Lecture note

NONLINEAR AND ADAPTIVE CONTROL

Instructor: Assoc. Prof. Dr. Huynh Thai Hoang

D
Department
t t off A
Automatic
t ti CControl
t l
Faculty of Electrical and Electronics Engineering
Ho Chi Minh City University of Technology
Email: [email protected]
Homepage:
p g http://www4.hcmut.edu.vn/~hthoang/
p g

26 August 2017 © H. T. Hoàng - HCMUT 1

 Course Objective
 To provide
T id students
t d t withith ffundamental
d t l kknowledge
l d
about nonlinear and adaptive control systems. The
students will learn approaches to analyze the
stability of nonlinear control systems and methods
to design controllers for nonlinear and unknown
plants
l t

26 August 2017 © H. T. Hoàng - HCMUT 2

 Course outline
 Ch t 1:
Chapter 1 Introduction
I t d ti
 Chapter 2: Nonlinear Control
 Chapter 3: Adaptive Control
 Chapter 4: Optimal Control (Optional)

26 August 2017 © H. T. Hoàng - HCMUT 3

 Textbooks and References
 Lecture
L t note:
t Nonlinear
N li and
d Ad
Adaptive
ti C Control
t l – Huynh
H h
Thai Hoang, Published in BKel
 References:
[1] Applied nonliear control
control, Jean-Jacques
Jean Jacques EE.
Slotine & Weiping Li, Prentice-Hall International
Editions Inc.
Editions, Inc 1991.
1991
[2] Adaptive control , Karl Johan Astrom & Bjom
Witt
Wittnmark k , Addison-Wesley
Addi W l P Publishing
bli hi
Company, second edition 2000.

26 August 2017 © H. T. Hoàng - HCMUT 4

 Grading scheme
 Homework
H k and
d exercise:
i 20%
 Quiz: 20%
 Midterm exam: 20%
 Final exam: 40%

26 August 2017 © H. T. Hoàng - HCMUT 5

 How to learn the course? ACTIVE LEARNING
(*) Aft
After 2 weeks
k we
tend to remember…
10% off what
h t we read
d R di
Reading
20% of what we hear Hearing words
30% of what we see Looking at pictures
Watching a movie
50% of what we Looking at an exhibit
see and
dhhear Watching a demostration PASSIVE
Seeing it done on location
70% off what
h t P ti i ti iin a di
Participating discussion
i
we say ACTIVE
Giving a talk
90% of what Doing a dramatic presentation
we say Simulating the real experience
and do Doing the real thing
(*) Edgar Dale, “Audio-Visual Methods in Teching,” Holt, Rinehart and Winston
26 August 2017 © H. T. Hoang - HCMUT 6
 Ch t 1
Chapter

INTRODUCTION

26 August 2017 © H. T. Hoàng - HCMUT 7

 Primitive control systems
 Before
B f the
th middle the 19th century:
iddl off th t control
t l systems
t
were designed based on heuristics and experiences.

26 August 2017 © H. T. Hoàng - HCMUT 8

 Primitive control systems

26 August 2017 © H. T. Hoàng - HCMUT 9

 Classical control (before 1960)
 From the
F th middle the 19th century
iddl off th t to
t the
th 1930s:
1930
mathematical foundation of control theory was
developed:
 Lyapunov (1892)
 Routh-Hurwitz (1895)
 From
F th 1930s
the 1930 tot 1940s:
1940 classical
l i l control
t l theory
th
was growing fast:
 Frequency response, Bode plot
 Nyquist plot,
plot Nichols plot,
plot
 Root locus method

26 August 2017 © H. T. Hoàng - HCMUT 10

 Characteristics of classical control
 SISO systemt
 Linear plants (or nonlinear plants working around
equilibrium points)
 Design in frequency domain
 Using plots in design
 Design
D i to t trade-off
t d ff between:
b t
 Performance and robustness
 Effects of uncertainties
 PID controllers

26 August 2017 © H. T. Hoàng - HCMUT 11

 Limitation of classical control theory
 Control
C t l performance:
f
 is not optimal
 is
i d degraded
d d iin th
the cases th
thatt nonlinear
li systems
t
working in a wide range.
 is
i nott ensured d when
h ththe plants
l t h have titime-varying
i
parameters.
 It is
i nott sure to
t have
h the
th design
d i performance
f if the
th
mathematical model have uncertainties.
 It is difficult to apply classical control theory in the
following cases:
 MIMO systems
 Nonlinear systems
 Time domain performance
26 August 2017 © H. T. Hoàng - HCMUT 12
 Example of complex systems
 C
Cranes, robot
b t arms

26 August 2017 © H. T. Hoàng - HCMUT 13

 Example of complex systems (cont.)
 Shi steering
Ship t i system
t

26 August 2017 © H. T. Hoàng - HCMUT 14

 Example of complex system
 Ai l
Airplane control
t l system
t

26 August 2017 © H. T. Hoàng - HCMUT 15

 Example of complex system (cont.)
 P
Process control
t l

26 August 2017 © H. T. Hoàng - HCMUT 16

 Modern control (1960 – present)
 Nonlinear
N li control
t l
 Optimal control
 Adaptive control
 Robust control

Thi course focuses

This f on nonlinear
li and
d adaptive
d ti control
t l

26 August 2017 © H. T. Hoàng - HCMUT 17

 Nonlinear control
 Mathematical
M th ti l fundamentals
f d t l off nonlinear
li control
t l theory
th
were developed in the 19th century. After 1960, nonlinear
control theory has been growing fast.
fast
 Approaches to analysis of nonlinear systems:
 Describing function method
 Phase plane method
 Lyapunov
y p theory
y
 Popov criterion
 Small gain theorem
 Methods to design nonlinear control systems
 Feedback linearization control
 Sliding mode control
 Back-stepping control,…

26 August 2017 © H. T. Hoàng - HCMUT 18

 Adaptive control
 Adaptive
Ad ti control
t l iis a control
t l method
th d iin which
hi h th
the
controller’s parameters are automatically adjusted online
so that the desired control performances are ensured
even if the working conditions are changed.
 In the 1950s: Model Reference Adaptive System
 In the end of 1960s: advance in control theory
contributed to the development
p of adaptive
p control ((state-
space equation, Lyapunov’s stability theory, dynamic
programming,…)
 In the 1970s and the beginning of 1980s: stability
analysis of adaptive control systems
 From the end of 1980s to the beginning of 1990s:
robustness analysis of adaptive control systems
 R
Recently:
tl llearning
i controlt lhhas b
been ddeveloped
l d
26 August 2017 © H. T. Hoàng - HCMUT 19
 General block diagram of adaptive control systems

Identification/
Estimation
Working
o g
conditions
Adjusment

y(t))
y(
r(t) C t ll
Controller Plant
u(t)

26 August 2017 © H. T. Hoàng - HCMUT 20

 Practical application of adaptive control
 Mostt off the
M th manufacturer
f t off controller
t ll (Siemens,
(Si
Omron, Autonics, Schneider, Precision Digital,...)
have auto-tuning controller for multi-range of
application such as temperature control
control, motor speed
control, process control,…
 Modern
M d Di t ib t d Control
Distributed C t l System
S t (DCS) softwares
ft
have adaptive control function.
 PCS7 (Siemens)
 Delta V (Emerson)

26 August 2017 © H. T. Hoàng - HCMUT 21

 Practical application of adaptive control

26 August 2017 © H. T. Hoàng - HCMUT 22

 Practical application of adaptive control

26 August 2017 © H. T. Hoàng - HCMUT 23

 Practical application of adaptive control

26 August 2017 © H. T. Hoàng - HCMUT 24

 Practical application of adaptive control

26 August 2017 © H. T. Hoàng - HCMUT 25

 Course outline
 Nonlinear
N li control:
t l
 Mathematical model of nonlinear systems
 Describing
D ibi ffunction
ti methods
th d
 Lyapunov stability
 Feedback
F db k lilinearization
i ti control
t l
 Sliding mode control
 Adaptive
Ad ti control t l
 Model reference control
 Model reference adaptive system (MRAS)
 On-linear parameter estimation
 Self-tuning regulator (STR)
 Gain scheduling control
 System analysis, design and simulation using Matlab
26 August 2017 © H. T. Hoàng - HCMUT 26
 Related theory
 Prerequisite
P i it course:
 Advanced mathematics (Linear Algebra,
Differential Equation, Numerical Methods,…)
 Introduction of Control Systems
 Follow-up advanced courses
 Nonlinear
N li control
t l
 Multivariable control
 Robust and optimal control
 Adaptive control

26 August 2017 © H. T. Hoàng - HCMUT 27

 Learning outcomes
On sucessful
O f l completion
l ti off this
thi course, students
t d t
should have the following knowledge and skills:
 Ability to analyze and design of nonlinear control
system using describing function method and
Lyapunov stability theory;
 Ability
y to design
g feeback linearization controller,
sliding mode controller;
 Ability to design model reference adaptive control
systems, seft-tuning regulators, gain-scheduling
controllers
 Ability to use Matlab in analysis and design of
nonlinear and adaptive control systems
26 August 2017 © H. T. Hoàng - HCMUT 28