FFT spectrum analysis
Ls Rs are shown in Fig.4 when passive filter invest into the system
ZL
only. The system power factor become 0.98 (cos=0.98) after
reactive power compensation by the capacitor of passive filter,
VT2 VT4
which has met the demand of power sector for this type of
load. The distortion of system current rate falls to 23.1% from
36.7%, the distortion rate is relatively large compared with the
Fig.2. The single phase model of system
national standard requirement for harmonic distortion rat
although it has decreased.

(a) System current waveform

(a) System current waveform

(b) FFT spectrum analysis

Fig.3. The system current wave and ite spectrum analysis (b) FFT spectrum analysis of system current waveform
System current distortion rate is 36.7% (THDi) known Fig.4. System current and its FFT analysis when PF is added
from figure 3.The waveform distortion is serious and its
content of every harmonic is shown in Table . The amplitude IV. THE PRINCIPLE OF FMFC AND REALIZATION
of fundamental wave is 404A, the amplitude of 3th, 5th and 7th Known from the analysis of above, the system has meet
harmonic current is 117A, 66.63A and 33.05A respectively the requirements of power factor through reactive power
which are the high content of harmonic current. The power compensation of passive filter, but the harmonic distortion rate
factor of system is 0.867 (cos=0.867) which is below the is relatively large, it cannot meet the requirements of national
power sector demand for this type of load (generally is 0.9 standard requirement for harmonic distortion rate. In this
0.95) [7]. design harmonic magnetic flux compensation theory is used to

11
isolate harmonic from system, thus improve the efficiency of

passive filter.
 I1  I 2 / k 0 (7)

Then the whole circuit is equivalent to the circumstance of

A. The Compensation Principle[11] m=1 in [10], fundamental wave magnetic potential in
According to the principle of magnetic flux controllable transformer core is
transformer [10], the transformer primary winding AX is in
series between the power system and harmonic source,  W1 I 1  W2 I 2 0 (8)
transformer secondary winding ax connects to the output of
active inverter. The series transformer and its T-shape Here (7), (8) is the condition of the fundamental magnetic
equivalent circuit show in Fig.5. The current of primary flux compensation, then (5) and (6) will simplify into
winding is the sum of fundamental wave current and nth
harmonic current, namely i1=i1(1)+i1(h). The injected U1 r1  jZ  L11  kM  I 1 (9)
fundamental current i2 (i2=i2(1)= ki1(1)) into the secondary

winding with the fundamental current component of the power U 2 r2  jZ  L22  M / k  I 2 = I 2  r2  jx2  (10)
utility are the same on frequency and inversely in phase. For
the fundamental and harmonics, there are two cases in terms From terminals AX and ax, the equivalent impedance of
of the law of superposition. the transformer is derived
1) For the fundamental:
U1
1
r1  jZ  L11  kM  Z1 
1
For the fundamental, from the transformer circuit, we can Z AX
r1  jx1 (11)
obtain the voltage equations as follows. I1

U1  r1  jZ L11  I 1  jZM I 2 (1)
Z ax 1

U2
r2  jx2 Z 2 
1
(12)

I2
U 2  r2  jZ L22  I 2  jZM I 1 (2)
Then the equivalent impedance of the secondary winding
Here, L11 and L22 are self-inductance of primary and secondary referred to primary winding is
winding respectively; M is mutual inductance between the ' '
primary and secondary winding; Z is fundamental angular
frequency.
 Z ax'  
1
U 2 / I 2  r  jx 
2
' '
2 Z 2'   
1 (13)

Here, Z1(1) is the fundamental leakage impedance of the

primary, the prime denotes referred quantities of secondary
winding to primary winding. The primary and secondary
impedance (11) and (13) suggest the equivalent circuit of the
series transformer to the fundamental shown in Fig.6.
(1)
I1 r1 x1 x2' r2' '
I2
A a
Fig.5. Series transformer and T-shape equivalent circuit

Equations (1) and (2) can be changed into (1)

U1

U '(1)
2

U1 r1  jZ  L11  kM  I 1  jZkM ( I 1  I 2 / k ) (3)

X x

U 2 r2  jZ  L22  M / k  I 2  jZ M ( I 1  I 2 / k ) (4) Fig.6. The fundamental equivalent circuit

2) For nth-order harmonic:

Further, equations (3) and (4) can be changed into

Since only a fundamental current is injected to the
U1  r1  jZ Ls1  I 1  jZkM (I 1  I 2 / k ) (5) secondary winding of the transformer, i2doesnt include any
order harmonic besides the fundamental current, that is to say,
i2(n)=0, thus the whole circuit is equivalent to the circumstance
U 2  r2  jZ Ls 2  I 2  jZM (I 1  I 2 / k ) (6)
of m = 0 in [10], the harmonic voltage of the transformer
Here, Ls1, Ls2 are the leakage inductance of primary primary winding is:
winding and secondary winding, respectively. Let ZLs1=x1, (n) (n) ( n)
ZLs2=x2, x1 and x2 represent the fundamental leakage inductive U1 r1  jnZ  L11  kM  I 1  jnZ kM I 1
reactance of the primary and secondary winding, respectively.  (n) (n)
 (14)
jZkM=Z1(1), Z1(1) is the fundamental magnetizing impedance of =  r1  jnx1  I 1  jnxm I 1
the transformer.
The nth harmonic impedance appears to the follow
If the injected fundamental current i2 satisfies
equation from the terminals AX.

12
 (n) when the load and system parameter are change.
U1
 Z (n)
AX Z ( n)
m (n)
 r1  jnx1   jnxm | nZ 
(1)
m
(15) Distortion Fundamental
currents costt LPF current i1(1)
I1 costt
sint
nt
sint
nt LPF
Here, Zm(1) is the excitation impedance of fundamental
wave. According to (14) and (15), the equivalent circuit of the
transformer to the nth-order harmonic can be represented by u costt
s sint
nt
PLL
Fig.7.
(n)
I1 r1 n x1 Fig.8. Fundamental current detection diagram based on the principle of phase
A discriminator

(n) n xm
U1

X
Fig.7. The harmonic equivalent circuit.

Known from the analysis of the above, as long as the

inverter tracks to the fundamental current of primary winding
i1(1) , injects the compensation fundamental current i2(1) into
secondary winding of transformer, and meets the condition of
fundamental magnetic flux compensation of the equation (7)
and (8), then the main magnetic flux fundamental wave is Fig.9. Distortion current waveform of system (up) and fundamental current
waveform detected (down)
close to zero. It impedance appears to be a low-impedance of
Z1(1) for fundamental wave from the view of terminal AX, but
for the nth harmonic, it appears to be a high-impedance of
nZm(1).
Apply this principle to the design, transformer is in series
between the passive filter and the power supply, the
fundamental wave presents to be a leakage impedance which
approximate to zero for transformer primary winding, but for (a)
the harmonic it appears to be a high excitation impedance,
thus improve the harmonic impedance to the whole system, so
it attains the aim of harmonic current isolation, forces more
and more harmonic current flow into passive filter, and
improve the efficiency of passive filter further.

B. The Realization of FMFC

In order to realize the fundamental magnetic flux (b)
compensation of series transformer in this design, it should
detect the fundamental wave current component from
distortion current firstly, and through a certain way of current
control to track the fundamental wave current to generate
fundamental wave compensate current which needed, and
realize the real-time compensation finally.
In this paper, the fundamental wave current detection
adopts the instantaneous fundamental wave current detection
based on the principle of phase discriminator which has the
simple principle real-time performance as well as high
accuracy [12-14]. The configuration of this principle is shown in
Fig.8. Using this method to detect the system fundamental (c)
wave current from distortion current is shown in Fig.9.
Fig.10. The waveform of system fundamental current and hysteresis tracking
The inverter circuit is used to generate fundamental wave current
compensation current. The single-phase voltage source The fundamental wave compensation current waveform
inverter circuit is adopted in this design, the current tracking generated by the inverter by tracking the fundamental wave
mode adopts hysteresis comparison control mode. Reference current of the system distortion current is shown in Fig.10.
[12] offers the principle and the switch frequency calculation. Fig.10 (a) is the fundamental current tracking information by
This control mode has good stability and good robustness intercepting one segment of the simulation waveform. Fig.10

13
(b) is the actual tracking current waveform by phase reverse pass transformer successfully, but most of the harmonic
for observing conveniently compared to the actual current is forced to flow into passive filter, so it can improve
fundamental current, through filtering the high frequency the efficiency of passive filter. In this design, because only the
switch burr by output filter, two waveforms can achieve the fundamental current component needs to be detected and
same basically. tracked, and the fundamental is a single frequency, the
detecting and tracking with rapid response is more easily
realized than harmonic detecting and tracking. Using
MATLAB to do simulation to the model, the simulation
results show that this design can filter out harmonic more
efficiency in power systems when the waveform distortion is
serious compared with using passive filter singly.

th REFRNCES
Fig.11. The 3 harmonic current wave
[1] R. C. Dugan, M. F. McGranaghan, H. W. Beaty, Electrical Power
Systems Quality, 1996, McGraw-Hill.
V. SIMULATION RESULTS
[2] R. D. Henderson and P. J. Rose, Harmonics: the effects on power quality
Fig.11 is the harmonic current through the three times and transformers, IEEE Transactions on Industrial Application, vol. 30,
single-tuning filter without adding FMFC transformer. Fig.12 pp. 528-532, May/Jun. 1994.
is the harmonic current through the three times single- [3] R. Redl, P. Tenti, and J, Daan van Wyk, Power electronics polluting
tuning filter when adding FMFC transformer. Fig.13 is the effects, IEEE Specter, vol. 34, pp. 32-39, May. 1997.
system current waveform after the fundamental magnetic flux [4] J. Arrillaga, and R. W. Neville, Power System Harmonics, 2nd Edition,
2003, John Wiley & Sons.
compensation in the system power supply side, through the
[5] D. Rivas, L. Moran, J. W. Dixon, J. R. Espinoza, Improving passive
simulation result by FFT spectrum analysis, it can obtain the filter compensa-tion performance with active techniques, IEEE Trans.
result that the system current distortion rate falls to 1.27% Ind. Electron. vol. 50, pp. 161-170, Apr. 2003.
from 23.1%, which approximates to sine wave and is enough [6] A. Boussaid, Y. Maouche, A.L. Nemmour, A. Khezzar, A new optimal
to satisfy the requirement of system to harmonic, here, the approac for improvement of active power filter using FPSO for
data of 23.1% is only the passive filter added. enhancing power quality, Electrical Power and Energy Systems,vol. 69,
pp. 188-199, Feb, 2015.
[7] F. Z. Peng, H. Akagi and A. Nabae, A new approach to harmonic
compensation in power system - a combined system of shunt passive and
series active filters, IEEE Transactions on Industrial Application, vol.
26, pp. 983-990, Nov./Dec. 1996.
[8] H. Fujita, and H. Akagi, Apractical approach to harmonic compensation
in power systemsSeries connection of passive and active filters, IEEE
Trans. Ind. Applicat., vol. 27, pp. 1020-1025, Nov./Dec. 1991.
[9] B. Singh, and V. Verma, An improved hybrid filter for compensation of
th
Fig.12. The 3 harmonic current when FMFC is added current andvoltage harmonics for varying rectifier loads, Int. J. Electr.
Power Energy Syst, vol. 29, pp. 312-321, Apr. 2007.
[10] D. Y. Li, Q. F. Chen, and Z. C. Jia, A novel principle of adjustable
reactor based on magnetic flux controlable[J], Proceedings of the CSEE,
vol. 23, pp. 161-120, Feb. 2003.
[11] D. Y. Li, Q. F. Chen, Z. C. Jia and C. Z. Zhang, A novel active power
filter with fundamental magnetic flux compensation, IEEE Transactions
on Power Delivery, vol. 19, pp. 799-805, Apr. 2004.
[12] H. Akagi, A. Nabae, and S. Atoh, Control strategy of active power filters
using voltage-source PWM converters, IEEE Trans. Ind. Applicat, vol.
Fig.13. System current wave when FMFC is added IA-22, pp. 460-465, May/June. 1986.
[13] A. Boussaid, A. L. Nemmour, L. Louze, and A. Khezzar, A novel
VI. CONCLUSION REMARKS strategy for shunt active filter control, Eleetric Power Systems Research,
vol. 123, pp. 154-163, Mar. 2015.
In this paper a design of passive filter combined with [14] RS. Herrera, P. Salmeron, H. Kim, Instantaneous reactive power theory
FMFC theory is proposed, namely a double winding applied to active power filter compensation: different approaches,
transformer is in series between the power supply and passive assessment and experimental results. IEEE Trans Ind Electron. vol. 55,
filter, the secondary winding generates fundamental current by pp. 184-196, Feb.2008.
active inverter with the principle of fundamental magnetic flux
compensation, thus the system fundamental wave current can

14
Proceedings of the 2015 International Conference on
Advanced Mechatronic Systems, Beijing, China, August, 22-24, 2015

Research on Large Supermarket Fresh Food Supplier

Evaluation and Selection based on SWOT-Entropy Weight
Fuzzy Comprehensive Model
Ying Yuan*Yi ZhangLong DiGang Gang WuZong Xiao Yang
College of Vehicle and Transportation Engineering
Henan University of Science and Technology
Luoyang, China
*
[email protected]

Abstract Aiming at the subjectivity and random problems In the actual operation of the supermarket, due to the
existing in traditional evaluation and selection processes of huge number of suppliers, the supply of commodities is also
supermarket fresh goods suppliers, and combined with the more diverse, thus the factors which affect the quality of the
perishable and vulnerable characteristics and the current
goods are complex[2]. Many previous studies on supplier
situation and development trend of supermarket fresh food, the
selection and evaluation defined numerous evaluation criteria
paper constructed the ladder type evaluation framework by
and selection frameworks for supplier selection, and the
using SWOT analysis method, which is used to study the external
environment and capacity of the suppliers of supermarket fresh indicators in the studies mostly adopts QCDS that is
goods. In view of the deficiencies of common comprehensive quality, cost, delivery, service principle[3]. It is not difficult
evaluation methods on weight determination , therefore, this to find that these existing index systems present more or less
study proposed a structured methodology for supplier selection one-sidedness, according to the characteristics of supermarket
and evaluation based on the Entropy Weight Fuzzy fresh goods supplier selection, (For example, only consider
Comprehensive Evaluation. In the end, the method was applied the influence of internal factors , while ignoring the influence
on the course of the fresh goods suppliers optimization in one
of the external environment). Therefore, Aiming at the
supermarket to validate the feasibility and correctness.
deficiencies(such as, single, strong subjective arbitrariness.
Keywords Large supermarket; Fresh goods; Supplier etc ) existing in traditional evaluation and selection processes
selection; SWOT analysis; Entropy Weight; Fuzzy Comprehensive of supermarket fresh goods suppliers[4, 5]. this study thus
Evaluation proposes a structured methodologySWOT for supplier
selection and evaluation by studying the internal factors and
I. INTRODUCTION
external environment of candidate suppliers comprehensively,
With the development of supermarket management and which relates to all links of procurement, transportation,
the swift of peoples consumption concept, supermarket has distribution, Etc. and based on this method the criteria and
become the main channel for customers to purchase fresh indicators of supplier selection are chosen to establish a
goods, which are indispensable commodity in supermarket comprehensive supplier evaluation index system .
chains[1]. And the management of fresh goods, which is the Subsequently, potential suppliers are screened through
main means for supermarket to implement the differentiated Entropy weight fuzzy comprehensive evaluation method.
competition, plays a key role in attracting customers and Finally, the feasibility and applicability of the proposed
stabling the passenger flow in supermarket. Especially in the method are verified by an example, which provides the theory
larger supermarkets is more remarkable. Hence, the bases for supermarket to choose the best supplier.
evaluation and selection of suppliers who provide fresh goods
to supermarket is an important problem, which is the large
supermarkets need to consider.

978-1-4673-7995-3/15/$31.00 2015
c IEEE 15
 II. SWOT ANALYSIS FOR SUPERMARKET FRESH FOOD D. Threat
SUPPLIERS
Fierce competition, potential substitutes, etc, to a certain
A. Strength extent, are threatening the development of suppliers. In the
a. Price: Some suppliers can supply to supermarket at course of the negotiations with supermarket, due to the lack of
wholesale price, which make supermarket have a price Industry Association Protection, suppliers are usually in a
advantage. weak position. In addition, the uncertainty of market and
natural risk, which also makes them face a great challenge.
b. Channel: Large suppliers have more sale channels and
control the sales terminal, which can help producers to III. ESTABLISHMENT OF EVALUATION CRITERIA AND INDEX
make their goods access to the supermarket in the most SYSTEM FOR SUPPLIER SELECTION

short time, and make mutual benefit. It is more important

A. The index system
to seasonal vegetables and fruit.
According to the characteristics of fresh food and current
c. The fixed assets scale: They have their own cold storage, situation of supermarket business[6,7], as well as based on the
transport vehicles, logistics infrastructure, and the above analysis of fresh suppliers by using SWOT method, the
continuous expansion of business, which makes their supermarket can build the evaluation index system framework
equipment become more perfect. according to the suppliers price, quality, service, innovation
d. Brand effect: Many suppliers have the quite high ability, delivery etc[8]. As shown in Figure 1
popularity in the market through years of operation, and 1) Price: It can reflects the information compared with the
have strong connections with the supermarket. market counterparts through the price, quantity discount,
transport cost, which are provided by suppliers.
e. The financing facilities: Because the mature supplier has
2) Delivery: It reflects suppliers delivery speed, including
many advantages, so both fresh base and manufacturers
the supplier delivery time and delivery accuracy etc.
use the conditions to sell their products, and provide
3) Quality: Its the ability index that is to ensure the quality
credit funds to the supplier.
of goods, including the qualified rate and the discount
B. Weakness rate.
At present, most fresh suppliers have small scale, lack of 4) R&D capability: With the application of information
professional knowledge etc, which affect the long-term technology in SCM, as well as the rapid development of
management of chain supermarket, which is not easy to form market demand, the suppliers have the market perspective
a standardized and stable relationship between supply and and the ability to develop and improve products, which
demand. meets the requirement of customers.
5) Service: This is an important index for measuring the
C. Opportunity
supplier service capability and their cooperation ability.
With the development of economy, peoples income has 6) Competence: Facing the fierce market competition and
improved. They ask higher requirements on the quality and various uncertain factors, dose the suppliers can
variety of goods. This promotes the supermarket scale and the effectively make use of advantageous factors, avoid the
scope of business gradually expanded, which makes the threat factors and stand out in the strong competitors?
supermarket business varieties tend to be diversified and 7) Business strategy: To formulate the strategy based on the
provides a good basis of market demand for suppliers. In external environment and their own conditions, and
additionsuppliers also through investment, self improvement, constantly adjust and implement. So as to improve their
internet technology and so on to improve their ability. At the own competitiveness.
same time, increasing guide of government support also 8) Organizational management: Its the first problem need to
provides a good environment for the supply enterprise. solve, which represents the strength of the suppliers.

16
 Evaluation Index System for Supplier
Selection

R&D Service Business 2UJDQL]DWLRQDO

Price Delivery time Quality Competence
Capability Strategy 0DQDJHPHQW

xLead Time xCustomer xCompetitors xIntegration

xGross xQuality xR&D Rate xInventory
xDelivery Complaints xBusiness Risk Strategy
Profit Rate Qualified xProductivity Turnover
Flexibility xAfter Sale xPerspective xDifferentation
xVolume Rate xImprovement Ratio
xOn-Time Service xQuick Strategy
Discount xDiscount Capability xCommunicat
Delivery Satisfaction Response to xMarket
Rate Rate ion
Rate xInfrastructure Market Cooperation

Supplier 1 Supplier 2
Supplier M

Fig. 1 Framework of evaluation indicators for supplier selection.

The index system from Figure.1considers not only the The above index system covers subjective and objective
comprehensive level of suppliers, but also the future
factors. Some indicators can be quantified directly, such as the
development trend and prospect, which can reflect the whole
situation of suppliers. price, delivery etc. Others are only peoples subjective feeling,
In practice, due to the difficulty of each index data acquisition,
B. The Calibration method of index
the calibration of each index uses the method listed in table 1.

TABLE I. THE CALIBRATION METHOD OF EVALUATION INDEX

Numbers Main criteria Sub-criteria Get method Characteristics

1 Price The unit price Query Objective
2 Delivery time On time delivery rate Query Objective
3 Quality Qualification rate Query Objective
4 R&D capability R&D Rate Scoring Subjective
5 Service Customer satisfaction Research Objective
6 Competence Handle drawbacks ability Scoring Subjective
7 Business Strategy Market Cooperation Scoring Subjective
8 Management Organizational Scoring Subjective

The objective indices can be obtained directly from each evaluating objects counted n, then forms an original indices
supplier information parameter or by using statistical method,
value matrix X ( xij ) mun
while the subjective indexes are mainly obtained by using 10
point scale. The higher the score, the better the index.
x11 x1m
(1)
IV. ENTROPY FUZZY COMPREHENSIVE EVALUATION MODEL X=
x xmn
A. Normalization of the original evaluating matrix n1

Suppose there are evaluating indexes counted m, Normalization this matrix to get Equation (2)

17
 y11 y1m m
(2)
as: wi (1  H i ) / (1  H i ) (6)
Y ( yij ) num i 1
y ynm
n1 The entropy method is objective weighting, so the
scientific weights should be combined the expert evaluation
Where, yij is the data of the jth evaluating object on the [9]
(subjective) with the entropy weight (objective) . Suppose
the weight of expert evaluation of ith indicator could be
indicator, and. yij [0,1] Among these indicators, to which
defined as O =(O1 , O2 ,, Om ) , thus the comprehensive
the bigger the better, there are

xij  min( xij ) weight of ith indicator could be defined as

yij (3) m
max( xij )  min( xij ) ti (Oi u wi ) / (Oi u wi ) (7)
i 1
While, the smaller the better, there are
Get the comprehensive weight vector : T (t1 , t2 ,, tm )
max( xij )  xij
yij (4)
max( xij )  min( xij ) D. Fuzzy comprehensive evaluation matrix

Compound operate weight and the fuzzy comprehensive

B. Definition of the entropy
evaluation matrix can get comprehensive evaluation result [10]:
In the m indicators, n evaluating objects evaluation
B TY =b j , ( j 1, 2, , n) (8)
problem, the entropy of ith indicator is defined as
n
V. EXAMPLE
Hi k fij ln fij (i 1, 2,, m) (5)
j 1 In order to promote the development of business and meet
the consumers demand, so as to improve their market
n
In which f ij yij / yij , k 1/ ln n , competitiveness, one supermarket needs to choose the best
j 1 one as its strategic partner from six suppliers. According to
the calculated steps, we can compute the entropy weight of
C. Definition of the weight of entropy
the evaluation indices, as is shown in table 2.
The weight of entropy of ith indicator could be defined

TABLE II. WEIGHT AND MEMBERSHIP SUMMARY OF THE INDEX

Suppliers
Index Get method
A B C D E F

Price 3.7 3.2 3.5 3.1 3.8 2.9 VWDWLVWLFV

Delivery time 92% 90% 88% 95% 94% 90% VWDWLVWLFV
Quality 89% 91% 85% 92% 95% 90% VWDWLVWLFV
R&D capability 8 8.5 9.1 7.8 8.2 7.5 6FRULQJ
Service 92% 94% 96% 90% 93% 93% VWDWLVWLFV
Competence 8.1 9.1 8.2 7.6 8.5 7.9 6FRULQJ
Business Strategy 8 7.3 6.9 9 8.2 6.9 6FRULQJ
Management 8.1 7.2 8.3 6.5 8.2 8.6 6FRULQJ

According to the original data in Table 2, get the evaluation matrix

18
 3.2 3.7 3.5 3.8 3.3 2.9 2) Different dimension indexes were standardized by using
90% 92% 95% 95% 94% 90%
the entropy weight fuzzy comprehensive evaluation

89% 91% 85% 92% 95% 90% method, and combined subjective weight given by

8.0 8.5 9.1 7.8 8.2 7.5 experts, which can effectively avoid the interference of
X =
92% 94% 96% 90% 91% 93% subjective factors, and overcome the limitations existing

8.1 9.1 8.2 7.6 8.5 7.9 in determination of the weight by a single method. This
8.0 7.3 6.9 9.0 8.2 6.9 method fully reflects the advantages of the objective and

8.1 7.2 8.3 6.5 8.2 8.6 subjective evaluation, and make the evaluation results
more true and reasonable.
Get the normalization matrix:
3) The methodology proposed in this article can help
0.6667 0.1111 0.3333 0.0000 0.5556 1.0000
0.2857 0.5714 1.0000 1.0000 0.8571 0.2857
supermarket select the most appropriate supplier and
evaluate supplier performance effectively and simply by
0.4000 0.6000 0.0000 0.7000 1.0000 0.5000
an example analysis, which has strong applicability and
Y 0.3125 0.6250 1.0000 0.1875 0.4375 0.0000
0.3333 extension.
0.6667 1.0000 0.0000 0.1667 0.5000

0.3333 1.0000 0.4000 0.0000 0.6000 0.2000 ACKNOWLEDGMENT
0.6000 0.3200 0.1600 1.0000 0.6800 0.1600

0.7619 0.3333 0.8571 0.0000 0.8095 1.0000 This project was supported by Henan Province Natural
Science and Research Program Projects (2011A580002).
Calculated the entropy of each index:
H (0.7992 0.8317 0.8705 0.8158 0.8161 0.8182 0.6166 0.8690) REFERENCES
Calculated the entropy weight of each index: [1] N.Chamhuri and P.J.B. Batt. Segmentation of Malaysian shoppers by
stroe choice behavior in their purchase of fresh meat and fresh
W (0.1349 0.1131 0.0870 0.1237 0.1236 0.1221 0.2576 0.0880) produce,[J]. Journal of Retailing and Consumer Services,2013,
20(6):516-528.
The weight of expert evaluation: [2] B. Wang, X.Z. Fu and S.K. Zhang, Research on the Evaluation on
Fresh Agricultural Product Supplier Based on Entropy Weight TOPSIS
O (0.1 0.15 0.2 0.15 0.1 0.1 0.1 0.1) Model,[J]. Journal of Anhui Agti. Sci. 2008,26(13):5626-5628.
[3] J.C. Yuan. Structured methodology for supplier selection and
Calculated the comprehensive weight of each index: evaluation in a supply chain,[J]. Journal of Ratailing and Consumer
T (0.1119 0.1408 0.1443 0.1540 0.1025 0.1013 0.2137 0.0730 ) Service,2013,20(6):516-528.
[4] A. Aksoy and N. Oztiirk. Supplier selection and performance
Finally calculated the comprehensive evaluation vector: evaluation in just-in-time production environments,[J]. Export System
B TY (0.4505 0.5600 0.4310 0.5714 0.6146 0.4030) with Applications, 2011,30(5):51-59.
[5] J. Lee, H. Cho and Y. S. Kim. Assessing business impacts of agility
That is, the final ordering of the results is criterion and order allocation strategy in multi-criteria supplier
selection,[J]. Expert System with Application, 2015,42:1136-1148.
F oC o A o B o D o E
[6] A. Rong, R. Akkerman and M. Grunow. An optimization approach
So, we can get that the comprehensive levels of E is the for managing fresh food quzlity throughout the supply chain,[J].
International Journal of Production Economics, 2011,131(1):421-429.
highest, which is the preferred partner for the supermarket. [7] J.C. Yuh. Structured methodology for supplier selection and
evaluation in a supply chain,[J]. Information Science, 2011,
VI. CONCLUDING REMARKS 181:1651-1670.
[8] B.E. Song, W.Y. Zhu and X.M. Li. Research on supermarket fresh
1) Based on the characteristics of fresh suppliers selection, food vendor selection model based on entropy combining weight and
topsis,[J]. Food & Machinery,2013,29(4):223-228.
this paper established a set of comprehensive evaluation [9] J.H. Li, X.H. Li and J. He. Application of Entropy Weight Fuzzy
index system through the SWOT analysis method. This Comprehensive Evaluation Mothod to Taxicab Style Optimization,[J].
Journal of Transportation and Communication, 2008,26(4):193-196.
system, which involved all aspects of the internal and [10] Z.H. Tian and S.C. Gao. Entropy-based of Fuzzy Comprehensive
external factors, reflected the comprehensive level of the Evaluation Method for Surface Water Quality Evaluation,[J]. Journal
of Anhui Normal University, 2012, 35(1):63-66.
candidate suppliers, and provided a theoretical basis for
supermarket to select the best supplier.

19
Proceedings of the 2015 International Conference on
Advanced Mechatronic Systems, Beijing, China, August, 22-24, 2015

A Novel Outer Rotor Axial Primary Magnetic

Circuit Permanent Magnet Generator
Xi-Mei LI Zong-Xiao YANG* and Dong-Liang SONG
Institute of Systems Science and Engineering, Henan Engineering Laboratory of Wind Power Systems
Henan University of Science and Technology, No.48 Xiyuan Road, Luoyang 471003, China
*Corresponding author Email: [email protected]

AbstractWind energy has attracted persistent attention as a Permanent magnet generator has a simple structure,
renewable energy, so it is very significant to research the reliable operation, small size, light weight, small loss, high
permanent magnet generator which is an essential part of wind efficiency and significant advantages. Its shapes, sizes, spatial
energy conversion system. This paper proposes a novel outer placement and direction of the permanent magnets in
rotor axial primary magnetic circuit permanent magnet combination has a variety of flexible, so that the permanent
generator. For the sake of small size and light weight with extra magnet generator of electromagnetic structure design, in theory,
low speed for direct coupling, the outer rotor structure is used. has a variety of changes[5-6]. The current conventional wind
The stator windings are adopted double fractional-slot turbine permanent magnet between the stator and the rotor has
concentrated to reduce the cogging torque. The most highlight is
utilized the structure in rotating tangential alternating with
a novel axial primary magnetic circuit. Through the air gap
magnetic flux density test experiments, there exists the linear
permanent magnets, however, this kind of generator is with the
relationship of the negative slope between the magnetic flux overall start reluctance torque and lower wind energy
density and the air gap. This novel type of magnetic circuit conversion rate so that energy cannot generate electricity under
structure makes greatly improved the sensitivity gap, so that the the breeze [7]. To solve the problems of the conventional
key technology such as structure design process route structure, we propose a novel outer rotor axial primary
determined and manufacturing become easier to achieve, magnetic permanent magnet generator combined with coaxial
significantly reduce generator costs, ease of construction direct drive vertical axis wind turbine. Through magnetic flux
practical application. Meanwhile, the novel outer rotor axial density and the air gap test experiments, it becomes easier to
primary magnetic circuit permanent magnet generator can achieve for the permanent magnet generator structure of design,
overcome the problems of the lager overall start reluctance process route determined, manufacturing and other key
generator torque and the lower wind energy conversion rate technology areas.
under the condition of small wind.
KeywordsWind energy; Novel axial primary magnetic circuit; II. WORKING PRINCIPLE
Outer rotor; permanent magnet generator.
The direct drive type vertical axis wind turbine consists of
four major: the upper part of three halberd vertical axis wind
turbine the permanent magnet wind turbine in the middle
I. INTRODUCTION the lower part of the support base and multi-stage storage type
Wind energy is a clean, safe, renewable, green energy, of power supply control system components. The lower part of
inexhaustible [1]. In recent years, there is a growing awareness the wind wheel through flange and the upper lid of a generator
of environmental problems caused by fossil energy, is connected rigidly with bolts, the outer rotor shell is fixed to
particularly global warming. Wind power gradually increases between the upper cover and lower cover. The generator is
the proportion of the world's total energy and has become the mainly composed of three parts including rotor, stator and the
fastest growing, the most promising way of energy air gap ,the pole of permanent magnet generator is constituted
development in the world. Large-scale development and by the permanent magnets arranged the outer rotor. While the
utilization of wind energy will reduce the consumption of fossil outer rotor generator is with wind wheel rotation, permanent
fuels and reduce greenhouse gas emissions and protect the magnets fixed to the outer rotor and stator coils together with
environment [2]. As an important part of wind energy vertical axis wind wheel cut magnetic field lines along with
conversion system, many types of generator concepts have wind wheel rotation movement. From the principle of
been used and proposed [3]. Not every place has a strong wind. electromagnetic induction, the conductor cutting magnetic
How to obtain energy from low wind speed is a challenge field line will generate the induced electromotive force in the
today. Low speed characteristics of direct drive permanent circuit, the coils windings via terminals interval leads to a
magnet synchronous generator can match low-power off-grid power load, then coils leads to two other locations of
wind energy conversion system, it have a good applied to additional power load, thereby constituting a three-phase AC
urban roof, remote mountainous areas and other low-speed output[8].
applications[4].

978-1-4673-7995-3/15/$31.00 2015
c IEEE 20
 III. STRUCTURE DESIGN multipole structure can be easily accommodated. Moreover,
the total length of the magnetic path is reduced. This not only
As its technological development, the structure of
permanent magnet generator has a various of forms, their effectively uses the PMs, but also offers a noticeable height
design criteria is the use of rare earth permanent magnet which reduction of the rotor yoke, resulting in a significantly reduced
has high coercively, increasing magnetic flux, reducing the total volume and weight.
armature reaction, high-speed operation to improve the 3) While the generator is running, the centrifugal force of
electromagnetic efficiency[9]. How to choose the right the magnets applies a pressure to the bonding media. Thus, the
structures of generator in many types needs to be screened reliability of the glued joints becomes higher.
against for different occasions. In this paper, in order to meet 4) The coil pitch equals the slot pitch so that the end
direct drive operating conditions under the low wind speeds, winding is shorter, and correspondingly, copper loss is lower
we design a novel axial primary magnetic structure of the than those in a long pole pitch machine.
outer rotor permanent magnet generators. Though experts in 5) As exposed to the natural wind flow, the outer rotor
the generator field have researched windings in many years, it has a much reduced temperature rise than does the stator.
has found that if the stator windings make use of double
concentrated fractional slot structure, the permanent magnet
generator will be more beneficial to play a good performance
at the running time [10]. Details are given below.
A. Axial Primary Magnetic Circuit
Currently, the traditional permanent magnet wind turbines ,
have problems of the lager overall start reluctance generator
torque and the lower wind energy conversion rate under the
condition of small wind .The paper provides an axial primary
magnetic circuit structure to overcome the shortcomings of
traditional magnetic structure, to achieve energy generation
with the breeze. The axial primary magnetic circuit has
formed a closed primary magnetic circuit in the main axis
direction of the generator by outer rotor permanent magnet
unit and the stator magnetic induction unit, while the magnetic
flux leakage direction is formed the air gap tangential Fig.1. The main axial magnetic field lines distribution
direction arranged as the way of SN pole interval reverse in
the in the circumference of the shaft section. The equal
interval of permanent magnets arranged in the direction of the
magnetic flux leakage is above five times, under 20 times
more than the main magnetic direction of the air gap. In order
to ensure the strength of the magnetic flux density along the
axial line of the main magnetic path direction, the relationship
between the magnetic flux density and air shows a negative
slope of the linear, not only improving the sensitivity gap
generator, but becoming easier to achieve the structure design
of permanent magnet generator, significantly to reduce in the
cost of the generator, to facilitate engineering application. The
Fig.1 shows the main axial magnetic field lines distribution by
software Magnet.
B. The Design of Outer Rotor Permanent Magnet
Another salient feature of the designed PMG is the outer-
Fig.2. The outer rotor structure model
rotor structure, the Fig.2 is outer rotor structure
model.Features of the simple outer rotor structure make the
C. The Stator Laminations and Coil Windings Structure
PM generator compact, light, rugged, and easy to be installed
at the top of a high tower. As can be noticed, the wound stator Armature winding of permanent magnet generator would
is stationary, located in the center of the machine while the be embedded in the stator slots; the common permanent
magnets are mounted evenly along the inner circumference of magnet generator stator slot has divided into four categories:
pear-shaped slot, the trapezoidal slot, half open slots and the
the rotor. Several advantages can be identified easily from the
opening slot. Low-power permanent magnet generator stator
outer-rotor structure, as follows[11].
slot is better suited equal tooth width, we choose pear-shaped
1) The blades of the wind turbine can be conveniently slot, which is equal tooth width slot and with smaller magnetic
bolted to the front face of the drum to realize the direct flux leakage harmonic components [12-13].
coupling between the wind turbine and the PM generator.
2) Because of the enlarged periphery of the out-rotor, the As the generator opened cogging begins to rotate, it will
changes in the stator teeth and slots corresponding gap

21
parameter, causing its gap magnetic field energy changing, so V0 H m / (Hm  H0 ) (1)
torque changes, that is cogging torque. In order to reduce the
cogging torque, AC servo permanent magnet motors adopt From the experimental data shown in Fig. 4, it seems that
fractional-slot concentrated windings, it is an indisputable there is a linear relationship between the main magnetic path
trend. Fractional-slot concentrated winding is a single or RI WKH PDJQHWLF LQGXFWLRQ % DQG WKH DLU JDS , so we choose
double layer stator windings of the span y = l, the stator slot the formula (2) which is the unitary linear regression as
number of per pole per phase q <l[14] .To make winding analysis of mathematical models to study the intrinsic
induced electromotive force amplitude greater ,we hope relationship of magnetic induction H and WKHDLUJDS. m

winding coefficient is larger, generally short coefficient is

larger than 0.866,namely,the short angle is not exceed
60 e, and set that the number of slots is a multiple of 3 to Hm AG  B (2)
comply with the three-phase symmetry. In this paper , the
stator design is 30 poles 27 slots , matching the number of In accordance with the mathematical model the
poles and slots, this form is not only with a bigger the winding formula(2) , we make the test data in Table 1 to process by
coefficient ,but also with smaller starting torque. Based on the using of the least squares method, we can obtain the following
above windings theory and design ideas, the Fig.3 shows the estimation parameters: the slope of A -300, the intercept of
expanded view of the windings. B 1300, and the correlation coefficient of 0.9043. Regression
line is also shown in Fig 4.

main magnetic flux

(mT)
1 mT = 10 GS

Data regression analysis model and results:

1100
+P $%$ % 
&RUUHODWLRQFRHIILFLHQWLV
1060

1020

980
Fig.3. The expanded view of the windings

IV. ELECTROMAGNETIC FIELD TEST 940

A. Test Steps The air gap

(mm)
900
Firstly, select any of the stator as a reference, and then 0.70 0.80 0.90 1.00 1.10 1.20
measure the magnetic flux density of each rotor and the stator
air gap, the main magnetic flux, the direction of leakage Fig.4. The test results of the main magnetic direction of the magnetic flux
magnetic flux in clockwise direction. Finally, record he density and the air gap
measured data. The obtained experimental data is shown in.
Test data from Table1 can estimate the leakage coefficient
of new three-dimensional magnetic structure according to the
equation (1), and the results are shown in the last line Table 1.

z Table 1 Test Date

$LUJDSmm 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20 1.25

The main magnetic circuit of

magnetic induction intensity-Hm 1081 1060 1074 1061 1039 1009 981 968 961 960 908 954
mT

The leakage magnetic circuit of

magnetic induction intensity-H0 154 212 198 176 192 137 136 164 228 163 178 327
mT

leakage coefficient-0 0.88 0.83 0.84 0.86 0.84 0.88 0.88 0.86 0.81 0.85 0.84 0.74

22
B. Analysis of Test Results ACKNOWLEDGMENT
According to the traditional theory, it shows that the The financial supports from the National Natural
magnetic field intensity is inversely proportional to the square Science Foundation of China (NSFC) under the grant
of the air gap. When the test uses Tesla of the Hall probe No.70671035 and No.71071078 are acknowledged gratefully.
minimum thickness of 1.1mm, the result has found that there The authors also gratefully acknowledge the helpful
seems to be a linear relationship existing between the comments and suggestions of the reviewers, which have
Analysis of Test Results. improved the presentation.
According to the traditional theory, it shows that the
magnetic field intensity is inversely proportional to the square
of the air gap. When the test uses Tesla of the Hall probe REFERENCES
minimum thickness of 1.1mm, the result has found that there
seems to be a linear relationship existing between the [1] W. Wu, V.S. Ramsden, T. Crawford and G. Hill,"Low-speed, high-
magnetic flux density and the air gap. After improving the torque, direct-drive permanent magnet generator for wind turbines,"
precision of Hall probes (Hall probe minimum thickness of 35th IAS Annual Meeting and World Conference on Industrial
0.2mm), the result of verification test confirms the correctness Applications of Electrical Energy, IEEE, 2000, 1:147-154.
of our hypothetical mathematical models again. Through the [2] N. Bianchi, S. Bolognani and P. Frare."Design criteria for high-
two actual test , the novel three-dimensional axial primary efficiency SPM synchronous motors,"[J].IEEE Transactions on Energy
Conversion, 2006, 21:396-404.
magnetic structure changes the traditional theory of the
[3] L.-B. Wang, Z.-X. Yang and L. Song,"Power efficiency of outer-rotor
relationship between magnetic flux density and the air permanent magnet wind generator," 2012 9th IEEE International
gap ,establishes a new mathematics model of the magnetic Conference on Networking, Sensing and Control,, IEEE Computer
field : The novel three-dimensional axial primary magnetic Society, 222-227.
structure, there exists a negative slope linear relationship [4] H. Li and Z. Chen."Overview of different wind generator systems and
between magnetic flux density and the air gap. their comparisons,"[J].IET Renewable Power Generation,2008,2:123-
138.
V. CONCLUDING REMARKS [5] K.T. Chau, C.C. Chan and C. Liu."Overview of permanent-magnet
brushless drives for electric and hybrid electric vehicles,"[J].IEEE
This paper describes a new type of topological structure Transactions on Industrial Electronics, 2008, 55:2246-2257.
generator formed of outer rotor permanent magnet, axial [6] H. Chen, R. Qu, J. Li and B. Zhao,"Comparison of interior and surface
primary magnetic and the double-fractional-slot concentrated permanent magnet machines with fractional slot concentrated windings
for direct-drive wind generators," 2014 17th International Conference
windings. This structure design solves the problem of the on Electrical Machines and Systems (ICEMS),2014 , 2612-2617.
traditional permanent magnet with the lager overall start [7] Y. Chen, P. Pillay and A. Khan."PM wind generator topologies,"[J].IEE
reluctance generator torque and the lower wind energy E Transactions. on Industry Applications,2005,41:1619-1626.
conversion rate under the condition of small wind. Through [8] Z.X. YANG, A Novel Outer Rotor Axial Primary Magnetic Circuit
the test of magnetic flux density in the axial primary magnetic Permanent Magnet Generator, [P]. Chinese, ZL201110215832.1, 2014-
structure, the magnetic flux density and air gap is a negative 01-01.
slope of the linear relationship. [9] J.A.Guemes, A.M. Iraolagoitia, M.P.Donsion and J.I. Del
Hoyo,"Analysis of torque in permanent magnet synchronous motors
Studies in this paper are preliminary and future researches with fractional slot windings," 2008 International Conference on
on the outer rotor permanent magnet generator may focus on Electrical Machines , Inst. of Elec. and Elec. Eng. Computer Society,
those aspects: 2008,1-4.
[10] S.-U. Chung, J.-M. Kim, D.-H. Koo, B.-C. Woo, D.-K. Hong and J.-Y.
1 By using the finite element software Magnet, simulate and Lee."Fractional slot concentrated winding permanent magnet
analysis for distribution of the internal magnetic field ;With synchronous machine with consequent pole rotor for low speed direct
drive,"[J].IEEE Transactions on Magnetics, 2012, 48:2965-2968.
the magnetic potential vector, explore the calculation of the
[11] Chen, J., Nayar, C.V. and Xu, L., "Design and finite-element analysis of
key parameters of the magnetic flux leakage coefficient, an outer-rotor permanent-magnet generator for directly coupled wind
calculate pole arc coefficient and optimize the size of the turbines, "2000 International Magnetics Conference,2000, 3802-3809.
design. [12] Y. Tani, S. Morimoto and M. Sanada, "Influence of number of poles,
2 Further studies on the outer rotor axial primary magnetic magnet arrangement, and current density on characteristics of inner and
PMG of the structural characteristics, based on prototype tests, outer rotor PMSMs," 2011 IEEE 9th International Conference on Power
Electronics and Drive Systems, 2011, 711-716.
analyze the aerodynamic characteristics of the motor, the
[13] G.T Paula, J.R.B.A Monteiro, T.E Portes Almeida and M. Patricio
power output characteristics and efficiency characteristics. Santana."Different Slot Configurations for Direct-Drive PM Brushless
Machines,"[J].IEEE Latin America Transactions, 2015, 13:634-639.
3 Off-grid wind turbine cannot be directly connected to the [14] Y. Yubo, W. Xiuhe, Z. Rong, Z. Changqing and D. Tingting,"Research
load, while it must charge the battery stored the energy of cogging torque reduction by different slot width pairing in permanent
through the rectifier at first, and then the battery will magnet motors," 8th International Conference on Electrical Machines
provide electric energy to the load, we hope to increase the and Systems, Conference, Location, Inst. of Elec. and Elec. Eng.
Computer Society, 2005,1:367-370.
electronic commentator composed by the diodes ,aiming to
convert AC to DC, achieving connected to the battery.

23
Proceedings of the 2015 International Conference on
Advanced Mechatronic Systems, Beijing, China, August, 22-24, 2015

Active Training Research of a Lower Limb

Rehabilitation Robot Based on Constrained
Trajectory
Dong Zhang, Hao Lu, Yongfei Feng, Peng Xu
Hongbo Wang Parallel Robot and Mechatronic System Laboratory of
Parallel Robot and Mechatronic System Laboratory of Hebei Province
Hebei Province; [email protected], [email protected]
Key Laboratory of Advanced Forging & Stamping [email protected], [email protected]
Technology and Science of Ministry of Education, Yanshan Razvan-Viorel Mihai, Luige Vladareanu
University, Qinhuangdao China Mechatronics Laboratory, Romanian Academy - Institute of
[email protected] Solid Mechanics, Bucharest, Romania
[email protected],
[email protected]

AbstractIn order to help patients achieve effective and The Lokomat is developed by Hocoma AG and it can simulate
coordinating multi-joints movement easily, the paper presents an the normal walking gait, help patients to exercise limb muscles,
active training method of a lower limb rehabilitation robot based and restore nervous system control of the walking ability [4].
on constrained trajectory. The realization process of the method The MotionMaker is developed by Swortec company and it has
is as follows. The patients movement intention is recognized by different rehabilitation training patterns based on patients
analyzing the signals from torque sensors installed in different rehabilitation stages [5]. Although researchers have
rehabilitation robot joints, and the moving velocity of the developed many kinds of rehabilitation robots, there are still
mechanical arm terminal is calculated through the establishment few rehabilitation robots that can be suitable for patients in
of the damping function and the design of constrained trajectory.
different injured levels. The suspended gait trainers help
The velocity of joints can be calculated through kinematic Jacobi
patients to improve their stand-up ability during the period of
matrix, and is used to drive the joints. The moving velocity of the
mechanical arm terminal on a constrained trajectory can be middle and late recovery. It has a good effect, but is not
controlled through adjusting the damping function, in order to suitable for patients in early stage. The existing sitting/lying
achieve the active rehabilitation training for different patients at gait trainers are not so intelligent that they need patients or
different levels. Finally, the feasibility of the method is validated physiatricians adjust the length of the robots legs manually or
by the experiment. just have sitting mode or lying mode. And the measurement of
joint torque is indirect. Therefore, a multi-joint lower limbs
Keywordsrehabilitation robot; active training; constrained rehabilitation robot (MJLLRR) is presented in this paper, as
trajectory; damping function shown in Fig. 1. The MJLLRR is mainly divided into left
mechanical leg module, right mechanical leg module, movable
I. INTRODUCTION seat and electric control box. Each leg on the leg module is a
series mechanical arm, including the hip joint, knee joint and
At present, many countries in the world have entered the ankle joint which correspond to the joints of human, and can
aging society, physiological recession results in decline of realize three degrees of the freedom movement in the human
limbs activity ability gradually with age and it brings a lot of sagittal plane. The length of the leg and the width between two
inconvenience in daily life, and patients with limbs dysfunction legs can be adjusted by electric to meet the demand of different
who are caused by cerebral apoplexy disease are increasing [1]. patients. The humanization design of movable seat can be
Medical theories and clinical medicine prove that for patients separated from the robot, which is convenient for patient to seat.
the correct and scientific rehabilitation training plays a very Patients can adapt themselves to the training for different level
important role in recovering and improving the limbs motor damage through adjusting the seat pitch angle.
function, in addition to early operation and essential drugs
treatment [2]. With the researches of scholars, many kinds of active
training methods are developed, such as impedance control
The research on lower limb rehabilitation robots has method [6], and force/position hybrid method [7]. Human
become a hotspot. It can be divided into wearable trainers, Adaptive Mechatronics (HAM) is presented by Tokyo Denki
suspended gait trainers and sitting/lying gait trainers. The University, It is an intelligent mechanical system that adapts
Robot Suit HAL is developed by University of Tsukuba and it themselves to human skill under various environments, impro-
is mainly used for patients with late rehabilitation [3].

978-1-4673-7995-3/15/$31.00 2015
c IEEE 24
 Electric and T 2 are respectively the rotation angle of hip joint and knee
control box Movable seat joint. In addition clockwise is negative and counterclockwise is
positive. The rotation shaft center of the hip joint is the origin
Right mechanical of the coordinate system. X and Y are respectively the
Left mechanical
leg module horizontal direction and the vertical direction.
leg module
No-load torque of the knee joint: The gravity G2 of the
crus is decomposed along the normal and tangential directions,
and only the normal component generates the torque because
the extension of tangential component intersects with the knee
joint axis. The normal component of G2 is G2 cos(T1  T2 ) , so
the no-load torque applies on the knee joint is M 2 :
Fig. 1 Multi-joint lower limbs rehabilitation robot
M2 R2G2 cos(T1  T2 ) (1)
-ves human skill, and assists the operation to achieve best
performance of the human-machine system [8]. However, these No-load torque of the hip joint: it is both affected by G1
control methods are very complicated and are not very suitable
and G2 , so the no-load torque caused by G1 is considered
for early stage of rehabilitation training. Therefore, active
training method of a lower limb rehabilitation robot based on firstly. The gravity G1 of thigh is decomposed along the
constrained trajectory is presented The method achieves normal and tangential directions, and only the normal
active training through limiting the movement of patient on the component generates the torque, so the no-load torque
constrained trajectory and implementing damping control on generated by G1 is G1 R1 cos T1 . The distance between the G2
patient force. Comparing to training on unconstrained and the hip joint shaft is l1 cosT1  R 2 cos(T1  T2 ) , then the
trajectory, training on constrained trajectory can help patients
achieve effective and coordinating multi-joint movement easily, torque generated by G2 for hip joint is
and it is more accorded with the normal exercise habits. In this G2 (l1 cosT1  R 2 cos(T1  T2 )) . So the no-load torque applies
paper, the circle which is relatively large in the working space on the hip joint is M 1 :
is chosen as constrained trajectory.
M1 G1 R1 cosT1  G2 (l1 cosT1  R 2 cos(T1  T2 )) (2)
The paper is organized as follows: in section II, the
movement intention of patient is recognized; in section III, After (1) and (2) are integrated, the calibration parameter
constrained trajectory is designed; in section IV, the damping equation of no-load torque can be written as
control strategy is established and the moving velocity of
patients is controlled; Finally, the experiments and results are M1 cos T1 cos(T1  T 2 ) G1 R  G2 L1
M 0 (3)
given.
2 cos(T1  T 2 ) G2 R2

II. RECOGNIZATION OF PATIENTS MOVEMENT INTENTION G1 R  G2 L1

M1 cos T1 cos(T1  T 2 ) 1 0 G2 R2
A. Calculation of No-load Torque of Mechanical Arm M 0 (4)
Hip joint and knee joint of lower limbs rehabilitation robot 2 cos(T1  T 2 ) 0 1 m1

(LLRR) are equipped with torque sensors. In the process of m 2
active training, torque that composes of two parts is measured
by torque sensors: one is no-load torque of mechanical arm,
and the other is generated by the force which is applied by Y
patients on the midpoint of the ankle joint shaft (MOTAJS).
Therefore, the two parts of torque are needed to be analyzed
respectively. This section analyzes no-load torque of
mechanical arm. A T2
According to the prototype of the MJLLRR, the mechanical
arm can be simplified as planar linkage mechanism with 2 R2 l2
l1 fy F
degrees of freedom that is consisted of hip joint and knee joint.
As shown in Fig. 2, points O , A and B are respectively hip G2
R1 G1 T1  T 2
joint, knee joint and ankle joint; l1 and l2 are respectively the
B fx
thigh length and the crus length; R1 and R2 are respectively the
T1
half of thigh length and half of the crus length; G1 and G2 are O
the thigh gravity and the crus gravity; F is the force that X
patients apply in the MOTAJS; f x and f y are respectively the Fig. 2. The static model of equivalent mechanism.
horizontal component and the vertical component of F . T1

25
 Because of the friction torque of each joint, the elasticity of III. THE DESIGN OF CONSTRAINED TRAJECTORY
belt, mechanical clearance and other factors which can also The circle treadmill exercise meets the training
influence the no-load torque, therefore, the (3) need additional requirements of patients better, therefore circle is designed for
amendments as (4). Abbreviating as: constrained trajectory. As shown in fig. 3, the long dashed line
M L(T ) u C (5) and the short dashed line are respectively the thigh length and
the crus length. The curve S1 represents the trajectory of the
The calibration parameters could be calculated by (6): MOTAJS when the hip joint rotates in its rotation range with
C L1 (T ) M (6) the hip and crus keeping parallel. The curve S 2 represents the
trajectory of the MOTAJS when the hip joint rotates in its
The calculation of calibration parameters is as follows: it rotation range with the knee bent to the maximum. The curve
sets several groups of joint angles, and controls mechanical S3 represents the trajectory of the end point when the knee
arm to complete no-load movement; then unknown parameters
joint rotates in its rotation range with the hip in the highest
can be calculated by equation (6) according to joints torque
and actual joints angle after stability movement, and then gets position. The curve S 4 represents the trajectory of the end
the average value of the parameters, finally obtains the point when the knee joint rotates in its rotation range with the
MJLLRR no-load torque equation. hip in the lowest position. The workspace of the MOTAJS is
enclosed by curve S1 , S 2 , S3 and S 4 . Circle O1 which is
B. Calculation of Patients Force tangential with curve S1 , S3 and S 2 is the constrained
This segment calculates the force that patients apply in the trajectory. Straight line PO1 and ON are the auxiliary lines.
MOTAJS.
Through the analysis of the geometric method, tangent
From Fig. 2, the coordinate values of point B can be equation between circle O1 and curve S1 is as follow:
obtained as follows:
x2  y 2 (lON  lO1N )2
xB l1 cos T1  l2 cos(T1  T 2 )
(7)
yB l1 sin T1  l2 sin(T1  T 2 ) Tangent equation between circle O1 and curve S 2 is as
Through the analysis of the geometric method, the hip joint follow:
torque is M1 (t ) : ( x  xP )2  ( y  yP )2 (lPM  lMO1 )2
M1 (t ) f y xB  f x yB (8)
Tangent equation between circle O1 and curve S3 is as
Similarly, the knee joint torque is M 2 (t ) : follow:

M 2 (t ) f y l2 cos(T1  T2 )  f x l2 sin(T1  T2 ) (9) x2  y 2 (lOL  lO1L )2

Combined (8) and (9), it can obtain: Among them:

M 1 (t ) fx lON l1  l2
M (t ) H (10)
2 fy lO1N lO1M lO1L r
Considering the joint torque applied by patients is the
difference between measured torque and no-load torque,
patients force solution can be expressed as: Y H

fx M (t ) M *  M1
F f H 1 H  1* (11) l2
y M 2 (t ) M 2  M 2 S3
S1
In the formula, H  represents the force Jacobian matrix P
inverse, M1 (t ) and M 2 (t ) are respectively the hip and knee
M
joint torques that are applied by patients; M 1* and M 2* are the l1 O1 N
measured torques; M 1 and M 2 are the no-load torques. Thus,
L
at any sampling time in the process of movement, the patients
O X
force applied at the MOTAJS can be calculated, according to
the sampling values of each joints torque sensor and joint S2
angle.
S4

Fig. 3. Constrained trajectory

26
 lPM l2 The damping function f has 3 stages: in the first stage,
F d 0.5 N , F is treated as shake, and the terminal velocity is
lOL l12  l2 2  2l1l2 cos T2 zero; in the second stage, 0.5N d F d 8N , the damping
function f is linear, the 't decreases fast and the terminal
xP l1 cosT1 velocity increases fast when F increase, it is easy for patients
to move and patients can build confidence; in the last stage,
yP l1 sin T1 F t 8N , the slope of f reduce gradually, the 't decreases
The following relations are obtained: slowly and the terminal velocity increase slowly when
F increase, it is hard for patients to move faster and patients
x2  y 2 (l1  l2  r )2 (12) can train their strength.
( x  l1 cosT1 )2  ( y  l1 sin T1 )2 (l2  r )2 (13) In the early rehabilitation training period of patients, weak
f is settled and with the development of rehabilitation training,
x2  y 2 ( l12  l22  2l1l2 cos T2  r )2 (14) f is adjusted for more strong. Through adjusting damping
function f , different mechanical impedance is settled, and
The circle center O1 ( x, y) , radius r can be solved out finally the active rehabilitation training is realized on limb at
through the above three equations. different levels for different patients.
Parametric equations of the circle can be expressed as
follow: B. The Calculation of the Joint Velocity
Obtaining the MAT velocity is not sufficient to control the
x  r cos T
xB mechanical arm, thus the joint velocity is needed.
(15)
yB y  r sin T
The equation between the terminal velocity and the joint
In order to achieve constrained trajectory, the constraint velocity is as follows:
circle is divided equally, the MOTAJS is controlled and then
followed by each equal points in sequence. X J (q ) q (20)
The coordinate of the adjacent equal points are respectively In (17), J (q) represents the kinematic Jacobian matrix, q
( xB (n 1)
y B (n 1) ) , ( xB (n)
y B (n) ) . represents the joint velocity.
The displacement increment of the adjacent equal points is To solve J(q) , both sides of (7) are taken the derivative
's : with time:

'xB xB ( n 1)  xB ( n ) r cos T( n 1)  r cos T( n ) l1 ssin T1  l2 sin(T1  T 2 ) l2 sin(T1  T 2 ) T1
X l
's (16) (21)
'y
B yB ( n 1)  yB ( n ) r sin T( n 1)  r sin T( n ) cos T1  l2 cos(T1  T2 ) l2 cos(T1  T2 ) T 2
l1 co
Y
The Jacobi matrix of the mechanical arm is got:
IV. THE ESTABLISHMENT OF DAMPING CONTROL MODEL
l1 sin T1  l2 sin(T1  T 2 )  l2 sin(T1  T 2 )
J (q ) l cos T  l cos(T  T ) (22)
A. Establishment of Damping Function
1 1 2 1 2 l2 cos(T1  T 2 )
X represents the mechanical arm terminal (MAT) velocity,
and it can be obtained as follow: The joint velocity is obtained:

's
's q J ( q)  X (23)
X (17)
't
C. Realization of Control Strategy of Active Training
't is the time interval which is spent by the MAT when it After (11), (17), (19) and (23) are combined, the
moves between the two adjacent equal points on the rehabilitation robot control algorithm expression is got as
constrained trajectory. 's is calculated in section III. In order to follows:
realize the damping control of MAT, it needs to establish
damping function. J (q) 's
q J (q )  X (24)
't f (F ) (18) f ( H  ( M *  M ))

The specific damping function f is designed as followed: Principle block diagram of rehabilitation robot active
training based on constrained trajectory is shown in Fig. 4.
200 F  2100 0.5 N d F d 8 N Specific work principle is as follows. In a sampling time, it
't (19) measures torque value of each joint by torque sensors, and
4000 / F F t 8N calculates no-load joints torque through each joint angle at the
current moment, then calculates the force which patients apply

27
 Lower Limb of the thigh and crus are given respectively:
Joint Torque Movement Lower l1 =380mm , l2 =325mm . The actual angle of hip joint and knee
Sensor Limb joint were recorded by position sensors during the experiment.
Neural
Feel Robot The actual trajectory of MAT can be calculated by (7). The
Control
Rehabilitation Movement Central actual trajectory and calculated trajectory of MAT are shown in
Robot Nervous Fig. 6.
Velocity Joint
Command Angular In order to show the error between actual trajectory and
Servo Motor Velocity Velocity Robot Inverse
calculated trajectory better, the position error and angle error
are defined in Fig. 7. The position error and angle error appear
Controller Limit Jacobian
when the actual point is not coincided fully with the calculated
point. The position error is the shortest distance between the
Working Constraint Equal Terminal
Space Trajectory Process Velocity
Joints Angle Terminal
No-load Movement Velocity
Joints Time Calculation
Torque Interval
Patient
Damping
Force
Function
Calculation
Measured Patient
Joints Force
Torque

Fig. 4. Principle block diagram of rehabilitation robot active training

Fig. 5. Experiment setup
on the MAT. The motion time interval of the MAT is obtained
through the damping function. The displacement increment of -80
adjacent equal points is got, after constrained trajectory is
processing equally. The terminal velocity is calculated through -100
the displacement increment and the time interval of motion. -120
The angular velocities of each joint of robot are mapped from
terminal velocity by using the inverse kinematical Jacobi -140 Calculated trajectory
Y(mm)

matrix of robot. Angular velocity commands are imported to -160

the servo motor controller after the velocity saturation
limitation, and the stable output of angular velocity is achieved -180 Actual trajectory
by driving servo motor. Finally the robot achieves active -200
rehabilitation training on patients limbs based on constrained
trajectory. -220

It is not difficult to find the principle of active training -240

based on damping control through the above steps and analysis.
Mechanical damping system is modeled from the dynamic 300 350 400 450 500
relationship between the interaction that patients apply on the X(mm)
rehabilitation robot and the robot kinematics. The interaction Fig. 6. The actual trajectory and calculated trajectory of MAT
force is transformed into controlled instruction of the
movement speed of the rehabilitation robot through the Calculated point
mechanical damping. The whole movement process behaves as Actual point
that robot passively follows and assists limb movement. The
robot can provide assistance timely according to the human Angle error
motion intention. At the same time, through adjusting the
damping function parameters, different mechanical damping
can be set, and finally the active rehabilitation training on limb Position error
is achieved at different levels.

V. EXPERIMENT AND RESULTS

The right mechanical leg module is used to do experiment. Cricle center
The experiment scene is given in Fig. 5. The human leg is Constraint trajectory
bound on the pedal and the human hip and knee joints well
match the corresponding joints of mechanical arm. The lengths Fig. 7. The position error and angle error

28
 To be clear, the rehabilitation training robot is mainly
required safety, stability while contacting with the human body.
0.08 It is no strict requirements on the track, so the error is within
0.06 the allowable range.
Relative position error

0.04 The terminal force and terminal velocity are recorded

during experiment. They are shown in Fig. 10. It can be found
0.02 out that terminal velocity is zero when terminal force is less
0 than 0.5N, and damping function f is working at the linear
-0.02
stage when terminal force is more than 0.5N, but third stage
of f dont appear because terminal force is less than 8N during
-0.04 experiment.
-0.06
0 10 20 30 40
VI. CONCLUSION
t(s)
In order to help patients achieve effective and coordinating
Fig. 8. The relative position error multi-joints movement easily, the paper presents an active
training method of a lower limb rehabilitation robot based on
0.2 constrained trajectory. The method achieves active training
through limiting the movement of patient on the constrained
0.1 trajectory and implementing damping control on patient force.
Through adjusting the damping function, it can achieve the
Relative angle error

0
active rehabilitation training for different patients at different
-0.1 levels. Finally, the feasibility of the method is validated by the
experiment.
-0.2

-0.3 ACKNOWLEDGMENT
-0.4 This work is supported by Special Research Found for the
Doctoral Program of Higher Education (no.20131333110006) ,
-0.5 the Natural Science Foundation of Hebei Province of China(no.
0 10 20 30 40
t(s) E2015203405), and the High Level Talent Project in Hebei
Province.
Fig. 9. The relative angle error

REFERENCES
2
[1] G.Z. Xu, A.G. Song and H.J. Li, System design and control technique
Terminal force
1.8 of robot-aided rehabilitation, Journal of Clinical Rehabilitation Tissue
Terminal velocity Engineering Research, vol. 13, pp. 717-717, 2009.
Terminal velocity(mm/s)

1.6 [2] D.W. Jin and J.Z Zhang, System design and control technique of robot-
aided rehabilitation, Modern Rehabilitation, vol. 4, pp. 643-644, 2000
Terminal force(N)

1.4 [3] Yano H, Kasai K, Saito H, et al. Sharing Sense of Walking with
Locomotion Interfaces[J].International Journal of Human Computer
1.2 8 Interaction, vol. 17,pp. 447-462.2005.
[4] Jezernik S, Colombo G, Keller T. Robotic Orthosis Lokomat: A
1 Rehabilitation and Research Tool[J]. Nearomodulation, vol. 6, pp. 108-
105, 2003
0.8 7 [5] Metrailler P, Blanchard V, Perrin I, et al. Improvement of Rehabilitation
Possibilities with the MotionMaker[C]. The First IEEE/RAS-EMBS
0.6 International Conference on Biomedical Robotics and Biomechatronics,
pp. 359-364,2006.
0.4 6
0 10 20 30 40 50 60 70 [6] N. Hogan, Impedance control: an approach to manipulation: part I
t(s) theory,Part IIimplementation, part IIIapplications, ASME
Transactions Journal of Dynamic Systems and Measurement Control B,
Fig. 10. The terminal force and terminal velocity vol. 107, pp. 1-24, 1985.
actual point and the constraint trajectory; the angle error is [7] MASON M T. Compliance and force control for computer controlled
manipulators[J].Systems, Man and Cybernetics, IEEE Transactions, vol.
phase difference between the calculated point and the actual 11, pp. 418-432,1981.
point. The relative position error and relative angle error are [8] Harashima F, Suzuki S. Human Adaptive Mechatronics-Interaction and
respectively shown in Fig. 8 and Fig. 9. In Fig. 8, the relative Intelligence[C]. Proceedings of the 9th IEEE International Workshop on
position error is negative when actual point is inside the circle, Advanced Motion Control, pp.1-8, 2006.
and it is positive when actual point is outside the circle. In Fig.
9, the relative angle error is very big in the beginning, because
the angles of calculated point and actual point are both small.

29
Proceedings of the 2015 International Conference on
Advanced Mechatronic Systems, Beijing, China, August, 22-24, 2015

Fuzzy Controller Design of the Wire Feeder of

Invasive Vascular Interventional Surgery Robot

Hongbo Wang, Shuaishuai Wang, Wen Zhang, Liyu Xue Yang

Xie School of Mechanical Engineering, University of Jinan
Parallel Robot and Mechatronic System Laboratory of Jinan, China
Hebei Province [email protected]
Key Laboratory of Advanced Forging & Stamping
Technology and Science of Ministry of Education, Yanshan
Luige Vladareanu, Radu Munteanu
University, Qinhuangdao China
[email protected], [email protected], Mechatronics Laboratory, Romanian Academy - Institute of
[email protected], [email protected] Solid Mechanics, Bucharest, Romania
[email protected], [email protected]

AbstractA fuzzy controller is established for the wire feeder performance of the wire feeder, the dynamic model of the wire
of the invasive vascular interventional surgery robot in this paper. feeder is constructed. Based on the dynamic model, the fuzzy
First of all, the mechanism of the wire feeder is introduced, and PID controller is established to improve the accuracy of
the axial motion is analyzed. Second, the dynamic model of the position tracking and the anti-jamming ability. The simulation
wire feeder is established. Next, the fuzzy PID controller is shows that the tracking error of the fuzzy PID controller is
established to improve the positional accuracy and dynamic smaller than normal PID controller and the fuzzy PID
performance. At last, the controller is simulated by MATLAB. controller has greater anti-jamming ability.
Comparing with normal PID controller, the fuzzy PID controller
has better dynamic properties and anti-jamming ability. This paper is organized as follows. In section 2, the
mechanism of the wire feeder is introduced, and the dynamic
Keywordsinvasive vascular interventional surgery; wire model of the wire feeder is constructed. In section 3, the fuzzy
feeder; fuzzy PID controller; adaptive control inference rules are analyzed, and the fuzzy PID controller of
the wire feeder is designed. In section 4, the simulation is
I. INTRODUCTION conducted. At last, the conclusion is obtained.
Vascular interventional surgery is that the doctor inserts the
catheter into blood vessels and cure lesions under the guidance II. DYNAMIC MODELING FOR WIRE FEEDER
of the Digital Subtraction Angiography (DSA) so that the
catheter reaches the embolism and malformed vascular to A. Mechancial Realization
dissolve the clot of narrow blood vessels or is used for other The wire feeder is used to clamp, move and rotate the
purposes[1-2]. Vascular interventional surgery has been the surgery guide wire. It consists of a fixed finger, a promoting
key method to cure the cardiovascular disease [3]. But there are finger and rotating finger as shown in Fig. 1. The fixed finger
obvious weaknesses of vascular interventional surgery, such as: is mounted on the front of the wire feeder. It is used to locate
the work environment has lot of X-ray, which will lead to the catheter that has been inserted into vessel and paves a way
occupational disease; it is easy to be fatigued for the doctor, for subsequent wires. The promoting finger locates in the
because of the complex surgery procedure; the surgery needs middle of the device. It is used to grasp the catheter or the
doctors with rich experience. These drawbacks limit the wide guide wire and bring them moving along axial direction. The
application of the vascular interventional surgery [4]. To solve rotating finger locates in the end of the robot. It is used to
this problem, the invasive vascular interventional surgery robot rotate the catheter and guide wire. These three fingers could
should be applied in vascular interventional surgery. Robotic deliver the catheter and guide wire to the lesion location.
manipulation of surgical instruments has many advantages,
such as precise positioning in accordance with the medical The feed schematic of the wire feeder is shown in Fig. 2.
images, executive action without vibration, working in the The driving force of the movement comes from a DC motor.
detrimental environments to human, repositioning or reaching The power outputted from the motor is transferred to a screw
multiple targets through a complex trajectory rapidly and through a pair of meshing gears. The screw rotation drives the
exactly [5]. screw nut moving back and forth. The promoting part of the
mechanism is fixed on the screw nut, so it moves back and
The wire feeder is the direct actuator of the invasive forth following the screw nut. The promoting finger is mounted
vascular interventional surgery robot system. So its dynamic on promoting part. So the promoting finger will follow the
properties and anti-jamming ability are the main indicators of motion of the screw nut. The promoting finger consists of
the robot motion system performance. To improve the double grippers. It could clamp catheter or the guide wire, and

978-1-4673-7995-3/15/$31.00 2015
c IEEE 30
bring them moving along axial direction. The selected motor is d 2T sx
maxon-RE13. Tsx1  J sx1  J sx 2 
 Tsx 3 (2)
dt 2
Fixed figner Promoting finger Rotating finger The kinetics equation of meshing gear could be written as:
d 2T sx
Tsx 3 J sx 3 Tsx 4 (3)
dt 2
d 2T sxc
Tsx 4 J sx 4  Tsx 5 (4)
dt 2
The kinetic equation of screw could be described as:
d 2T sxc
Tsx 5  Tt
J sx 5 (5)
Fig. 1 Wire feeder. dt 2
d 2T sxc
Tsx 5 J sx 5  Tt (6)
Meshing dt 2
gears Torque Tt of screw could be described as:
Gripper
P
Ft Tt (7)
2S
The force between screw and the promoting finger could be
described as:
d 2 xs
Maxon motor Ft  Ff  Fe
Mt (8)
dt 2
The friction Ff, between promoting finger and catheter,
Fig. 2 Feed schematic of wire feeder. includes viscous friction and coulomb friction. It could be
written as:
xs
dxs
ms Fe Ff Pv
 Pc M t g (9)
jsx5 dt
jsx4 The relational equation, between the position xs of the
Tsxc isx5 promoting finger and the angle of the screw, could be
Tsx
Tsx jsx3 described as:
jsx1 jsx2
p
T sxc
xs (10)
Fig. 3 Simplified axial motion
2S
The force Fe is the axial resistance between catheter and
guide wire. Supposing the motion of catheter and guide is
B. Dynamic model of wire feeder along the axis of blood vessel, the catheter and guide is flexible,
Robot system is a nonlinear and uncertain system. So the and the effecting of Fe is very small. So Fe could be ignored.
accurate dynamic model is difficult to establish. In this section,
Based on previous equations, the following equation
to construct the approximate dynamic model of the wire feeder,
describing the relation between the moment of the motor and
its axial motion is simplified as shown in Fig. 3. Dynamic
the axial displacement could be obtained:
model follows the following assumptions(1) Robot system is
a linear system. (2) Ignore uncertain factors, such as friction  J s  J sx1  J sx 2  J sx 3  isx 
from bearings. (3) The weight of catheter and guide line is very 2S d 2 xs
Tsx 2 
2

small. So their inertia and weight could be ignored. (4) The p J sx 4  J xs 5  mt p
dt
frictional resistances, which generate by the axial motion of the 2S (11)

catheter and guide ling along the blood vessel, are very small.
And part of them could be ignored. p 2S dxs p
Pv  isx Bs  Pc mt g
Based on these assumptions, the kinetic equation of the 2S p d t 2S
wire feeder could be deduced as following: Where, p is the lead of the screw. Js, Jsx1, Jsx2, Jsx3, Jsx4, Jsx5
are the rotational inertia of the motor, joint, transmission rod,
The motor kinetic equation could be described as: driving gear, driven gear and screw, respectively. isx is the
d 2T sx dT transmission ratio of the gear. mt is the mass of the promoting
Tsx J s  Bs sx  Tsx1 (1)
dt 2 dt finger. xs is the axial displacement. v, c is the coefficient of
The kinetic equation of joint and transmission rod could be coulomb frication and viscosity friction for the promoting
established as following: finger and guide rail. Bs is the damping coefficient of the motor.

31
 III. CREATING FUZZY CONTROL RULES
In this section, the fuzzy PID controller is designed for the
wire feeder. The key point of designing a fuzzy PID controller
is to find out fuzzy relations between three parameters of PID
and the errors e, ec. Based on the fuzzy rules, PID parameters
could be adjusted to fit the corresponding error e and error
change ec.
To realize the online self-tuning of the three PID
parameters through fuzzy PID controller, it needs the expert
experiences to establish the fuzzy inference rules table. The
range of the input linguistic variables e and ec are {NB, NM,
NS, O, PS, PM, PB}. The range of output linguistic variables
Kp, Ki, Kd are {NB, NM, NS, O, PS, PM, PB} [6]. Fuzzy
inference rules tables of Kp, Ki and Kd are established as
following:

TABLE I. Fuzzy inference rules for Kp a) Nonlinear relation for Kp

e/ec NB NM NS ZO PS PM PB
NB PB PB PM PM PS ZO ZO
NM PB PB PM PS PS ZO NS
NS PM PM PM PS ZO NS NS
ZO PM PM PS ZO NS NM NM
PS PS PS ZO NS NS NM NM
PM PS ZO NS NM NM NM NB
PB ZO ZO NM NM NM NB NB

TABLE II. Fuzzy inference rules for Ki

e/ec NB NM NS ZO PS PM PB
NB NB NB NM NM NS ZO ZO
NM NB NB NM NS NS ZO NS
NS NB NM NS NS ZO PS PS
ZO NM NM NS ZO PS PM PM
PS NM NS ZO PS PS PM PB
PM ZO ZO PS PS PM PB PB
PB ZO ZO PS PM PM PB PB

TABLE III. Fuzzy inference rules table for Kd b) Nonlinear relation for Ki

e/ec NB NM NS ZO PS PM PB
NB PS NS NB NB NB NM PS
NM PS NS NB NM NM NS ZO
NS ZO NS NM NM NS NS ZO
ZO ZO NS NS NS NS NS ZO
PS ZO ZO ZO ZO ZO ZO ZO
PM PB NS PS PS PS PS PB
PB PB PM PM PM PS PS PB

Based on the fuzzy inference, the value of Kp, Ki, Kd

could be abtained. The real values of PID parameters Kp, Ki, Kd
could be described as the linear superposition between Kp,
Ki, Kd and the initial value Kp0, Ki0, Kd0 as show in formula
12.
K p K p 0  'K p

Ki K i 0  'K i (12)
K K d 0  'K d
d c) Nonlinear relation for Kd
Based on upper fuzzy inference rules of PID parameters,
Fig. 4 are drawn to describe the nonlinear relations between Kp, Fig. 4 Nonlinear relations for PID parameters
Ki, Kd and e, ec.

32
 There are lots of methods for narrating fuzzy variables. In IV. SIMULATION BASED ON MATLAB
this section, the COA method is adopted: Since the robot system is a nonlinear and uncertain system,
so it is difficult to establish its accurate dynamics model. In this
u P  x , y ,u 
i u k k i
section, the simulation is based on the linear formula 14. To
uFC  xk , yk  i
(13) verify the dynamic properties and anti-jamming ability of the
P  x , y ,u 
i
u k k i
fuzzy PID controller, the fuzzy PID and the normal PID model
Where, uFC(xk, yk) is the real output from control system. ui are constructed in the fuzzy kit of MATLAB. Comparing
is the discrete element of the output fuzzy set. u(xk, yk, ui) is outputs from two models, it concludes that the fuzzy PID
the membership function for the corresponding element. controller has better performance.
The upper formula is the discretization form of the centroid Based on the formula 11, the dynamic equation could be
defuzzification method. The process for narrating fuzzy obtained. Values of parameters in formula 11 are shown in
variables is offline, and the control strategy table is saved in the following table. Then, the open loop transfer function could be
computer. When the control system is running, the control obtained.
system will read the control strategy table to modify Kp, Ki
xs (s) 28.7923
and Kd. (14)
Tsxc (s) s  0.0077 s
2

TABLE IV. Mechanical parameters of the wire feeder (g/mm2)

Js Jsx1 Jsx2 Jsx3 Jsx4 Jsx5 mt(kg) v p(mm) Bs c isx

0.54 65.13 23.95 408.28 4176.56 398.46 124.80 0.2 1 0.02 0.1 4.44

Fig. 5 Control system model of the fuzzy PID and notmal PID controller

The fuzzy inference system of the robot is constructed 1.5 

Input
through 2-dimensional Mamdani controller. Fuzzy inference Output of fuzzy PI D
rule is constructed according to previous tables. The method 1
Output of norm al PID
Tracking displacem ent(mm)

of Max-min is adopted with fuzzy Control Strategy. The

0.5
method of COA is adopted for narrating fuzzy variables. The
range of fuzzy variables e, ec, Kp, Ki, Kd is designed
0
according to position command, initial error and experience.
To simplify the calculation, the shape of membership function
-0.5
uses the triangle [7-8].
The model for this fuzzy system is constructed in -1

MATLAB as shown in Fig.5. Comparing the fuzzy PID with

normal PID, the scope shows the outputs from fuzzy PID and -1.5 
0 0.5 1 1.5 2
normal PID. Initial parameters of the fuzzy PID could be
obtained by using formula Z-N [9]. Fig. 6 Tracking the sinusoidal signal

33
 0.06  To verify the anti-jamming ability of the fuzzy PID model,
Fuzzy P ID err or
0.04 the system could be simulated by unit step response of the wire
Normal P ID error
feeder. After the wire feeder running for 0.5s, Step signal with
0.02
amplitude 1 mm will be added. The tracking displacement and
the tracking error of the fuzzy PID and normal PID controller
Tracking e rror(mm)

0
will be obtained as shown in Fig.8 and Fig. 9. The tracking
-0.02
displacement with the time is shown in Fig. 8. The tracking
-0.04 error with the time is shown in Fig. 9. Comparing with the
-0.06 normal PID controller, the fuzzy PID has many advantages,
such as smaller overshot, fast response, and shorter time to
-0.08
adjust the error.
-0.1
Comparing the normal PID controller with the fuzzy PID
-0.12 
0 0.5 1 1.5 2
controller, the fuzzy PID controller could track the
Tim e(s) displacement better which will improve the positional accuracy
Fig. 7 The error of tracking the sinusoidal signal of the wire feeder. And it has the greater ability for anti-
jamming, which means it will make fast responses for the
To verify the position tracking performance of the fuzzy jamming.
PID model, the system can be simulated by tracking the
sinusoidal signal with amplitude 1 and frequency 0.1rad/s. V. CONCLUSION
Then the output results from fuzzy PID and normal PID are
In this paper, the dynamic model for the axial motion of the
shown in Fig. 6 and Fig. 7. The tracking displacement with the
vascular interventional surgery robot is constructed. The axial
time is shown in Fig. 6. The tracking error with the time is
motion has many nonlinear factors, so the accuracy for slave
shown in Fig. 7. Based on Fig. 6 and Fig. 7, the track error for
hand tracking the position is not guaranteed by using the
fuzzy PID is better than normal PID, so the fuzzy system has
normal PID control. The fuzzy PID controller is designed for
better dynamic property.
this robot to track the position. The simulation shows the fuzzy
1.6  PID controller is better than normal PID controller and the
Input tracking performance is improved.
1.4 Output of fuzzy PID
Output of normal PID
Tracking displacement(mm)

1.2 ACKNOWLEDGMENT
1 This work is supported by Special Research Found for the
Doctoral Program of Highe r Education (no.20131333110006),
0.8 the Natural Science Foundation of Hebei Province of China (no.
0.6
E2015203405), and the High Level Talent Project in Hebei
Province.
0.4

0.2
REFERENCES
0  [1] W. S. Lu, W. Y. Xu, J. Zhang, et al. Application study of medical robots
0 0.5 1 1.5 2
Time(s) in vascular intervention. The International Journal of Medical Robotics
and Computer Assisted Surgery, 2011, 7(3), 361-366.
Fig. 8 Tracking the step signal
[2] J. Li, N Zhou, S. Wang, et al. Design of an integrated masterslave
0.6  robotic system for minimally invasive surgery[J]. The International
Fuzzy PID error
Journal of Medical Robotics and Computer Assisted Surgery, 2012, 8(1):
0.4 77-84.
Normal PID error
[3] D. Steven, H. Servatius, T. Rostock, et al. Reduced fluoroscopy during
0.2 atrial fibrillation ablation: benefits of robotic guided navigation[J].
Journal of cardiovascular electrophysiology, 2010, 21(1): 6-12.
Tracking error(mm)

0 [4] J. Guo, S. Guo, N. Xiao, et al. A novel robotic catheter system with
force and visual feedback for vascular interventional surgery[J].
-0.2 International Journal of Mechatronics and Automation, 2012, 2(1): 15-
24.
-0.4
[5] R. Saneifard, R. Saneifard. A method for defuzzification based on
-0.6 centroid point[J]. Official J Turk Fuzzy Syst Assoc, 2011, 2: 36-44.
[6] Sun Z, Li Z, Chen L, et al. Creating fuzzy control rules based on data
-0.8 mining[C]//2010 Sixth International Conference on Natural
Computation. 2010, 3: 1635-1639.
-1  [7] L. Da, D. Liu. Accuracy experimental study of the vascular
0 0.5 1 1.5 2
Time(s) interventional surgical robot propulsive mechanism[C]//Complex
Medical Engineering (CME), 2011 IEEE/ICME International
Fig. 9 The error of tracking the step signal Conference on. IEEE, 2011: 412-416.

34
[8] Y. Li, S. Tong, T. Li, et al. Adaptive fuzzy control of uncertain [9] K.S.Tang, K. F. Man, G. Chen, S. Kwong. "An Optimal Fuzzy PID
stochastic nonlinear systems with unknown dead zone using small-gain Controller". IEEE Transactions on Industrial Electronics. vol. 48, NO.4,
approach[J]. Fuzzy Sets and Systems, 2014, 235: 1-24. pp. 757-765, August 2001.

35
Proceedings of the 2015 International Conference on
Advanced Mechatronic Systems, Beijing, China, August, 22-24, 2015

Advanced Intelligent Walking Robot Control through

Sliding Motion Control and Bond Graphs Methods
Ionel-Alexandru GAL and Luige VLADAREANU* Hongbo WANG
Romanian Academy, Parallel Robot and Mechatronic Sys. Lab. of Hebei Province
Institute of Solid Mechanics of Romanian Academy IMSAR Key Laboratory of Advanced Forging & Stamping
Dept. Robotics and Mechatronics Technology and Science of Ministry of Education, Yanshan
Bucharest, Romania University, Qinhuangdao, China
[email protected] [email protected]

Hongnian YU Mingcong DENG

School of Design, Engineering and Computing, The Graduate School of Engineering
Bournemouth University, Dorset BH12 5BB, UK Tokyo University of Agriculture and Technology
yuh@bournemouth. ac.uk 2-24-16 Nakacho, Koganei Tokyo, 184-8588, Japan,
[email protected]

Abstract In this paper, we tested a dynamic control law on a The method uses the analogy effort-flow to describe a
2DOF robot walking leg. To achieve this, we modeled the system process that is graphically represented like some elementary
using bond graphs, and the Sliding Control Method for real time components that have one or multiple ports. These ports are the
control. To increase performance of the robot trajectory tracking places where the systems components interact with each other
control a fuzzy gain adjustment inside the dynamic control [8] (Figure 1).
method have been used and the new results include the systems Every physical process is described by using these few
influence on the sliding parameter. The research results lead to elementary components: effort and flux sources (SE and SF);
improve of the walking robot movement control on unstructured buildup of effort and flow (I and C); power dissipation (R);
and bumped surfaces.
power transformation (TF transformers and GY gyration);
Keywords Sliding Motion Control; Robotics; Bond Graphs;
effort and flux nodes (e and f).
Simulation, Intelligent Walking Robot Control. By using this type of modeling allowed by the bond graphs,
we made a virtual system that is a model of, and simulates, a
walking robots leg with 2 degrees of freedom. This was
I. INTRODUCTION conducted in the Bond Graph simulation environment 20-Sim.
Before testing on a real kinematic structure, one needs to By using this virtual system we could test a dynamic control
design and verify his control laws in a virtual testing law based on Sliding Control Method [2, 10]. This type of
environment. For this we have chosen to use the Bond Graph simulation represents a new method of approaching the design
modeling approach [1, 2]. To use the Bond Graphs modeling and control of walking robots that are controller through a
one can find it rather hard unless he has a database with dynamic control method.
different models previously developed by him or others. The Sliding Motion Control (SMC) is a robust control
Because robotic systems have a strong coupling between their method that has adjustable parameters, and can be easily
dynamics and because model uncertainties make motion modified to adapt on different mechatronic systems. This
control a difficult problem we used Bond Graph modeling to method is used because its dynamic behavior can be modified
reduce these effects. according to specific options [12, 14] and because its closed
Through modeling, one can design a system starting from loop response is imperturbable for a certain group of
its components and can simplify the problem at hand. The uncertainties [6] including nonlinear systems, MIMO systems
reason is because modeling can abstract certain subsets of a and systems that work on discrete time. The SMC has also
model attributes. Thus one can focus on the relevant issues of many advantages and disadvantages [14] and several methods
the problem [3, 5]. that were developed to improve its performances through
In 1959 at MIT, Henry Paynter has designed the well- intelligent approaches [4, 13, 15-17], including fuzzy logic
known approach called Bond Graphs. After that, Kamopp and control and neural network control.
Rosenberg published improvements to the original method to This Sliding Control Method is composed of two main
be used in practice but Breedveld formulated the framework parts. The first one, is formed by the control law which uses the
and Damic and Montgomery [5] continued to develop the system inverse dynamic model, meaning that the SMC in
method which leads to the current form of Bond Graphs. directly dependent on the whole dynamic system. The second
one is the discontinuous and robust term that introduces a high
* Corresponding author

978-1-4673-7995-3/15/$31.00 2015
c IEEE 36
control gain into the system, to minimize and to neglect system Also, we need the numerical values of the physical
uncertainties and disturbances. characteristics of the robot to test the control system. Thus, in
In this paper, well present a modified dynamic control Table 1, these values are presented and are chosen closer to
SMC, developed with the help of bond graphs and for which reality.
we used a fuzzy control law to amplify the command signal for
each joint/motor. The purpose of this paper is to improve a Bond graphs are mainly used in modeling the mechatronic
walking robot movement control on unstructured and bumped system because this component part of the system is the most
surfaces and to achieve improved tracking performances of difficult to simulate. Thus, by using the bond diagram from
position reference. Fig.3, we could simulate the behavior of a joints motor.
Having the diagram of a motor, we could simulate the
rotation joints of the robot. Because the system is simulated,
the initial angular values from the robot joints will always be
zero. This is why, we must always take into account that the
first motion a robot must take is a homing movement.
Fig. 1. A generic element for bond graphs

II. SYSTEM DESCRIPTION

The designed control system was applied to a walking
robots leg that has 2 degrees of freedom with rotation joints.
This system is presented in Figure 2 [7, 9].

Fig. 3 Bond diagram that simulates the behavior of a motor

III. THE DYNAMIC CONTROL USING SLIDING CONTROL

METHOD
For the control system used, we have the main dynamic
equation below:

 H  q  q  C  q, q  q  G  q   W d W 

where, H  q  nun is the inertial matrix, the vectors

q, q, q n are the position, speed and angular acceleration
within the robots joints. The matrix C  q, q  nun represents
Fig. 2. The robot leg structure the Coriolis and centrifugal forces, and G  q  n is the
gravitational vector. Also, W d n represents the disturbances
TABLE I. ROBOT DIMENSIONS
and dynamics vector that are not computed, and W n is the
Parameter Value Type
desired torque.
m1 0.04 kg
Knowing that the system error is:
m2 0.44 kg
m3 0.22 kg
 e qd  q  
l1 0.1 m
l2 0.5 m we chose the sliding surface as given in equation (4) (as
l3 0.7 m shown by S.E. Shafiei in [4]). This value of s tends to 0 when
g 9.8 m/s2 the error e also tends to 0, meaning that lim(s)=0 when
e o0.
The above kinematic structure corresponds to the equations
system in equation (1), in which mi is the mass and li is the
length of each link and g is the gravitational acceleration. t
 s
e  O1e  O2 edt 
0

xM l2 sin T1  l3 sin T2
 y  
M l1  l2 cos T1  l3 cos T2

37
 By using equation (4) we can rewrite the equation (2) and To calculate the Kfuzzy amplification matrix, we need two
obtain equation (5), by replacing the value of q with the sum inputs, s and s for which we have the membership
q+s, which will have the value of the desired value of qd functions in Figure 4a and Figure 4b.
when s equals zero. The value of q becomes qd because
we need to calculate the torque of a desired angular position. The abbreviation from Figure 4a and Figure 4b stands for:
N=Negative, Z=Zero, P=Positive, V=Very, B=Big,
M=Medium, S=Small. The values of these membership
 H  q  q  s   C  q, q  q  s   G  q   W d W  functions are presented in [11]. The control diagram used in the
Sliding Control Simulation is the one presented in Figure 5. In
the control diagram (Figure 5) we can observe two main parts:
By replacing the terms of equation (4) and (5) we get: the sliding control area and the motors joint control. The
diagram has three function blocks that can be attributed to the
 Hs Cs  f  W d  W   sliding control method. The first block is the one that computes
the value of f from the equation (8). This block uses the
values that were computed by many blocks: the reference
Results that we can compute the torque: generation blocks for each joint; the error computation block;
the function block that computes the dynamic matrix [11].
 W f  KV s  Ksat  s   

Where,

t
 f

H  qd  O1e  O2 e   C qd  O1e  O2 edt  G  

0

The equation (6) presents the force estimation, and

t
Kv s
K v e  K v O e  K v O edt is the exterior PID tracking
0
loop, Kv,K are positive diagonal matrix and are built so Fig. 4 (a) Member function for the input s
that the stability conditions are fulfilled. The lambda 1 and 2
matrix were adjusted to get the best results. The sat(s)
function used is the sat function, in which M is a constant that
defines an area around the sliding surface in which the control
parameter K is lowered proportional to the constant M:

1, s d M

s s
 sat , M  s  M  
M M
1, M d s

Fig. 4 (b) Member function for the input s

 K N K fuzzy , KV NV K fuzzy  
The second block that forms the Sliding Control Method is
the block that computes the value s according to equation (4)
and it contains the PID computation of the error e, equation
20 0 10 0 (3). The third SMC control block is the one that computes the
 O1 0 20 , O2 0 10  
reference torque for each joint, by using the equation (7). This
block gets its dynamic data from the block that computes the
To calculate the two matrix K and Kv we use a fuzzy value of f, gets the value of s for a better tracking control,
control law and two constant matrix, N and Nv: and also uses the fuzzy matrix, and also the value of the
sat(s) function. The sat(s) function has the role to keep the
end-effector in a stable area around the reference, by
50 0 5 0 decreasing its amplification value when the end-effector is
 N 0 50 , Nv 0 10   inside the target area.

38
 Fig. 5 The Dynamic Leg Control using Sliding Control and Bond Graphs

By using all these control blocks, we compute the reference respectively joint 2. In these two figures, the control error has
torque values that will be used by the motor joint control three peaks. The first one is due to homing motion at the start
blocks. These values are inputs to the joint PD (proportional of simulation, and the other two are perturbations introduced
derivative) feedback loop, which regulates the torque for each within the reference to test the systems response to rapid
joint torque. For these output signals we added a band filter of change in the reference signal.
+/-100 so the motor real inputs are not too big. Using the
presented control blocks, we have made an internal torque The obtained error in the case of joint number 1 is starting
control feedback loop and an external joint position feedback to stabilize after 0.25 seconds, to an angular error lower than
loop control, by using a dynamic approach along with the +/- 0.015 radians, meaning 0.86 degrees, after which the
Sliding Control Method. system continues to lower the error, but in a slower pace. In the
case of joint number 2, the system reaches in the first 0.25
seconds from the introduced disturbance a +/- 0.05 radians
IV. RESULTS AND CONCLUSIONS error meaning 2.87 degrees, but which continues to decrease in
The control system was made and simulated with the help a slower pace, better presented in Figure 7.
of the 20-Sim modeling environment which has the possibility
of using bond graphs. In Figures 6a, 6b and 7 we presented the
error resulted from the dynamic control modeling with the help
of Sliding Control Method and Bond Graphs.
Research goals involve applied Sliding Motion Control
(SMC), such as design of the advanced intelligent walking
robot control joint with the virtual projection method, known
as Vladareanu-Munteanu method, for design and develop a
real-time control system.

Fig. 6 (b) Joint 2 - Angular Error (rad)

Fig. 6 (a) Joint 1 - Angular Error (rad)

In testing the control system, we chose 2 sinusoidal signals

of 2 and 2.5 radians amplitude. In Figures 6a and 6b we present
Fig. 7. Joint 2 Zoomed in angular Error (rad)
the obtained results according to the simulation for joint 1 and

39
 As presented above, Figure 6b presents the error from the because it uses only simple effort-flow elements to model any
joint with the highest error, but which has a higher reference components. This type of simulation is very accurate, when
for the same frequency of the sinusoidal signal, resulting a done right, and we can build and simulate any type of control
higher movement speed. laws which use these simple elements. Moreover, this type of
simulation doesnt need a solver as in the case of Matlab
The maximum simulation time was chosen at 10 seconds. Simulink, which determines the time step and tolerance of
After 3.5 seconds from the disturbance, the error drops under errors. This means that when simulating certain types of
0.01 radians (0.57 degrees) and after another 1 second the error mechanical and electrical systems, the Bond graph modeling
drops under 0.0075 radians (0.43 degrees) and after that it will can achieve better results
oscillate around the zero axis.
The main conclusion is that this kind of control is one that
tracks very well a uniform reference signal. In the case where
we have sudden changes in the reference signals, the system
may need a higher time to reach a smaller error. These
problems occur because the control system behaves like a
predictive one, which uses the reference speed and acceleration
but also the error speed to eliminate the overrides in the
tracking signals so the robot can track and reach the reference
position as smoothly as possible.

Fig. 8. (a) Joint 1 with value of s

Fig. 8. (b) Joint 2 with value of s

To be able to analyze the sliding control method, we

presented the values of SMC parameter s during the Fig. 9 (a) Joint 1 - Angular Error (rad)
simulation. Figures 8(a) and 8(b) presents the reference, the
tracking signal, the error and the value of s. Thus, one can
observe how the sliding control method really works, by
noticing how the computed value of the PID that returns the
s value oscillates for different values of the reference signal.
When the two disturbances are introduced in the reference
system, the value of s has an instantaneous high amplitude
peak, due to its derivative part. For the difference between the
two s signals (joint 1 and joint 2) we can observe that for
joint 2, the disturbance signal records a bigger peak but it also
maintains a higher value (over +/-10) for another 0.125
seconds, where the signal of joint 1 presents an amplitude peak
only at the moment of disturbance.
To verify our results, we simulated the same conditions in a
Matlab Simulink simulation for two simulated motors. The Fig. 9 (b) Joint 2 - Angular Error (rad)
achieved controller results are presented in Figures 9(a) and
9(b). We can see that the overall conditions and tracking is The second conclusion is that the Bond graph simulations
roughly the same, but the error oscillates between +-0.013 can provide a more accurate simulation for the electric motors.
radians (+-0.745 degrees), which is almost double from the By having a custom library of elements one can create and
error obtained by using Bond graph approach. simulate a certain system in a shorter time. This translates into
a decrease in time and used resources, for testing new and
The difference in simulation results as we have seen in the improved control laws in order to prove research data.
case of Bond graphs compared to Matlab Simulink, is more
accurate when simulating mechanical and electrical parts,

40
 ACKNOWLEDGMENT 7600, vol. 10, no. 2, pp. 146-153, 2010, doi:
10.4316/AECE.2010.02026.
This work was accomplished through the Partnerships [8] Octavian Melinte, Gal I.A.. Bond graph modeling for haptic interface
Program in priority fields - PN II, developed with the support robot control, Proceedings of the European Computing Conference
of MEN-UEFISCDI, PN-II-PT-PCCA-2013-4, ID2009, (ECC 11), Paris, France, April 28-30, 2011, pag: 364-369, ISBN: 978-
VIPRO project no. 009/2014, Romanian Academy and FP7- 960-474-297-4.
PEOPLE-2012-IRSES RABOT project no. 318902. [9] Gal I.A., Luige Vladareanu, Mihai S. Munteanu, Octavian Melinte. PID
sliding motion control by using fuzzy adjustment, SISOM 2012 and
Session of the Commission of Acoustics, Bucharest 25-26 May 2012.
[10] Vladareanu L., Capitanu L. Hybrid Force-Position Systems with
REFERENCES Vibration Control for Improvment of Hip Implant Stability, Journal of
Biomechanics, 45, S1, S279, ISSN 0021-9290
[1] Damic V., Montgomery J.. Bond Graph Based Automated Modelling
Approach to Functional Design of Engineering Systems, Proceedings of [11] Gal I.A., Luige Vladareanu, Radu I. Munteanu. Sliding mode control
the International Conference Mechanics in Design'98, Nottingham Trent with bond graph modeling applied on a robot leg, Revue roumaine des
University GB, 1998, 377-386. sciences techniques Srie lectrotechnique et nergtique, 2015 Issue 2.
[2] Vladareanu L., Velea L.M., Munteanu R.I, Curaj A., Cononovici S., [12] Utkin, V. I. Sliding Modes and their Application in Variable Structure
Sireteanu T., Capitanu L., Munteanu M.S. Real time control method and Systems, MIR Publishers, Moscow, 1978.
device for robots in virtual projection, patent no. EPO-09464001, [13] Vladareanu V., andru O., chiopu P., andru A., Vladareanu L.
18.05.2009, EP2105263. Patent OSIM 123527/30.04.2013 Extension Hybrid Force-Position Control of Mechatronics Systems,
[3] Vladareanu V., Ton G., Vladareanu L., Smarandache F. The Navigation First International Symposium of Extenics, Beijing 2013.
of Mobile Robots n Non-Stationary and Non-Structured Environments, [14] Zhang M., Yu Z, H. H., Z.Y, The Sliding Mode Variable Structure
Int. Journal of Advance Mechatronic Systems 01/2013; 5(4):232- 243. Control Based on Composite Reaching Law of Active Magnetic
DOI: 10.1504/IJAMECHS.2013.057663, ISSN online: 1756-8420, ISSN Bearing, ICIC Express Letters, vol.2, no.1, pp.59-63, 2008.
print: 1756-8412 [15] Kaynak, O., Erbatur K., E.M., The Fusion of Computationally Intelligent
[4] Seyed Ehsan Shafiei (Editor). Advanced Strategies for Robot Methodologies and Sliding-Mode Control: A Survey, IEEE Trans. On
Manipulators, ISBN 978-953-307-099-5, Published by Janeza Trdine, Industrial Electronics, vol. 48, no. 1, pp. 4-17, 2001.
Croatia. [16] Vladareanu V., Sandru O.I., Vladareanu ., Yu H.N. Extension
[5] Damic V., Montgomery J.: Mechatronics by Bond Graphs, Springer Dynamical Stability Control Strategy for the Walking Robots,
Publishing. International Journal of Technology Management,SKIMA 2013,
[6] Lin F. J., Wai R. J., Kuo R. H., Liu D. C. A comparative study of sliding Inderscience Publisher, ISSN online: 1741-5276, ISSN print: 0267-5730
mode and model reference adaptive observers for induction motor drive. [17] Pop N., Vladareanu L., Popescu I.N., Ghi C., Gal I.A., Cang, S., Yu
Electr. Power Syst. Res. 44, 163174, 1998. H.N., Bratu V., Deng M., A numerical dynamic behaviour model for 3D
[7] Vladareanu, L., Tont, G., Ion, I., Munteanu, M. S., Mitroi, D. Walking contact problems with friction, Computational Materials Science,
Robots Dynamic Control Systems on an Uneven Terrain, Advances in Volume 94, November 2014, Pages 285-291, ISSN 0927-0256,
Electrical and Computer Engineering, ISSN 1582-7445, e-ISSN 1844- http://dx.doi.org/10.1016/j.commatsci.2014.05.072.

41
Proceedings of the 2015 International Conference on
Advanced Mechatronic Systems, Beijing, China, August, 22-24, 2015

Versatile Intelligent Portable Robot Platform for

Flexible Robotic Cells with AGV
Luige VLADAREANU, Cristian SPIRLEANU and Mihaiela Mingcong DENG
ILIESCU*
Romanian Academy, The Graduate School of Engineering
Institute of Solid Mechanics of Romanian Academy IMSAR Tokyo University of Agriculture and Technology
Dept. Robotics and Mechatronics,Bucharest, Romania 2-24-16 Nakacho, Koganei Tokyo, 184-8588, Japan
[email protected], [email protected] [email protected]

Hongnian YU Weizhong GUO, Feng GAO

School of Design, Engineering and Computing, School of Mechanical Engineering, Shanghai Jiao Tong
Bournemouth University, Dorset BH12 5BB, UK University, No. 800, Dongchuan Road Shanghai, China
yuh@bournemouth. ac.uk [email protected], [email protected]

Abstract The paper studies the flexible robotic cells in Modern assisted design is achieved with design tools using
cooperation with automated guided vehicle (AGV), in the dedicated software like Computer Aided Design (CAD),
presence of obstacles, at constant or variable speed and variable Computer Aided Manufacturing (CAM) or Computer Aided
load, aiming to optimizing the interaction between AGV and Engineering (CAE) in parallel with the use of simulation
flexible robotic cell components. Overall system performance is programs (e.g. structural and operational analysis,
analyzed by using modeling tools for discrete event systems like manufacturing processes optimization). Resource planning has
Generalized Stochastic Petri Net (GSPN). The interaction an important role, starting with how clients requests are
between AGV and flexible robotic cell components is managed, followed by automated scheduling of materials list
implemented through communication messages using serial data
according to items requested from manufacturing process flow
received from an optical XY encoder, communication protocol
and then launching the items in production based on their
receive function is modeled with GSPN. Improving of the
stability performances and real time motion control are analyzed priority level and time interval availability criteria when the
and the virtual projection method is adopted using the Versatile items are part of a manufacturing subansamble.
Intelligent Portable Robot Platform VIPRO. The obtained For the FMS characteristics improvement, the distributed
results, validated on the experimental RTOS robotic platform intelligence plays a significant role. The use of distributed
and DMQX language extension for robotic applications, lead to control technologies offers the optimal solution through
higher performance in relation to interaction optimization,
communication based controllers which ensure both local and
decrease the flexible cells cycle time, increase mobility and
remote control functions inside a SCADA system [4-6].
stability of the AGV and also the development of new
technological capabilities of the control systems. In the paper is presented a new control strategy based on
communication methods (message exchanges) between the
Keywords intelligent control, virtual projection method, real system components by programming language extension which
time robot control, automated guided vehicle AGV, Generalized contains functions designed for real time operating systems
Stochastic Petri Net, flexible robotic cell (RTOS).

I. INTRODUCTION The control methods presented in this paper combine the

communication protocols applied in artificial intelligence
The paper presents a new approach in flexible cell type (multi agent systems, MAS) with industrial communication
robotic systems using distributed control. Manufacturing protocols (e.g. Modbus, OPC server).
systems have gradually evolve from manual production to
automated lines and serial flexible cells, toward nowadays The paper, also presents a robotic platform with innovative
flexible manufacturing systems (FMS) [1-3]. The industrial hardware-software controllers using real time operating system
robot represent the main component of flexible manufacturing and extension language with data acquisition for later robotic
systems due to its execution capability of discontinued system discrete states evolution and performance analysis.
automated processes having a high complexity level. The By using Petri Nets, the operations flows were modeled for
industrial robot can be re-programmed and is easy adaptable to different configurations of flexible robotic cells using DMQX
manufacturing process, involving lower costs in terms of robotic controllers [7, 8]. The collaboration was achieved
workmanship, energy and materials. between system components as MAS system approach, and
* Corresponding author

978-1-4673-7995-3/15/$31.00 2015
c IEEE 42
between robot components as autonomous agents having the CPN modeling language is a general use language, for example
purpose of different tasks implementation at local level. it is not directed to a specific category but to a large system
classes having in common the process concurrence property.
Through message based communication, an improved Anyway, CPN can also be applied for systems modeling where
collaboration between flexible cell components was achieved processes concurrence and communication represent important
for complex robotic tasks implementation. From the characteristics (e.g. manufacturing systems and multi-agent
programming point of view, the scope was to obtain algorithms systems).
optimization and source code simplicity by keeping the
flexibility characteristic and developing adjacent functions The CPN modeling of the mMbus ASCII communication in
inside the language extension. a network with data loss is shown in Figure 1.
As result, a distributed system for flexible robotic cell with The states noted St, Si and Sr, are connected to the
virtual projection method was implemented, integrated in transitions Tt, Ti and Tr through unidirectional arcs. The states
Versatile Intelligent Portable Robot Platform VIPRO, and are represented with oval forms, marked with one or more
using a communication with modified Mbus protocol with tokens. The token associated value at a given time is identified
DQMX language extension for RTOS robotic controllers. by color. By using the CPN representation convention, the
name of states and transitions is placed inside of the graphical
II. MODELING CPN AND GSPN FOR COMMUNICATION forms. The states are grouped depending on communication
PROTOCOLS IN FRC sequence type:
Colored Petri Nets (CPN) represent a graphical language St = {Transmisie, CaracterTransmis,
used for concurrent systems models construction and their Urmtorul_caracter, ContorCar, RecAcc } (1)
Si = {A, B, C, D, EroareCar, FaraAcc, TP, TC }
properties analysis [9]. CPN is a Discrete Events Systems Sr = { Receptie, RecCar, UrmRec}
(DES) modeling language which combines Petri Net
capabilities with those of the high level programming The colors set for the states noted Transmission, A and B is
languages, compared with simple Petri Nets which ensure the defined as <INT x DATA>, and represents the possible
foundation of graphical notation and the basic primitives for combinations set in matrix (INT, DATA), used for data packet
modeling the concurrence and synchronization. The CPN modeling. For the Reception state is defined <DATA> type,
Modeling Language (CPN-ML) is based on Standard ML used to data packet load modeling and represents the set of
functional programming language [10], which gives the basic character arrays of <STRING> type. For the states
elements for data types defining, for data manipulation CharacterTransmitted, FollowingCharacter, ContorCar,
description and for creating compact and parametric models. ErrorCar, C and D was defined a <INT> type colors set.

Fig. 1 CPN modelling for mMbus-ASCII communication in a network with data loss

43
 The data packet used for sequence communication In this context the VIPRO platform brings back the
evaluation contains de description of a command having > virtual robots into real world, based on virtual projection
start identifier, the address 01, Mbus register number 04, method.
function identifier read sensor from DMQX extension
language having the arguments 01,07,MbusASCII and at This method enables the development in a virtual
the end >terminator character. Thus, the initialisation of environment of mobile robots mechatronic systems that
Transmission state is defined with 6 tokens of characters comminicate with real robotic systems through high speed
arrays type which are concatenated by the ++operator, in interfaces. The result will consist in the development of a
the following form : versatile, intelligent, protable robotic platform, VIPRO,
which will ensure the improvement of motion stability
1` (1,>)++ perfromaces in, both, virtual and real enmvironment, for
1` (2,01)++ flexible robotic cell.
1` (3,04)++ (2)
1` (4,read_sensor)++ The VIPRO platform architecture (Figure 2), in
1` (5,01,07,mMbusAscii)++ correlation with the virtual projection method, has as main
1` (8,>) component the intelligent control interface module. It uses
advanced control strategies adapted to robot environment,
By using this description and composition way for the such as: extended control (extenics), neutrosophic control
communication messages, the CPN model implements the human adaptive mechatronics, etc., implemented by IT&C
DMQX language extension function, with the suppression of techniques, fast-processing and real time communication
limitations imposed by a classical infrastructure from a ones. This module contains mainly the interface for
distributed system which uses a less robust protocol but intelligent neutrosophic control by integrating the Robot
having a large scale usage - Mbus, and adding improved Neutrosophic Control (RNC) method, Extended Control
characteristics (e.g. the retransmission of sequences from the Interface through Extenics (ICEx) and Haptic Robot Control
data packet in case that unpredictable errors were Interface (HRC).
encountered during the communication process).
Robots Neutrosophoc Control Interface (ICNs)
integrates the robot control neutrosophic method, Robot
III. OPTIMIZATION OF INTELLIGENT CONTROL METHODS
Neutrosophic Control (RNC), known as Vladareanu-
USING THE VIPRO PLATFORM
Smarandache method [13].
The intelligent control method allows optimization of
interaction between AGV and flexible robotic cell
components.

Fig.2 The architecture of the VIPRO control system

ICNs uses the neutrosophic logic to fuse information have to consider the systems passive compliance. The
from robot sensors in order to obtain a high accuracy result generalized space (coordinates) where robot works can be
of sensor data. This result is further used in decision making defined in a six degree of freedom constant space,
process of the robot. The force position hybird control of representing force constraints in normal direction to the
flexible cell robots equipped with the compliant joints also working surface and force constratints in tangential direction.

44
 When there is the fusion of information received from problems. The systems extended control enables the
different sensors, a certain rate of conflict is possible to optimization of control process applying extenics theory,
occur. In this case, the robot will use the fuzzy and metrics and extended function to extenics space defined by
netrosophic logics [11-13]. So, real time dynamic the extenics logic. Real time operating consists in
neutrosophic fusion is considered and an autonomous robot implementing the robots extended control inteface (ICEx)
can make a descision at any moment. In order to combine into the basic scheme.
information, there can be applied information fusion theories
(Dezert-Smarandache theory, Dempster-Shafer theory, trust The scientific fundamentals are based on extenics theory
transfer model of Smets) together with different fusion rules, for defining position extended distance, U(X,Xo), force
such as: conflicts proportional distribution, hybrid rule, etc. extended distance,U(Fa,XFo), dependence function
K(X, Xo, XCR ) of the current position signal, X, relative to
More, there can also be applied fuzzy logic / sets and the positive standard interval of reference position, Xo and
neutrosophic logic / sets so that to obtain a combination positive transtion interval of position, XCR, respectively, the
model for the information received from robot sensors, dependence function, K(Fa, XFo, XFCR ), of force signal, Fa,
using neutrosophic logic operators (N-norm si N-co-norm) relative to the positive standard interval of reference force,
representing generalisation of fuzzy operators (T-norm si T- XFo and positive transition interval of force, XFCR, This is
co-norm). how, there is obtained an extended transformation through a
Neutrosophic method of robots control, Robot relationship including a dependence function extended to
Neutrosophic Control (RNC), known as Vladareanu- measurement universe, U, by position extended distance,
Smarandache method [13], due to the application of U(X,Xo) and force extended distance,U(Fa,XFo) thus, leading
neutrosophioc logic and Dezert Smarandache Theory in to the optimization of interaction between AGV and flexible
robots control (DSmT), does stand for a new method robotic cell components.
combining fussy theory and information fusion.
IV. DESIGN OF THE VIPRO OPEN ARCHITECTURE SYSTEM
By implementing the RNC method inot VIPRO platform,
it has been developed a new control interface for Lately, the mobile robots have got special attention
improvement of flexible robotic cell and AGV interaction from, both, research scientists, and manufacturing industry
performances. specialists, so that there resulted important hardware and
The interface of robots extended control integrates the software development
extenics methods of robots real time control, by applying the
extenics theory for solving contradictory problems [14-18].
Extenics is a science dealing with solving contradictory

Fig.3. Virtual projection method by Vladareanu-Munteanu applied to VIPRO Platform

Real time control robots with remotely controlled development of different robots types with sensing,
network and human operator abilities have a highly transport and manipulation abilities for different application
important role in the concept and performance improvement [20, 21].
of flexible robotic cells. Huge research resulted in the

45
 Research activities performed for developing 3D virtual MCS load controller module with the role to ensure the load
environment platforms lead to the gain of important of the m MS servo-actuator modules.
expertise on robots motion in virtual enrironments, based on An MCS load controller module receives the XRP and
high abilities of navigation, cooperation, obstacles XRF, position and force reference and a XRS reference
avoidance, simulation of high reliability virtual signal to generate loads to MS servo-actuator modules, from
environments but, there is no interaction of virtual the ICMF multi-function control interface with the role to
environment : virtual robot real robot. ensure the real time control and the load of the m MS load
actuator modules.
The VIPRO platform mentioned in this paper and
evidenced in Figure 2, is based on the virtual projection In Figure 4, there are presented the simulation results for
method, known as Vladareanu-Munteanu method see the simplified GSPN model of receiving in RF mMbus
Figure 3 [21, 22]. The notations stand for the following: communication protocol, with average number of tokens for
SCMC is the classic control system, which drives the servo- the places and the other states coresponding to other
actuators MS1, MSm,-with m being the number of the variables (e.g. cmd_index, buf[] ) used for modelling.
robots degrees of freedom, and receives signals of TM1-
Receive_USARTdata, Verify_USART_Errors, getc_USART,
TMm measure. The methods refers to an open arhitecture
real time control device for robots. It is aimed at improving Verify_(cmd_index,ch), cmd_receive, respond_to_cmd
control performances and fit for application in control
systems of nano/micro/macro manipulators and robots.
The problem solved by the virtual projection method is
that of enabling to desgin, test and experiment different
control methods on a real time control system, on-line, even
if there is no real mechanical structure [23-25]. There is its
virtual projection and, thus, it is possible to efficiently
improve performnaces of robots control systems, that
already exist.
A corresponding number of load actuators are rigidly
coupled to the servo-actuator modules receiving control
signals from a MCS load controller module which ensures
the load of the MS servo-actuator modules. Also, a number
of m AS load actuator modules rigidly coupled to the m Fig. 4 Simulation results for the simplified mMbus GSPN model
MS servo-actuator modules receive control signals from a

Fig. 5 Robotic platform with DMQX controllers the software architecture

46
 The software architecture implemented for the robotic State space and invariant analysis which indicated that
platform testing (Figure 5) highlights the connection the net is covered by positive T-invariants and bounded,
between software objects for a demo scenario using DMQX with the following P-invariants equation:
robotic controllers,. There are the next requirements:
trajectory generation, objects manipulation and positioning M(buf [cmd_index] = ch) +
feedback loop control for the RC servo motors of the robotic M(cmd_index = 1) + M(cmd_index = 2) +
manipulators. M(cmd_receive) + M(getc_USART) + (3)
M(Receive_USARTdata) +
The simulation results using GSPN for the simplified M(respond_to_cmd) + M(Verify_(cmd_index, ch))
model of receiving with RF mMbus protocol is presented in + M(Verify_USART_Errors) = 1
Figure 6.

Fig. 6 PIPE simulation results for GSPN simplified model of the receiving

V. RESULTS AND CONCLUSIONS software) and intelligent control interfaces (fuzzy,

information fusion, multi-agent, hybrid force position
The user program for robotic application using DMQX dynamic control, robot neutrosophic control, dynamics and
controllers is developed using Microsoft Visual Integrated adaptive, robust and iterative learning control, etc.)
Development Environment (IDE), with functions implemented through IT&C techniques on fast time and high
implemented through language extension that can be called data processing PC server, in order to improve the stability
from customizable Graphical User Interface (GUI). The performances and real time motion control.
DMQX controllers are connected into a wireless For the wireless communication (using 2.4GHz free
communication network, using mBus protocol for the band), the mMbus-RF protocol provides the standard
message transparency during receiving or transmission functionality of Mbus-RTU or ASCII protocol, but
sequences. additionally implements a mechanism for receiving the
The scope is to develop a dynamic versatile intelligent messages of each active node from the network, by
portable robot platform by using 3D virtual representation, configuring additional data registers. The supervisor node
on a PC with high graphic processing power and advanced received information is processed and then displayed in a
programming languages, of robots through mechanical graphical form or is logged on data storage device for later
structure modeling. It is build an open architecture system data analysis and to evaluate the performance of the robotic
made of a robot classic control system (with embedded system.

47
 Starting from the principle of using communication based
digital sensors, for increasing the coordination precision of
the robotic arm movements, a feedback control loop was
implemented for RC servo motors acting, considering the
information provided by an accelerometer for the cinematic
inertia. In Figure 7 is presented the AGV prototype using XY
optical encoder and DMQX controller.
The components are identified as follows :
(1) independent direction wheels controlled by RC
servo motors
(2) accumulators for autonomous supply
Fig.7. The AGV prototype using XY optical encoder and DMQX (3) DMQX robotic controller
controller (4) communication port for programming and testing
the robotic controller
For programming the operations that are involved by the In Figure 8a, is presented the variation along XY axes for
robotic application, the flexible cell is considered as a the positioning information provided by the image sensor of
discrete event system DES. Based on previously simulated XY optical encoder, during the repetition of AGV trajectory
models with instruments for Petri Nets (CPN or GSPN), the which was predefined by the user program configured with
states and the transitions of the system are defined, At tthe graphical user interface. The repetition number defined by
end, through source code recompilation, the robotic the program is rt=10. In Figure 8b are presented the results
application can be verified or simulated by running the obtained for the same trajectory and the same rt repetitions,
custom user program. but with implemented corrective control for the M1 and M2
RC servo motors that drive the AGV wheels.
The obtained studies and research lead to the conclusion
that the innovative VIPRO platform is conceptually
competitive with other similar virtual application platforms
such as CDA, CAM, CAE, Solid Works or MatLab,
Simulink, COMSOL, Lab View, etc., but will also allow the
design, testing and experimentation of intelligent control
methods in real time, integrating the classical robot system in
modeling and simulation of the robot, thereby providing the
opportunity for the VIPRO platform to enter the IT market as
a new component among existing.

ACKNOWLEDGMENT
This work was accomplished through the Partnerships
a) Program in priority fields - PN II, developed with the support
of MEN-UEFISCDI, PN-II-PT-PCCA-2013-4, ID2009,
VIPRO project no. 009/2014, Romanian Academy and FP7-
PEOPLE-2012-IRSES RABOT project no. 318902.

REFERENCES
[1] Damic V., Montgomery J.. Bond Graph Based Automated
Modelling Approach to Functional Design of Engineering Systems,
Proceedings of the International Conference Mechanics in Design'98,
Nottingham Trent University GB, 1998, 377-386.
[2] Vladareanu L., Velea L.M., Munteanu R.I, Curaj A., Cononovici S.,
Sireteanu T., Capitanu L., Munteanu M.S. Real time control method
and device for robots in virtual projection, patent no. EPO-
09464001, 18.05.2009, EP2105263. Patent OSIM 123527/30.04.2013
[3] Craig, J.J.: "Introduction to Robotics, Mechanics and Control",
Addison - Wesley Publishing, New York, 1986
[4] R. Holubek, D.R.D. Sobrino, R. Ruzarovsky : "Analysis of the
b) Communication Methods of an iCIM 3000 System within the Frame
of Research Purpose", World Academy of Science, Engineering and
Fig.8 Positioning precision variation : a) without trajectory adjustment b) Technology,Vol.7, 2013
with trajectory adjustment using feedback from optical sensor
[5] Vladareanu V., Ton G., Vladareanu L., Smarandache F. The
Navigation of Mobile Robots in Non-Stationary and Non-Structured

48
 Environments, Int. Journal of Advance Mechatronic Systems [17] Vladareanu V., andru O., chiopu P., andru A., Vladareanu L.
01/2013; 5(4):232- 243. DOI: 10.1504/IJAMECHS.2013.057663, Extension Hybrid Force-Position Control of Mechatronics Systems,
ISSN online: 1756-8420, ISSN print: 1756-8412 First International Symposium of Extenics, Beijing 2013.
[6] P. Kotl, A. Mudrikov, D.R. Delgado Sobrino : Material Flow in [18] Vladareanu V., Ton G., Vladareanu L., Smarandache F., The
Flexible Production Systems, Proceedings in Manufacturing Navigation of Mobile Robots n Non-Stationary and Non-Structured
Systems, vol.5, no.4, 2010 Environments, Int. Journal of Advance Mechatronic Systems
[7] C.Spirleanu, L.Vladareanu, Mingcong Deng: Positioning control 01/2013; 5(4):232- 243. DOI: 10.1504/IJAMECHS.2013.057663,
through command messages inside a flexible robotic cell with AGV, ISSN online: 1756-8420, ISSN print: 1756-8412
The 20th International Conference on Control Systems and Computer [19] Kaynak, O., Erbatur K., E.M., The Fusion of Computationally
Science (CSCS20-2015), UPB, Romania, 27-29 May 2015 Intelligent Methodologies and Sliding-Mode Control: A Survey,
[8] M. Iliescu, C. Spirleanu, L.Vladareanu: Flexible cell system for IEEE Trans. On Industrial Electronics, vol. 48, no. 1, pp. 4-17, 2001.
grinding process optimization of 40c130 thermal spray coating , [20] Kajita S., Kanehiro F., Kaneko K., Fujiwara K., Harada K., Yokoi K.,
MSE - Sibiu (2015) Hirukawa H., Biped Walking Pattern Generation by using Preview
[9] J.R. Celaya, A. Desrochers, R.J. Graves :Modeling and Analysis of Control of Zero-Moment Point , Proceedings of the 2003 Int.
Multi-agent Systems using Petri Nets, Journal of Computers vol.4, Conference on Robotics and Automation, Taipei, 2003, p. 1620-1626.
nr. 10, 2009 [21] Vladareanu L, Lucian MV, Munteanu RI, Curaj A, Cononovici S,
[10] J.D. Ullman : Elements of ML Programming, Prentice Hall, 1998 Sireteanu T, Capitanu L, Munteanu MS. Real time control method
and device for robot in virtual projection, EU patent no. EPO-
[11] Vladareanu V., Ton G., Vladareanu L., Smarandache F., The 09464001, 18.05.2009, OSIM 123527/30.04.2013
Navigation of Mobile Robots n Non-Stationary and Non-Structured
Environments, Int. Journal of Advance Mechatronic Systems [22] Vladareanu L., Capitanu L., Hybrid Force-Position Systems with
01/2013; 5(4):232- 243. DOI: 10.1504/IJAMECHS.2013.057663, Vibration Control for Improvment of Hip Implant Stability, Journal
ISSN online: 1756-8420, ISSN print: 1756-8412 of Biomechanics, 45, S1, S279, ISSN 0021-9290
[12] Vladareanu V., Schiopu P., Cang S and Yu H.N, Reduced Base [23] Luige Vladareanu, Radu I. Munteanu, Shuang Cang , Hongnian Yu,
Fuzzy Logic Controller for Robot Actuators, Applied Mechanics and Hongbo Wang, Victor Vladareanu, Zeng-Guang Hou, Octavian
Materials Vol. 555 (2014) pp 249-258 (2014) Trans Tech Melinte, Xiaojie Wang, Guibin Bian, Yongfei I Feng, Haptic
Publications, Switzerland doi:10.4028/ interfaces for the rescue walking robots motion in the disaster areas,
www.scientific.net/AMM.555.249 patent, OSIM A2014 00577 , 29/07/201
[13] Smarandache F., Vladareanu L., Applications of Neutrosophic Logic [24] Pop,N., Vladareanu,L., Popescu,I.N., Ghi,C., Gal,I.A., Cang, S.,
to Robotics - An Introduction, The 2011 IEEE International Yu,H.N., Bratu, V., Deng,M., A numerical dynamic behaviour
Conference on Granular Computing Kaohsiung, Taiwan, Nov. 8-10, model for 3D contact problems with friction, Computational
2011, pp. 607-612, ISBN 978-1-4577-0370-6 Materials Science, Volume 94, November 2014, Pages 285-291,
ISSN 0927-0256, http://dx.doi.org/10.1016/j.commatsci.2014.05.072.
[14] Cai Wen. Extension Set and Non-Compatible Problems, Journal of
Scientific Exploration, 1983, (1): 83-97 [25] Kagami S., Nishiwaki K., Kuffner J., Kuniyoshi Y., Inaba M. and
Inoue H., Online 3D vision, motion planning and bipedal locomotion
[15] Yang Chunyan, Cai Wen, Extension Engineering, Science Press,
control coupling system of humanoid robot H7, Proceedings of the
Beijing, 2002.
2002 IEEE/ RSJ International Conference on Intelligent Robots and
[16] Florentin Smarandache, Generalizations of the Distance and Systems, Lausanne, Switzerland, 2557-2562 (2002)
Dependent Function in Extenics to 2D, 3D, and n-D, viXra.org,
http://vixra.org/pdf/1206.0014v1.pdf, 2012.

49
Proceedings of the 2015 International Conference on
Advanced Mechatronic Systems, Beijing, China, August, 22-24, 2015

Haptic Intelligent Interfaces for NAO Robot Hand

Control

Octavian Melinte, Luige Vladareanu*, Ligia Munteanu Zeng-Guang Hou , Gui-Bin Bian,
Dept. Robotics and Mechatronics State Key Laboratory of Management and Control
Institute of Solid Mechanics of Romanian Academy for Complex Systems
Bucharest, Romania Institute of Automation, Chinese Academy of Sciences,
[email protected], [email protected]* Beijing, 100900, China, [email protected]

Hongnian Yu, Shuang Cang Hongbo Wang

School of Design, Engineering and Computing, Parallel Robot and Mechatronic Sys. Lab.
Bournemouth University, Yanshan University
Dorset BH12 5BB, UK Qinhuangdao, China
yuh@bournemouth. ac.uk [email protected]

Abstract. The paper presents a new approach on haptic Due to the compliant feature of impedance controllers and
interface control for NAO robotic hand. The haptic teleoperation the possibility of using force feedback, the most widely used
of NAO robot hand raises some issues, mapping the haptic device control diagram in haptic systems and the one which will be
dynamic being the most important. The mapping was achieved implemented in this research is the impedance control.
through Neural Network by determining a dynamic gain of robot
Impedance control will be discussed for haptic interfaces in
haptic feedback. The results lead to the achievement of the haptic
intelligent interfaces for the NAO robot hand control which can the virtual environment and a straightforward approach will be
be integrated into an innovative haptic robot control system. used for the teleoperation experiment.
A number of environmental factors can act on the haptic
Keywords humanoid robots; haptic interfaces; intelligent devices, such as loads or other forces, and their influence can
control; Nao robot hardly be anticipated. The influence of such disorders may be
a cause of significant deviations from the reference motion of
I. INTRODUCTION the robot in relation to the reference signals.
Haptic interfaces can be implemented in robotic
Humanoid robots represent the most interesting and difficult
applications which use the force feedback from the hands end
area of robotics. This type of robots spread very fast and in the
effector. These applications are useful when the robot is
near future they will accomplish complex tasks, thus replacing
remotely operated, the operator being able to receive tactile
the human in various applications. The state of the art research
information, besides the video and audio as if on site.
in this area focuses on developing intelligent decision making
Humanoid robots controlled by haptic devices can be used in
algorithms, increasing the stability of humanoids robots in
rescue operations, in case of natural disasters or in areas where
unstructured environments, optimizing the planning path, the
access is difficult.
battery efficiency, the increase in autonomy etc. The most
Haptic teleoperation was studied in various forms and
important research on humanoid robots developed so far is
applications. Among them Manic and Odrej [7] achieved
carried for HRP Robot, ASIMO, HOAP, and QRIO [1] - [6].
haptic teleoperation of a mobile robot. In addition to self-
In addition to this research, this paper aims to examine the
organizing algorithm for orientation, this paper proposes a
possibility of implementing a bidirectional communication
fuzzy algorithm mapping between the mobile robot motion
between human operator and a humanoid robot, in this case,
and haptic device. In [8] NAO robot is teleoperated through a
the NAO robot. The main purpose of the research in this paper
network of sensors attached to the human operator joints. This
is the developing of a haptic teleoperated system which
method does not take into account the kinematic model of the
enables the perception of the environment (real or virtual) in
two systems. Another work that discusses about the NAO
which the robot hand is moved to the human operator.
robot teleoperation without taking into account the kinematic
There are two types of haptic interactions: interactions with
model is the one made by Adam et. al [9]. In paper [10, 11]
virtual environments and interactions with real environments
the haptic teleoperation is discussed in terms of linking the
(haptic teleoperation). In this paper we discuss about both
haptic system with the teleoperated one.
possibilities, however the experiments will be performed for
In our approach we will use a simulated haptic device,
haptic teleoperation of NAO hand.
similar to the ones used in similar research (Phantom device),
* Corresponding author with four degrees of freedom, of which we use only three, that

978-1-4673-7995-3/15/$31.00 2015
c IEEE 50
will interact with NAO robot hand. In order to achieve good transposed matrix JT. By closing the control loop a decrease of
results, the haptic device is equipped with encoders on each inertia due to the force from human operator is desired.
joint, while the robot hand has three capacitive sensors on the Further on, the obtained torque is compared with the one from
top of the hand in addition to the joint encoders. Also, in terms user, yielding a torque which will be used to determine the
of degrees of freedom, NAO robot hand has six degrees of position of the motors, and to calculate the dynamic terms of