IEEE Transactions on Mechatronics

Dynamic Characteristics Study of Electro-hydraulic Load-

sensing Control System

Journal: Transactions on Mechatronics

Manuscript ID TMECH-06-2021-12056

Manuscript Type: Regular paper

Date Submitted by the

18-Jun-2021
Author:

Complete List of Authors: Wu, Jiaming; Zhejiang University, Mechanical Engineering

Wang, Feng; Zhejiang University, Mechanical Engineering; University of
Minnesota, Mechanical Engineering
Xu, Bing; Zhejiang university, State Key Laboratory of Fluid Power
Transmission and Control
Sun, Zongxuan; University of Minnesota, Mechanical Engineering;

Pneumatics & hydraulics < Actuators and sensors, Modeling and design,
Keywords:
Design methodology for mechatronics < Modeling and design

Are any of authors IEEE

No
Member?:

Are any of authors ASME

Yes
Member?:
Page 1 of 18 IEEE Transactions on Mechatronics

1
2 Responses to Editors and Reviewers
3
4 The authors sincerely appreciate the comments and suggestions from editors and reviewers to improve the paper quality. The
5
modifications and updates are highlighted in the paper.
6
7
8
Senior Editor
9
10
11 Your paper is interesting. Indeed, load sensing variable displacement pump with pressure difference feedback has its obvious
12
13 advantage compared with hydraulic-mechanical load-sensing pump, but this kind pump is not firstly proposed. The
14 contributions of the paper are design theory, control and experiment validation. Based on the comments of the reviewers and
15 TE recommendation, the paper needs further to increase the novelty and improve the mathematical model, control method and
16
17 experiment.
18
19 The contribution of this paper is further clarified in the updated introduction section. The pipeline model is added and the
20
pump displacement dynamics model is updated in the analysis and simulation. The simulation and experimental studies are
21
22 extended and highlighted in the paper.
23
24
25 Technical Editor
26
27
28 This paper describes a hydraulic load-sensing control system with experimental validation. It is of interest for engineering
29 applications. Based on the review comments, the TE would like to recommend “reject & resubmit”. Before the resubmission,
30
31 the following issues should be well addressed:
32
33
1) The pipe model needs to be improved;
34
35
36 The pipeline model has been updated and highlighted in the subsection “System Modeling” of section II. The fluid
37
compressibility, friction and inertia are considered in the pipeline model where the pipeline is divided into multiple
38
39 segments and nodes. The fluid compressibility is calculated in each node. The fluid friction and inertia are calculated in
40 each segment. This pipeline model is used in the simulation studies.
41
42
43 2) A more thorough literature survey needs to be done with critics that lead to the proposed research;
44
45
46 The “Introduction” section has been updated and highlighted to provide more existing studies of electro-hydraulic load-
47
sensing system. The previous studies of eLS system mainly focus on the system function, controller design and system
48
49 efficiency. However the relationship between system parameters and control performance has not yet been studied,
50
including the influence of pump displacement dynamics on load-sensing control performance, and the influence of preset
51
52 pressure margin on load velocity control performance.
53
54
55 3) The system nonlinearity needs to be addressed;
56
57
58 The pump displacement dynamics is influenced by many factors, including swash plate mass, control pressure and control
59 valve dynamics. The pump displacement control system is a high-order dynamic system. In many studies the relationship
60
between pump displacement change and pump control voltage is presented by a second-order system. This nonlinear
 IEEE Transactions on Mechatronics Page 2 of 18

1
2 model is appropriate to present the pump displacement change in terms of pump control voltage, as illustrated in the
3 pump technical manual. The pump model has been updated and highlighted in the subsection “System Modeling” of
4
5 section II.
6
7
8 The proportional directional valve used in this study is a zero lap high performance valve. To simplify the analysis, the
9 valve dead zone and hysteresis are not considered. In many studies a second-order system is used to present the valve
10
11 opening in terms of control voltage.
12
13
14 The pipeline model used in this study is a nonlinear model where the fluid compressibility, friction and inertia are
15 considered.
16
17
18 4) The experimental validation needs to be corrected.
19
20
21 The pipeline model is added and the pump displacement dynamics model is updated in the analysis and simulation. The
22
23 simulation and experimental studies are updated and highlighted in the paper.
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Page 3 of 18 IEEE Transactions on Mechatronics

1
2 Reviewer 1
3
4
5 1) In my opinion, pressure margin is deeply influenced not only by pump bandwidth but also by losses in pipes associated
6 with delivered flow. In my opinion pipe model is not well described. Both in amesim/siemens simcenter (where a pipe
7
8 model is clearly introduced) and also for the exp. layout.
9
10
The pipeline model has been updated and highlighted in the subsection “System Modeling” of section II. The fluid
11
12 compressibility, friction and inertia are considered in the pipeline model where the pipeline is divided into multiple
13 segments and nodes. The fluid compressibility is calculated in each node. The fluid friction and inertia are calculated in
14
each segment. This pipeline model is used in the simulation studies.
15
16
17 2) Also another limit of the proposed approach is represented by the fact that pressure margin impose the regulation of pump
18
19 pressure respect to a single actuator... but if you limit the usage of a single pump to control a single actuator you can
20 directly consider an electrohydraulic solution in which speed of the pump is controlled by an electric motor as in the
21
22 reference [1].
23
24
25 There are three configurations of electronically-controlled flow source based on electric motor and pump in the eLS
26 system. The first one is to use a variable speed motor and a fixed displacement pump. It is known as electric-hydrostatic
27
28 drive and is attractive in the electrified powertrain owing to its easy speed control. However, the flow rate adjustment
29 becomes slow due to large motor shaft inertia in high power level. Moreover there are oscillation problems in high
30
31 frequency drive and low efficiency in low frequency drive. To obtain faster flow rate adjustment, a variable displacement
32 pump combined with a variable speed motor is used. However this increases system control complexity and cost. A simple
33
34 and cost-effective method is to use a constant speed motor to drive a variable displacement pump.
35
36
37 The configuration of eLS system is dependent on the system performance requirement, power level, control flexibility and
38 cost. These discussions are added and highlighted in the introduction section.
39
40
41 3) For what concern modelling of valves and plants I also suggest to consider contribution/paper [2] also considering cited
42
43 references such as H.E. Merrit. I believe that authors should consider to better emphasize advantages of their approach
44 respect to overcited references [1-2]. Also in Merrit and Karnopp authors should find ref. concerning modelling of pipes
45
46 adopted also by Simcenter Amesim.
47
48
49 The system models are updated including: 1) the pump model is presented by a second-order system. This nonlinear
50 model is appropriate to present the pump displacement change in terms of pump control voltage, as illustrated in the
51
52 pump technical manual; 2) the pipeline model is added where the fluid compressibility, friction and inertia are considered.
53
54
55
56
57
58
59
60
 IEEE Transactions on Mechatronics Page 4 of 18

1
2 Reviewer 2
3
4
5 The manuscript is focused on the important topic of electro load-sensing system and analysis. However, the study has some
6 major flaws that cannot convince the readers.
7
8
9 1) The experimental study is using a fixed displacement pump to animate the load-sensing displacement pump with some
10 electrical feedback. The two system may have very different dynamics. The variable displacement pump with hydraulic
11
12 control piston can demonstrate lots of nonlinearity. Even the simulation and test results seems correlating well. But it has
13 nothing meaningful in practice.
14
15
The pump displacement dynamics is influenced by many factors, including swash plate mass, control pressure and control
16
17 valve dynamics. The pump displacement control system is a high-order dynamic system. In many studies the relationship
18 between pump displacement change and pump control voltage is presented by a second-order system. This nonlinear
19
model is appropriate to present the pump displacement change in terms of pump control voltage, as illustrated in the
20
21 pump technical manual. The pump model has been updated and highlighted in the subsection “System Modeling” of
22 section II.
23
24
25 The reason why using a fixed displacement pump with a bypass proportional directional valve is to simulate variable
26 displacement pump with different pump displacement dynamics by controlling the opening of proportional directional
27 valve. This is easier than using different variable displacement pumps. The contribution of this study is to provide a
28
guidance for eLS system design by investigating the quantitative relationship between pump displacement dynamics and
29
30 load-sensing control performance.
31
32
2) The controller signal Upu is just a value or amount of signals in the paper. But in practice, the Upu has to be very strong
33
34 hydro-mechanical or electro-mechanical forces to jog the swash plate. It means the mechanism of producing the forces
35 and force controlling will be the major topic in the electro load-sensing pump system. However, the paper did not mention
36 and discuss at all.
37
38
39 The pump displacement control system is a high-order dynamic system as the pump displacement dynamics is influenced
40 by many factors, including swash plate mass, control pressure and control valve dynamics. In many studies the
41
relationship between pump displacement and pump control voltage is presented by a second-order system that is the
42
43 result of the factors mentioned above. This nonlinear model is appropriate to present the pump displacement change in
44 terms of pump control voltage, as illustrated in the pump technical manual. The pump model has been updated and
45
highlighted in the subsection “System Modeling” of section II.
46
47
48 3) Some previous literatures and studies have already presented detail design of electro pressure, load-sensing and torque
49 control for open-circuit variable displacement pumps. I suggest the authors should review and include those studies.
50
51
52
53 The existing studies of variable displacement pump control have been added and highlighted in the subsection “System
54
Modeling” of section II.
55
56
57 “In many studies the relationship between pump displacement and pump control voltage is presented by a second-order
58
system that is the result of the factors mentioned above [25]. This nonlinear model is appropriate to present the pump
59
60 displacement change in terms of pump control voltage, as illustrated in the pump technical manual.”
Page 5 of 18 IEEE Transactions on Mechatronics

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
 IEEE Transactions on Mechatronics Page 6 of 18

1
2 Reviewer 3
3
4
5
6 Notably, the following comments need to be carefully considered.
7
8 1) The overall English level needs to be greatly improved.
9
10
11 The paper has been polished by a native English speaker.
12
13
2) There are many variables in this manuscript, so it is better to gather them together at the beginning of the text.
14
15
16 The variables are defined right after each equation to give a better reading. This also conforms to the writing style of the
17 papers in this journal.
18
19
20 3) The model established is simple and common, not challenging.
21
22
23 The system models are updated including: 1) the pump model is presented by a second-order system. This nonlinear
24 model is appropriate to present the pump displacement change in terms of pump control voltage, as illustrated in the
25
26 pump technical manual; 2) the pipeline model is added where the fluid compressibility, friction and inertia are considered.
27 The pipeline is divided into multiple segments and nodes. The fluid compressibility is calculated in each node. The fluid
28
29 friction and inertia are calculated in each segment. This is the same model being used in AMESim.
30
31 4) The research background needs to be further strengthened. And there are many nearly developed advanced control
32
33 strategies that can overcome the model uncertainties of the electro-hydraulic servo systems. So they can be employed as
34 load velocity controllers or/and load sensing controllers.
35
36
The authors agree that the performance of LS control or load velocity control can be improved by using advanced control
37
38 or setting appropriate system parameters, e.g., pump displacement dynamics and preset pressure margin. In this study
39 the focus is the influence of system parameters on the system performance and a simple feedback control with PID
40
controllers is used.
41
42
43 5) Some nonlinear characteristics of the proportional directional valve, especially the dead zone characteristic, have not
44 been fully considered.
45
46
47 The proportional directional valve used in this study is a zero lap high performance valve. To simplify the analysis, the
48
valve dead zone and hysteresis are not considered. In many studies a second-order system is used to present the valve
49
50 opening in terms of control voltage. This nonlinear model is appropriate to present the valve opening in terms of valve
51
control voltage, as illustrated in the valve technical manual.
52
53
54 6) Simulation step size and controller parameters should be clearly listed.
55
56
57 In the simulation a variable-step solver is used. The controller parameters have been added in Table 1.
58
59
7) Some experiments covering low frequency commands and high frequency commands should be carried out.
60

In the experimental study the load velocity frequency ranges from 1 to 30 Hz, and the range is enough to cover the load
Page 7 of 18 IEEE Transactions on Mechatronics

1
2 dynamics. The pump displacement natural frequency ranges from 2 to 10 Hz, corresponding to the step response rise
3 time from 260 to 50 ms.
4
5
8) The format of the references is not standard.
6
7
8 The reference format has been updated to follow the guideline of the journal.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
 IEEE Transactions on Mechatronics Page 8 of 18

1
2
3
4
5
6
7
Dynamic Characteristics Study of Electro-hydraulic
8
9
Load-sensing Control System
10
11 Jiaming Wu, Feng Wang, Bing Xu and Zongxuan Sun
12
13
14 1Abstract—Hydraulic load-sensing control system is an efficient pressure, the LS compensator adjusts pump displacement so
15 system solution to working function of mobile machines. By that the pump outlet pressure has a constant pressure margin
16 sensing the highest load pressure and adjusting the displacement over load pressure [2]. A pressure limiter is to limit maximum
17 of a hydro-mechanical load-sensing pump, the system pressure is
always maintained a constant pressure margin over highest load pump outlet pressure when the pressure is reached [3]. The
18 pressure. However it has poor system dynamic performance and hydraulic LS control system is efficient since it only provides
19 limited flexibility due to hydro-mechanical load-sensing pump. needed flow rate and there is no relief valve in the system [4].
20 An electro-hydraulic load-sensing control system where the However, a notable issue of hydro-mechanical load-sensing
21 pressure control is achieved electronically and only a simple system is the instability problem due to pump outlet pressure
22 displacement controlled pump is needed, offers better system oscillation. Kim found that a large pressure margin in the
performance and larger flexibility. The existing studies focus on
23 hydraulic LS system not only decreases system efficiency but
system application and control strategies of electro-hydraulic
24 load-sensing system, and the system dynamic characteristics have also causes system instability [5][6]. Book proposed a method
25 not yet been researched. Therefore, in this paper the dynamic to improve system stability by adjusting the damping ratio of
26 characteristics of electro-hydraulic load-sensing control system hydro-mechanical feedback control [7]. Birgitta studied the
27 are studied. The influence of pump displacement dynamics on stability of hydro-mechanical LS system with multiple
load-sensing control performance is investigated. The quantitive
28 actuators, and found the instability is due to the interaction
relationship between constant pressure margin and load velocity
29 control performance is also identified. A hydraulic load-sensing between mechanical structure and hydraulic system [8] [9]. Wu
30 control test rig is developed. The system dynamic characteristics investigated the influence of proportional directional valve
31 are verified through both simulation and experimental results. opening on system stability of hydro-mechanical LS system at
32 different system operating regions [10][11].
33 Index Terms—Electro-hydraulic load-sensing control, dynamic A LS solution with electronic control, electro-hydraulic
characteristics, pump displacement control, hydraulic load- load-sensing (eLS) control system becomes popular due its
34 sensing system, mobile machines
35 simple system structure and control flexibility. The schematic
36 is shown in Figure 1 (b). It has some advantages over hydro-
37 I. INTRODUCTION mechanical LS system: 1) the load pressure is sensed via
38 pressure transducers instead of shuttle valve and LS pipeline,
39
40
T HERE has been an increasing demand to reduce the fuel
consumption of mobile machines since they consume
considerable amount of fossil fuel each year. These mobile
improving system response by eliminating pressure dynamics
in the hydro-mechanical feedback loop [12]; 2) it employs an
41 electronically-controlled flow source and a digital LS
machines usually include drivetrain and working hydraulics to controller, allowing for flexible system control [13].
42 achieve transportation and working function. Despite of
43 There are three configurations of electronically-controlled
versatile functions of these machines, the working hydraulics flow source based on electric motor and pump in the eLS
44 consume a large portion of engine power. An efficient solution
45 system. The first one is to use a variable speed motor and a
to working hydraulics is to use a hydraulic load-sensing (LS) fixed displacement pump. It is known as electric-hydrostatic
46 control system where the system pressure is always slightly drive and is attractive in the electrified powertrain owing to its
47 higher than load pressure (constant pressure margin). easy speed control [14]. However, the flow rate adjustment
48 A typical hydraulic LS system consists of a hydro- becomes slow due to large motor shaft inertia in high power
49 mechanical LS pump, a proportional directional valve, a level. Moreover there are oscillation problems in high
50 shuttle valve and a LS pipeline, as shown in Figure 1 (a) [1]. It frequency drive and low efficiency in low frequency drive [15].
51 is essentially a pressure feedback control system. The highest To obtain faster flow rate adjustment, a variable displacement
52 load pressure is selected by shuttle valve and fed to LS pump pump combined with a variable speed motor is used. However
53 through pipeline. By comparing load pressure and pump outlet this increases system control complexity and cost. A simple
54
55
This work is supported in part by the National Natural Science Foundation Jiaming Wu, Feng Wang and Bing Xu are with the State Key Laboratory of
56 of China under Grant 51875509 and 91748210, and in part by the NSFC- Fluid Power and Mechatronic Systems, School of Mechanical Engineering,
57 Shanxi Joint Fund under Grant U1910212. Zhejiang University, Hangzhou 310027, China.
58 Corresponding author: Feng Wang (e-mail: dieter@[Link]). Zongxuan Sun is with the NSF Engineering Research Center for Compact
and Efficient Fluid Power, Department of Mechanical Engineering, University
59 of Minnesota, Minneapolis, MN 55455 USA.
60
Page 9 of 18 IEEE Transactions on Mechatronics

1
2
3
4 and cost-effective method is to use a constant speed motor to eLS system and proposed an anti-saturation flow/pressure
5 drive a variable displacement pump [16]. switched control method to eliminate control performance
6 Many studies have been conducted on the eLS system with deterioration [19][20].
7 constant speed motor and variable displacement pump. Rico The eLS system has been commercialized by some major
8 proposed an energy management strategy for eLS system in a fluid power manufactures. Bosch Rexroth released a
9 forklift to achieve pressure control, flow sharing, steering prior, commercial eLS system applicable to various applications.
10 anti-stall and high pressure protection, reducing system Owing to its flexible algorithm and parameter setting, a
11 complexity considerably [17]. Wang proposed an adaptive seamless interaction between off-road vehicle and working
12 backstepping sliding mode control for eLS system, saving function is achieved [21]. Parker also proposed an eLS solution
13 large amounts of energy compared with fixed displacement where the pressure margin is flexible according to different
pump system [18]. Xu studied the controller saturation in the functions [22].
14
A B
15
LS
16 Shuttle Valve Pipeline
17 A B Pressure
Transducers
18 Proportional P
U
P
U
19 Directional
Valve
20
LS Compensator
21 Proportional
22 Directional
Pressure Valve
23
Limiter
24 P Load Sensing
25 U
Controller
26
Displacement
27 ICE Control Piston
Electronically-
28 M
controlled Flow Source
29
30 Hydro-mechanical LS Pump
31 (a) Hydro-mechanical load-sensing system (b) Electro-hydraulic load-sensing system
32 Figure 1 Schematic of hydro-mechanical LS system and electro-hydraulic LS system
33
34 The previous studies of eLS system mainly focus on the dynamics and load-sensing control performance. The rest of
system function, controller design and system efficiency. the paper is organized as follows. Section II presents the
35
However the relationship between system parameters and mathematical model and control schematic of eLS system.
36
control performance has not yet been studied. One aspect is Section III shows simulation studies. Section IV presents the
37
that the load-sensing control performance is strongly affected experimental results.
38
by pump displacement dynamics since the pressure dynamics
39
at pump outlet is largely influenced by pump outlet flow rate II. ELECTRO-HYDRAULIC LOAD-SENSING SYSTEM
40
change. There are many factors to determine pump
41 A. System Overview
displacement dynamics, including swept volume of control
42 piston and flow capacity of control valve in the pump [23]. The The eLS system studied in this paper consists of a variable
43 rise time of pump displacement from zero to full displacement displacement pump, a proportional directional valve, a single-
44 ranges from one hundred to several hundreds milliseconds rod hydraulic cylinder, three pressure transducers (system
45 depending on the pump size, e.g., 120 ms for 40 cc/rev pump, pressure, load pressure and return pressure) and a piston
46 400 ms for 250 cc/rev pump (from Rexroth datasheet) [24]. velocity transducer, as shown in Figure 2. There are two
47 Another aspect is that the preset pressure margin in the eLS controllers in the system: load-sensing controller and load
48 system is usually set empirically, e.g., 20 bar. There is no select velocity controller. The former is to control pump outlet
49 criterion of this pressure margin. Larger pressure margin pressure by adjusting pump displacement, the latter is to
50 enables better system performance however lowers system control piston velocity by adjusting the opening of
51 efficiency. There is a trade-off between system performance proportional directional valve.
52 and efficiency. B. System Modeling
53 Therefore in this paper the dynamic characteristics of eLS Variable displacement pump
54 system are studied, including the influence of pump
55 The pump displacement dynamics is influenced by many
displacement dynamics on load-sensing control performance,
56 factors, including swash plate mass, control pressure and
and the influence of preset pressure margin on load velocity
57 control valve dynamics. In many studies the relationship
control performance. The contribution of this study is to
58 between pump displacement and pump control voltage is
provide a guidance for eLS system design by investigating the
presented by a second-order system that is the result of the
59 quantitative relationship between pump displacement
60
 IEEE Transactions on Mechatronics Page 10 of 18

1
2
3
4 The viscous friction coefficient 𝜆 is related to Reynolds
Vb Cylinder Vl
P
Fl U Pressure Transducer
5 V number and pipeline roughness in the surface [29].
Velocity Transducer
6 Ab Al U
V
x Proportional directional valve
7 U
pb 2 pl 2
P
U

8 vref vl
P
U The proportional directional valve used in this study is a
Load Velocity Ⅲ Ⅱ
9 Controller pb1
zero lap high performance valve. To simplify the analysis, the
pl1
10
B A valve dead zone and hysteresis are not considered. In many
uv Proportional
pref studies a second-order system is used to present the valve
11 T P
Directional
Valve opening in terms of control voltage [30][31]:
12 Qb Ql ps 2 Load Sensing
P
13 U
Controller
2𝜉 𝜔 𝜔 𝑥 𝑘 𝜔 𝑢 (5)
14 Ⅰ
ps1
15 upu where 𝑥 is valve opening, 𝑢 is valve control voltage, 𝜔 and
Qs
16 e 𝜉 are natural frequency and damping ratio of valve opening
E- Electronically-
17 motor controlled Pump dynamics, 𝑘 is valve control gain.
18 The inlet flow rate of pipeline Ⅱ and outlet flow rate of
19 pipeline Ⅲ are given by:
20 Figure 2 Schematic of electro-hydraulic load-sensing system
21 factors mentioned above [25]. This nonlinear model is ⎧𝐶 𝐴 𝑥 𝑝 𝑝 , 𝑖𝑓 𝑢 0
22 appropriate to present the pump displacement change in terms 𝑄 (6)
23 of pump control voltage, as illustrated in the pump technical ⎨𝐶 𝐴 𝑥 𝑝 𝑝 , 𝑖𝑓 𝑢 0
24 ⎩
manual. The pump displacement dynamics is given by [26][27]:
25
26 2𝜉 𝜔 𝜔 𝑉 𝑘 𝜔 𝑢 (1) ⎧𝐶 𝐴 𝑥 𝑝 𝑝 , 𝑖𝑓 𝑢 0
27 𝑄 (7)
where 𝑉 is pump displacement, 𝑢 is pump control voltage, ⎨𝐶 𝐴 𝑥 𝑝 𝑝 , 𝑖𝑓 𝑢 0
28 ⎩
29 𝜔 and 𝜉 are natural frequency and damping ratio of second-
30 order system, 𝑘 is control gain. where 𝐶 is flow coefficient, 𝐴 𝑥 is throttling area, 𝑝 , 𝑝 ,
31 Given an estimated damping ratio, the pump displacement 𝑝 are pressures at valve port P, A and B, 𝑝 is tank pressure.
32 dynamics is presented by its natural frequency. Hydraulic cylinder
33 The pump outlet flow rate, 𝑄 , is given by: The pressure dynamics of the chambers on both sides of
34 𝑄 𝜔 𝑉 𝑘𝑝 (2) hydraulic cylinder are given by:
35
36 where 𝜔 is pump shaft speed, 𝑘 is leakage coefficient, 𝑝 is 𝑄 𝐴𝑣 (8)
37 pump outlet pressure.
38 Pipeline 𝐴 𝑣 𝑄 (9)
39 There are three pipelines in the system: the pipeline at pump
40 where 𝑝 , 𝑝 , 𝑉 , 𝑉 are pressures and chamber volumes on
outlet (Ⅰ), two pipelines between proportional directional valve
41 both sides of hydraulic cylinder, 𝐴 and 𝐴 are acting areas of
and hydraulic cylinder (Ⅱ and Ⅲ). To show the pressure
42 piston chamber and rod chamber, 𝑣 is load velocity.
dynamics in these pipelines, the fluid compressibility, friction
43 Ignoring frictional and viscous forces, the piston dynamics
and inertia are considered in the pipeline model where the
44 is:
pipeline is divided into multiple segments and nodes. The fluid
45 compressibility is calculated in each node. The fluid friction 𝑚 𝐴𝑝 𝐴 𝑃 𝐹 (10)
46 and inertia are calculated in each segment [28].
47 where 𝑚 is piston mass, 𝐹 is piston load force.
48 𝑄 𝑄 (3)
C. Controllers Design
49
𝑝 𝑝 𝜆 𝑠𝑔𝑛 𝑄 (4) There are two controls in the eLS system: load-sensing
50
control and load velocity control. The former is to control
51
where 𝑝 is node pressure, 𝛽 is fluid bulk modulus, 𝑉 is node pump outlet pressure by adjusting pump displacement so that
52
volume, 𝑄 and 𝑄 are inlet and outlet flow rates of each a constant pressure margin over load pressure is achieved. The
53
node, 𝑄 is flow rate of each segment, 𝐴 is cross-sectional area latter is to control piston velocity by adjusting the opening of
54
of pipeline, 𝜌 is fluid density, 𝐿 is segment length, 𝑝 and proportional directional valve. The block diagram of eLS
55
𝑝 are inlet and outlet pressures of each segment, 𝜆 is system is shown in Figure 3.
56
viscous friction coefficient, 𝐷 is pipe diameter, s𝑔𝑛 𝑄 is the In the LS control, the reference input is desired system
57
sign function of flow rate 𝑄 (flow direction). pressure that is the sum of load pressure and constant pressure
58 margin. A pressure feedback control is used where a simple
59
60
Page 11 of 18 IEEE Transactions on Mechatronics

1
2
3
4
5 Load velocity control
vlref
6 _
Load-sensing control uv Velocity ev +
7 controller
8 _ Qb pb
9 pref + psref +
ep LS up Qs ps Proportional
Hydraulic
Hydraulic vl
10 controller
Pump Pipe
valve
Ql cylinder pl piston
+ and pipes
11
12 Fl
13
14
15 Figure 3 Block diagram of electro-hydraulic load-sensing control system
16 proportional-integral-derivative (PID) controller is to generate 𝜔 . The natural frequency at the peak point increases with load
17 pump control signal. In the load velocity control, the reference force frequency. The results provide a quantitative relationship
18 input is desired piston velocity and a feedback control is used.
19 A proportional-integral (PI) controller is to generate valve
20 control signal. There are some interactions between the two
21 feedback control loops. The load flow rate is a disturbance in
22 the LS control and the pump outlet pressure is a disturbance in
23 the load velocity control.
24
25 III. SIMULATION STUDY
26 A. Simulation Parameter
27
To investigate dynamic characteristics of eLS system, a
28 physical equation based dynamic simulation model is
29 developed. To conduct a comparison between simulation and
30 experimental results, the simulation parameters are obtained
31 from the developed test bench, as shown in Table 1. A variable-
32 step solver is used for the calculation. The simulation studies
33 include: 1) the relationships between pump displacement
34 dynamics and LS control performance; 2) the relationship
35 between preset pressure margin and load velocity control
36 performance.
37 Figure 4 System pressure, load pressure, pressure margin and
38 B. Load-sensing Control Simulation load flow rate with the input of variable load velocity and
39 The system pressure, load pressure, pressure margin and force
40 load flow rate with the input of variable load velocity and force
41 are shown in Figure 4. The load velocity has a step change at
42 1 s and a sinusoidal change between 7 s and 10 s. The load
43 force has a step change at 3 s and a sinusoidal change between
4 s and 6 s. Results show the effectiveness of the LS control
44
and the pressure margin is around the preset value of 20 bar.
45 There are some small step changes when the load velocity or
46 load force has a step change. Sinusoidal changes occur when
47 the load velocity or load force has a sinusoidal change (4 ~ 6 s,
48 7 ~ 10 s). The pressure margin changes are reflected on the
49 load flow rate changes. This conforms to previous analysis that
50 the load flow rate is a disturbance in the LS control. Both the
51 load velocity and load force changes contribute to load flow
52 rate changes.
53 The pressure margin fluctuation in terms of pump
54 displacement natural frequency at different load force
55 frequencies are shown in Figure 5. The pump displacement
56 natural frequency ranges from 2 to 10 Hz, corresponding to the Figure 5 Pressure margin fluctuation with different pump
57 step response rise time from 260 to 50 ms. Given certain load displacement natural frequencies (load force: 𝐹 30600
force frequency, the pressure margin fluctuation first increases 2000 𝑠𝑖𝑛 2𝜋𝑓𝑡 N, 𝑓 load force frequency, preset pressure
58
and then decreases with pump displacement natural frequency margin ∆𝑝 20 bar)
59
60
 IEEE Transactions on Mechatronics Page 12 of 18

1
2
3
4 Table 1 Simulation parameters of electro-hydraulic load-sensing control system
5
6 Property Parameter Value Unit
7 Variable displacement pump Full displacement 10.0 cm3/rev
8 Natural frequency 2, 6, 10 Hz
9 Damping ratio 0.7 —
10 Control gain 1.0 cm3/(rev∙V)
11 Leakage coefficient 3 10 m3/(Pa∙s)
12 Electric motor Rated speed 1460 rpm
13 Pipeline Internal diameter 12.7 mm
14 Length of pipe Ⅰ 3.0 m
15 Length of pipe Ⅱ and Ⅲ 1.0 m
16 Number of pressure nodes 5 —
17 Proportional directional valve Maximum throttling area 5.79 mm2
18 Natural frequency 80.0 Hz
19 Damping ratio 0.8 —
20 Hydraulic cylinder Piston stroke 409.0 mm
21 Piston diameter 62.4 mm
22 Rod diameter 41.7 mm
23 Piston mass 22.52 Kg
24 Fluid property Bulk modulus 1387.7 MPa
25 Density 870 Kg/m3
26 Kinematic viscosity 4.6 10 m2/s
27 Load-sensing controller Proportional gain 1.0 10 V/Pa
28 Integral gain 4.0 10 V/(Pa∙s)
29 Differential gain 5.0 10 V∙s /Pa
30 Load velocity controller Proportional gain 5 V∙s /m
31 Integral gain 500 V/m
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Figure 6 Pressure margin fluctuation with different pump Figure 7 Pressure margin fluctuation with different load force
49
displacement natural frequencies (load velocity: 𝑣 0.02 frequencies (load force: 𝐹 30600 2000 𝑠𝑖𝑛 2𝜋𝑓𝑡 N,
50 0.001 𝑠𝑖𝑛 2𝜋𝑓𝑡 m/s, 𝑓 load velocity frequency, preset
51 𝑓 load force frequency, 𝜔 pump displacement natural
pressure margin ∆𝑝 20 bar)
52 frequency, preset pressure margin ∆𝑝 20 bar)
53 between pump displacement natural frequency and pressure
load velocity frequencies are shown in Figure 6. Results show
54 margin fluctuation. This not only gives an estimation of
similar trends with Figure 5.
55 pressure margin fluctuation with pump displacement dynamics,
The pressure margin fluctuation in terms of load force
56 but also provides a pump selection guidance to achieve desired
frequency at different pump displacement natural frequencies
pressure margin fluctuation. The pressure margin fluctuation
57 are shown in Figure 7. Given a pump displacement natural
in terms of pump displacement natural frequency at different
58 frequency, the pressure margin fluctuation first increases and
59 then decreases with load force frequency. By selecting a pump
60
Page 13 of 18 IEEE Transactions on Mechatronics

1
2
3
4 with higher displacement natural frequency, the pressure C. Load Velocity Control Simulation
5 margin fluctuation is smaller when the load force frequency is The preset pressure margin in the LS control is selected
6 relatively low. This conforms to the real case where the load empirically in practice, e.g., 20 bar. It is to set enough pressure
7 force dynamics is not high. drop across the proportional directional valve to assure system
8 The pressure margin fluctuation in terms of load velocity control performance. Previous studies show that the system
9 frequency at different pump displacement natural frequencies
control performance deteriorates when this pressure drop
10 are shown in Figure 8. The trend is similar to Figure 7. The
decreases in the hydro-mechanical LS system. On the other
11 peak point load velocity frequency 𝜔 at different pump
hand, large pressure drop leads to severe throttling loss and
12 displacement natural frequency 𝜔 can be calculated based on
thus low system efficiency. There is a trade-off between system
13 the system model in section Ⅱ: control performance and energy efficiency [6][32].
14 In this study the influence of preset pressure margin on load
15 𝑉 𝐾 𝐾 𝜔 𝛽
𝜔 𝜔 velocity control performance in the eLS system is investigated.
16 𝑉 The bode plots of load velocity control at different preset
17 pressure margins are obtained, as shown in Figure 9. As the
18 where 𝑉 is volume of pipeline Ⅰ, 𝐾 , 𝐾 , 𝐾 are PID gains in
preset pressure margin increases from 5 to 30 bar, the cut-off
19 the load-sensing control.
frequency (-3 dB) increases significantly, while the phase
20 Based on the system parameters in Table 1, the peak point
delay decreases. Results show that a larger preset pressure
load velocity frequency is 2.33 times of pump displacement
21 margin allows for a broader bandwidth of load velocity control.
natural frequency. The analysis results and simulation results
22 are compared, as shown in Table 2. It shows a good agreement
23 between analysis and simulation results. There are small
24 discrepancies between two results. This is explained that the
25 peak point load velocity frequency calculation does not take
26 return pressure dynamics, fluid inertia and friction in pipeline
27 into consideration.
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42 Figure 9 Bode plots of load velocity control at different preset
43 pressure margins (pump displacement natural frequency
44 Figure 8 Pressure margin fluctuation with different load 𝜔 10 Hz)
45 velocity frequencies (load velocity: 𝑣 0.02
46 0.001 𝑠𝑖𝑛 2𝜋𝑓𝑡 m/s, 𝑓 load velocity frequency, 𝜔
47 pump displacement natural frequency, preset pressure margin
48 ∆𝑝 20 bar)
49
Table 2 The peak point load velocity frequency at different
50 pump displacement natural frequencies
51
52 Pump disp. Peak point load velocity
53 natural frequency (Hz)
54 frequency (Hz) Analysis Simulation
55 2 4.66 4.70
56 6 13.98 13.60 Figure 10 Cut-off frequencies of load velocity control at
57 10 23.30 22.90 different preset pressure margins (pump displacement natural
58 frequency 𝜔 10 Hz)
59
60
 IEEE Transactions on Mechatronics Page 14 of 18

1
2
3
4 The cut-off frequencies of load velocity control at different  Valve-controlled hydraulic cylinder under test: it includes a
5 preset pressure margins are shown in Figure 10. The cut-off proportional directional valve, a single-rod hydraulic
6 frequency increases with preset pressure margin. This shows cylinder (driving) and two relief valves. The hydraulic
7 that the load velocity control performance is improved with a cylinder (driving) is the control object in this study;
8 larger preset pressure margin, however this is achieved by  Hydraulic loading control unit: it includes a single-rod
9 decreasing system efficiency. This is consistent with the result hydraulic cylinder (loading), a proportional relief valve and
10 in the hydro-mechanical LS system. a solenoid on-off valve. The hydraulic loading control unit
11 is to simulate the load force acting on the piston of hydraulic
12 IV. EXPERIMENTAL STUDY cylinder (driving) under test. The load force is controlled by
13 adjusting the relief pressure of proportional relief valve. The
A. Test Bench Setup
14 solenoid on-off valve is used to reset the piston of hydraulic
15 To verify analysis and simulation results, a test bench of eLS
cylinder (driving) to initial position.
16 system is developed, with the schematic shown in Figure 11. It
The test bench photograph is shown in Figure 12 and the
17 consists of three subsystems:
system parameters are shown in Table 1. Some parameters are
18  Simulated variable displacement pump: it includes a fixed identified experimentally and used in the simulation study.
19 displacement pump driven by a constant speed electric
motor, and a bypass proportional directional valve at pump B. Simulated Variable Pump Outlet Flow Rate
20
outlet. By controlling the opening of proportional A fixed displacement pump with a bypass proportional
21
directional valve, the outlet flow rate of a variable directional valve is employed to generate desired outlet flow
22
displacement pump with displacement dynamics is rate of a variable displacement pump. Given desired pump
23
simulated;
24
Valve-controlled Hydraulic Hydraulic Loading
25 Cylinder Under Test Control Unit
26 Hydraulic Hydraulic
27 Cylinder (driving) Cylinder (loading)
Velocity
28 Pressure U
V
U Transducer
Transducer P

29
30 U Pressure
31 Transducer
P

Relief Proportional Relief

32 Valve Directional Valve 1 Valve
33
Pressure Transducer U

Proportional
34
P

Relief Valve
35
Solenoid On-
36 off Valve
37
38
39 Proportional
Electric Fixed Directional
40 Motor Pump Valve 2
41
42 Simulated Variable
Displacement Pump
43
44 Figure 11 Test bench schematic of electro-hydraulic load-sensing control system
45
46 Hydraulic Cylinder (loading)
47
48 Hydraulic Cylinder (driving)
Proportional Directional Valve 2
49
50 Proportional Relief Valve
51 Relief Valve
52
53 Tank
54 Proportional Directional Valve 1
55
Fixed Pump
56
Solenoid On-off Valve
57
58
59 Figure 12 Test bench photograph of electro-hydraulic load-sensing control system
60
Page 15 of 18 IEEE Transactions on Mechatronics

1
2
3
4 displacement dynamics, the corresponding time-varying outlet determine the outlet flow rate of fixed displacement pump at
5 flow rate is simulated by controlling the opening of bypass given outlet pressure in Eq. (11). The pressure drop – flow rate
6 proportional directional valve. The desired time-varying outlet characteristics of proportional directional valves under
7 flow rate of variable displacement pump, 𝑄 𝑡 , is given by: different valve control voltages are obtained experimentally, as
8 shown in Figure 14. The relationship between the valve
9 𝑄 𝑡 𝑄 𝑝 𝑡 𝑄 𝑥 𝑡 ,𝑝 𝑡 (11) opening and control voltage is obtained. With the valve
10 characteristics, the valve control voltage is calculated given the
where 𝑄 is outlet flow rate of fixed displacement pump, 𝑄
11 desired valve flow rate and actual valve pressure drop.
and 𝑥 are flow rate and opening of bypass proportional
12 directional valve.
13 The flow rates in Eq. (11) are all time-varying. The outlet
14 flow rate of fixed displacement pump is influenced by pump
15 outlet pressure due to its leakage. The flow rate of bypass
16 proportional directional valve is determined by valve opening
17 and pressure drop across the valve. To accurately generate the
18 desired outlet flow rate of variable displacement pump, the
19 flow rates of fixed displacement pump and bypass proportional
20 directional valve are identified experimentally.
21
22
23
Figure 15 Simulated variable pump outlet flow rates with
24
different displacement dynamics (𝜔 pump displacement
25
natural frequency)
26
27 The simulated variable pump outlet flow rates at different
28 pump displacement dynamics are shown in Figure 15. The
29 pump displacement natural frequency ranges from 2 to 10 Hz,
30 corresponding to the step response rise time from 260 to 50 ms.
31 The step response rise times are consistent with analysis results.
32 This shows the effectiveness of proposed method.
33 C. Load-sensing Control Test
34
The comparisons of pressure margin fluctuation between
35 Figure 13 Outlet flow rates of fixed displacement pump at
different outlet pressures simulation and test at different load velocity frequencies are
36
shown in Figure 16. The test results conform to simulation
37 results well. The possible reasons of small discrepancies
38
39
40
41
42
43
44
45
46
47
48
49
50
51 Figure 14 Pressure drop – flow rate characteristics of
52 proportional directional valve 2 under different valve control
53 voltages
54 The outlet flow rates of fixed displacement pump at different
55 outlet pressures are measured, as shown in Figure 13. The Figure 16 Comparisons of pressure margin fluctuation
56 measurements are performed at constant fluid temperature to between simulation and test at different load velocity
57 reduce the influence of fluid temperature on the fluid viscosity frequencies (load velocity: 𝑣 0.02 0.001 𝑠𝑖𝑛 2𝜋𝑓𝑡 m/
58 and flow rate measurement. By numerical fitting, the pump s, 𝑓 load velocity frequency, 𝜔 pump displacement
59 leakage coefficient is obtained. The measured data are used to natural frequency, preset pressure margin ∆𝑝 20 bar)
60
 IEEE Transactions on Mechatronics Page 16 of 18

1
2
3
4 between two results include: 1) the fluid bulk modulus in the simulation results. A larger preset pressure margin allows for
5 simulation and test may be not exactly the same; 2) the wall better load velocity control performance however the system
6 expansion of pipeline in the test is not considered in the efficiency is lower since the pressure drop across proportional
7 simulation. directional valve is larger. This shows a trade-off between load
8 Results show that given certain pump displacement natural velocity control performance and system efficiency.
9 frequency, the pressure margin fluctuation first increases and
10 then decreases with load velocity frequency. The peak point V. CONCLUSIONS
11 load velocity frequency increases with pump displacement The dynamic characteristics of eLS control system are
12 natural frequency. This shows that a variable pump with fast studied in this paper. Compared with conventional LS system
13 displacement dynamics is helpful to reduce pressure margin powered by a complex hydro-mechanical LS pump, the eLS
14 fluctuation, showing consistence with simulation results. control system employs a simple electronically-controlled
15 flow source. By using pressure transducers and digital LS
D. Load Velocity Control Test
16 controller, it offers better system performance and larger
17 The load velocity tracking performance at different preset
flexibility. The system dynamic characteristics studied in this
18 pressure margins are obtained experimentally, as shown in
paper include: the influence of pump displacement dynamics
19 Figure 17. Given desired sinusoidal load velocity, the actual
on LS control performance and the influence of preset pressure
20 load velocity at different preset pressure margins (10, 20 and
30 bar) are compared. The actual load velocity with a larger margin on load velocity control performance. A test bench of
21 eLS system is developed to verify analysis and simulation
preset pressure margin has less attenuation in amplitude and
22 results.
smaller delay in phase. This shows that a larger preset pressure
23 The quantitative relationship between pump displacement
margin is helpful to improve load velocity control performance.
24 dynamics and LS control performance is obtained. The
25 pressure margin fluctuation first increases and then decreases
26 with pump displacement natural frequency. This not only gives
27 an estimation of pressure margin fluctuation with pump
28 displacement dynamics, but also provides a pump selection
29 guidance to achieve desired LS control performance. By
30 selecting a pump with higher displacement natural frequency,
31 the pressure margin fluctuation is smaller when the load
32 dynamics is relatively slow. The influence of preset pressure
33 margin on load velocity control performance is also studied.
34 Results show the system with a larger preset pressure margin
35 allows for better load velocity control performance however
36 this results in lower system efficiency.
37
38 Figure 17 Load velocity tracking performance at different ACKNOWLEDGMENTS
39 preset pressure margins (desired load velocity: 𝑣 0.02 This research is conducted in the State Key Laboratory of
40 0.005 𝑠𝑖𝑛 20𝜋𝑡 m/s) Fluid Power and Mechatronic Systems at Zhejiang University.
41
42 REFERENCES
43 [1] J. Nurmi and J. Mattila, “Global energy-optimal redundancy resolution
44 of hydraulic manipulators: Experimental results for a forestry
manipulator,” Energies, vol. 10, no. 5, pp. 647, 2017.
45 [2] B. Xu and M. Cheng, “Motion control of multi-actuator hydraulic
46 systems for mobile machineries: Recent advancements and future trends,”
47 Frontiers of Mechanical Engineering, vol.13, no.2, pp. 151-166, 2018.
[3] H. C. Pedersen, T. O. Andersen, and M. R. Hansen, “Load sensing
48 systems-a review of the research contributions throughout the last
49 decades,” in Internationales Fluidtechnisches Kolloquium (IFK), pp.
50 125-139, 2004.
[4] M. Axin, B. Eriksson, and P. Krus, “Flow versus pressure vontrol of
51 pumps in mobile hydraulic systems,” in Proceedings of the Institution of
52 Mechanical Engineers, Part I: Journal of Systems and Control
53 Figure 18 Comparisons of load velocity cut-off frequency Engineering, vol. 228, no. 4, pp. 245-256, 2014.
between simulation and test results at different preset [5] S. D. Kim, H. S. Cho, and C. O. Lee, “Stability analysis of a load-sensing
54 hydraulic system,” in Proceedings of the Institution of Mechanical
55 pressure margins (pump displacement natural frequency Engineers, Part A: Journal of Power and Energy, vol. 202, no. 2, pp. 79-
56 𝜔 10 Hz) 88, 1988.
[6] S. D. Kim and H. S. Cho, “Suboptimal controller design method for the
57 The comparisons of load velocity cut-off frequency between energy efficiency of a load-sensing hydraulic servo system,” Journal of
58 simulation and test results at different preset pressure margins Dynamic Systems, Measurement and Control, Transactions of the ASME,
vol. 113, no. 3, pp. 487-493, 1991.
59 are shown in Figure 18. The test results conform well to
60
Page 17 of 18 IEEE Transactions on Mechatronics

1
2
3
4 [7] R. Book and C. E. Goering, “Load sensing hydraulic system simulation,” [32] Q. Hu, H. Zhang, S. Tian, and X. Qin, “Performances analysis of a novel
5 Applied Engineering in Agriculture, vol. 13, no. 1, pp. 17-25, 1997. load-sensing hydraulic system with overriding differential pressure
[8] B. Lantto, J. O. Palmberg, and P. Krus, “Static and dynamic performance control,” in Proceedings of the Institution of Mechanical Engineers. Part
6 of mobile load-sensing systems with two different types of pressure- C, Journal of Mechanical Engineering Science, vol. 231, no. 23, pp.
7 compensated valves,” SAE transactions, pp. 251-265, 1990. 4331-4343, 2017.
8 [9] B. Lantto, P. Krus, and J. O. Palmberg, “Dynamic properties of load-
sensing systems with interacting complex mechanical loads,” ASME J.
9 Dyn. Syst. Meas. Control, vol. 115, pp. 525-530, Sep. 1993. Jiaming Wu received B.S. degree in
10 [10] D. Wu, “Modeling and experimental evaluation of a load-sensing and
mechanical engineering from Zhejiang
11 pressure compensated hydraulic system,” PhD Thesis, University of
University, Hangzhou, China, in 2019. He
Saskatchewan, Saskatchewan, Canada, Dec. 2003.
12 [11] D. Wu, “Steady state analysis of the load sensing systems,” in is currently pursuing his Ph.D. degree in
13 Proceeding of the Second International Ph. D. Conference, 2002. mechanical engineering at Zhejiang
14 [12] T. Wang and Q. Wang, “An energy-saving pressure-compensated
University, Hangzhou, China.
hydraulic system with electrical approach,” IEEE/ASME Trans.
15 Mechatronics, vol. 19, no. 2, pp. 570–578, Apr. 2014. His main research focuses on modeling
16 [13] D. Lovrec and S. Ulaga, ‘‘Pressure control in hydraulic systems with and control of hydraulic load-sensing
17 variable or constant pumps?’’ Exp. Techn., vol. 31, no. 2, pp. 33–41, 2007. system for off-road vehicles.
[14] P. Andrea, L. Pugi, F. Vinattieri, R. Capitani, and C. Annicchiarico,
18 “Design and Testing of an Innovative Electro-Hydraulic Actuator for a
19 Semi-Active Differential,” in Proceedings of the Institution of
20 Mechanical Engineers, Part D, Journal of Automobile Engineering, vol.
232, no. 11, pp. 1438-1453, 2018.
Feng Wang received B.S., M.S., and
21 [15] D. Lovrec, M. Kastrecvc, and S. Ulaga, “Electro-hydraulic load sensing Ph.D. degrees in mechanical engineering
22 with a speed-controlled hydraulic supply system on forming-machines,” from Zhejiang University, Hangzhou,
23 Int. J. Adv. Manuf. Technol., vol. 41, no. 1112, pp. 1066-1075, Jun. 2008. China, in 2003, 2005, and 2009.
[16] W. Backe and B. Zähe, “Electrohydraulic load sensing,” SAE
24 transactions, pp. 303-311, 1991.
In 2009 he became a Postdoctoral
25 [17] R. H. Hansen, T. O. Andersen, and H. C. Pedersen, “Development and Associate in the Department of
26 implementation of an advanced power management algorithm for Mechanical Engineering at University of
electronic load sensing on a telehandler,” in Proc. Bath/ASME Symp. Minnesota, Minneapolis, MN, USA,
27 Fluid Power Motion Control, pp. 537-550, 2010.
where he joined the NSF Engineering Research Center for
28 [18] W. Wang and B. Wang, “An Energy-Saving Control Strategy with Load
Sensing for Electro-Hydraulic Servo Systems,” Strojniski Compact and Efficient Fluid Power (CCEFP). In 2016 he
29 became an Assistant Professor in the School of Mechanical
Vestnik/Journal of Mechanical Engineering, vol. 62, no. 12, pp. 709-716,
30 2016. Engineering at Zhejiang University, Hangzhou, China, where
31 [19] B. Xu, M. Cheng, W. Liu, and H. Yang, “A new electrohydraulic load
he joined the State Key Laboratory of Fluid Power and
32 sensing control system for hydraulic excavators,” in 8th International
fluid power conference (IFK), vol. 1, pp. 553-565, Mar. 2012. Mechatronics Systems. His research interests include
33 [20] M. Cheng, J. Zhang, B. Xu, and R. Ding, “An electrohydraulic load modeling and control of intelligent hydraulic powertrains for
34 sensing system based on flow/pressure switched control for mobile heavy-duty vehicles and large utility wind turbines.
machinery,” ISA transactions, vol. 96, pp. 367-375, 2020.
35 [21] U. Lenzgeiger and U. Maier, “Electronic load sensing for tractors,” in
36 11th International Fluid Power Conference, 2018.
37 [22] “Electronic Load Controlled Pump eLCP.” [Online]. Available:
Bing Xu received Ph.D. degree in
[Link]
38 F%20files/HY02-8057%[Link], Accessed: 9 May 2021. mechanical engineering from Zhejiang
39 [23] N. D. Manring and V. S. Mehta, “Physical limitations for the bandwidth University, Hangzhou, China, in 2001. He
40 frequency of a pressure controlled, axial-piston pump,” ASME J. Dyn. is currently a Professor in the school of
Syst. Meas. Control, vol. 133, pp. 061005-1-061005-12, Nov. 2011.
41 [24] “Axial piston variable pump A4VSO.” [Online]. Available:
mechanical engineering and also the
42 [Link] Director of the State Key Laboratory of
43 groups/industrial-hydraulics/pumps/axial-piston-pumps/variable- Fluid Power Transmission and Control at
pumps-open-circuit/a4vso, Accessed: 9 May 2021. Zhejiang University.
44 [25] J. Koivumäki and J. Mattila, “Adaptive and Nonlinear Control of
45 Discharge Pressure for Variable Displacement Axial Piston Pumps,”
He has authored or co-authored more
46 Journal of Dynamic Systems, Measurement, and Control, vol. 139, no. than 200 journal and conference papers, and authorized 49
10, 2017. patents. His research interests include fluid power components
47 [26] R. H. Hansen, A. M. Iversen, M. S. Jensen, T. O. Andersen, and H. C. and systems, mechatronic systems design and motion control
48 Pedersen, “Modeling and control of a teletruck using electronic load
for mobile machinery. Prof. Xu is a Chair Professor of the
49 sensing,” in ASME 2010 10th Biennial Conference on Engineering
Yangtze River Scholars Program and a science and technology
Systems Design and Analysis, pp. 769-778, 2010.
50 [27] G. P. Jayaraman and S. V. Lunzman, “Parameter estimation of an innovation leader of the Ten Thousand Talent Program.
51 electronic load sensing pump using the Recursive Least Squares
52 algorithm,” in 49th IEEE Conference on Decision and Control (CDC).
IEEE, pp. 3445-3450, 2010.
53 [28] I. E. Idelchik, “Handbook of hydraulic resistance,” Washington, 1986.
54 [29] H. E. Merritt, “Hydraulic control systems,” New York: Wiley, 1967.
55 [30] W. Thayer, “Transfer functions for Moog servo-valves,” Moog technical
bulletin 103, New York: Moog, Inc, January 1965.
56 [31] L. Pugi, E. Galardi, G. Pallini, L. Paolucci, and N. Lucchesie, “Design
57 and Testing of a Pulley and Cable Actuator for Large Ball Valves,” in
58 Proceedings of the Institution of Mechanical Engineers. Part I, Journal
of Systems and Control Engineering, vol. 230, no. 7, pp. 622-639, 2016.
59
60
 IEEE Transactions on Mechatronics Page 18 of 18

1
2
3
4
5 Zongxuan Sun (M’01) received the B.S.
6 degree in automatic control from
7 Southeast University, Nanjing, China, in
8 1995, and the M.S. and Ph.D. degrees in
9 mechanical engineering from the
10 University of Illinois at Urbana–
Champaign, Champaign, IL, USA, in
11
1998 and 2000, respectively.
12
He is currently the Director of the Engineering Research
13 Center for Compact and Efficient Fluid Power and a Professor
14 of mechanical engineering with the University of Minnesota,
15 Twin Cities, Minneapolis, MN, USA. He was a Staff
16 Researcher (2006–2007) and a Senior Researcher (2000–2006)
17 with the General Motors Research and Development Center,
18 Warren, MI, USA. He has been the author or coauthor of more
19 than 130 refereed technical papers and received 22 U.S. patents.
20 His research interests include controls and mechatronics with
21 applications to transportation propulsion systems. Dr. Sun is a
22 recipient of the Charles E. Bowers Faculty Teaching Award,
23 George W. Taylor Career Development Award from College of
24 Science and Engineering, University of Minnesota, NSF
CAREER Award, SAE Ralph R. Teetor Educational Award,
25
two best paper awards, Inventor Milestone Award, Spark Plug
26
Award, and Charles L. McCuen Special Achievement Award
27 from GM R&D.
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60