Tài liệu Hệ thống điều khiển mờ - Thiết kế và phân tích P12 docx

Wang Copyright 䊚 2001 John Wiley & Sons, Inc.ISBNs: 0-471-32324-1 Hardback ; 0-471-22459-6 Electronic CHAPTER 12 NEW STABILITY CONDITIONS AND DYNAMIC FEEDBACK DESIGNS This chapter presen

Kazuo Tanaka, Hua O Wang Copyright 䊚 2001 John Wiley & Sons, Inc.

ISBNs: 0-471-32324-1 Hardback ; 0-471-22459-6 Electronic

CHAPTER 12

NEW STABILITY CONDITIONS AND

DYNAMIC FEEDBACK DESIGNS

This chapter presents a unified systematic framework of control synthesis

w1᎐5 for dynamic systems described by the Takagi-Sugeno fuzzy model Inxcomparison with preceding chapters, this chapter provides two significantextensions First we provide a new sufficient condition for the existence of aquadratically stabilizing state feedback PDC controller which is more generaland relaxed than the existing conditions Second, we introduce the notion of

dynamic parallel distributed compensation DPDC and we provide a set ofsufficient LMI conditions for the existence of quadratically stabilizing dy-namic compensators

In this chapter, the notation M) 0 stands for a positive definite

Ž

u t , respectively Another notable point regarding the notation is that we

will use p t or p instead of z t as premise variables This is because z is

used as performance variables in Chapters 13 and 15 which are based on thesetting presented in this chapter The symbol xX denotes the transposedvector of x We often drop the p and just write h , but it should be kept in i

mind that the h ’s are functions of the variable p i

The summation process associated with the center of gravity

parameter p is a measurable external disturbance signal which does not

p s f x Using this interpretation, equations 2.3 and 2.4 describe a

nonlinear system As a slight modification to this interpretation, we canassume that the parameter p is a function of the measurable outputs of the

p using a combination of these approaches.

12.1 QUADRATIC STABILIZABILITY USING

Let us consider the Lyapunov function candidate V x s x Px, where

P) 0 Taking the time derivative of this function along the flow of

given in 17 and 20 It is also weaker than the LMI condition given in 25 ,

in which caseT i j becomes t I The above theorem can be further relaxed if i j

we know the structure of the fuzzy membership function:

䢇 Sometimes there is no overlap between two rules, that is, the product ofthe h and the h may be identically zero In this case, the above i j

theorem can be relaxed by dropping the condition 12.7 corresponding

to the i and j in 12.7

䢇 If only s - r rules can fire at the same time, then the conditions of this

theorem can be further relaxed to only require that all the diagonal

s = s principal submatrices of T are negative definite.

12.2 DYNAMIC FEEDBACK CONTROLLERS

In this section we introduce the concept of a DPDC, and we derive a set ofLMI conditions which can be used to design a stabilizing DPDC

In order to derive the LMI design conditions, it is useful to begin with aparameter-dependent linear model described by the equations

A parameter-dependent dynamic compensator is a parameter-dependentlinear system of the form

x t s AŽ Žp x t q B Ž Žp y t , Ž

u t s CŽ cŽp x t q D. cŽ cŽp y t Ž Ž12.12.

Defining the augmented system matrix

A p q B p DŽ Ž cŽp C p Ž B p CŽ cŽp.

A c lŽp s.

B cŽp C p Ž A cŽp.and the augmented state vector

T

x c lŽ t s x Ž t x cŽ t ,the resulting closed-loop dynamic equations are described by the equation

x Ž t s A Žp x Ž t Ž12.13.

˙c l c l c l

The system 12.11 is said to be quadratically stabilizable via an

s-dimen-sional parameter-dependent linear compensator if and only if there exists an

s-dimensional parameter-dependent controller and a positive definite matrix

P c l) 0 such that

P A c l c lŽp q A. T c lŽp P. c l- 0 Ž12.14.

Remark 39 If we fix the value of p, equation 12.14 represents a sufficient

condition for the existence of a set of linear, time-invariant controller

We will first partition the constant matrices P and Py 1 into components:

I P11

⌸ s P ⌸ s2 c l 1 T

0 P12

Ž

Equation 12.14 will hold if and only if

⌸1T P A c l c lŽp.⌸ q ⌸1 T1A T c lŽp P. c l⌸ - 0.1This equation can also be rewritten as

⌸T2A c lŽp.⌸ q ⌸1 T1A T c lŽp.⌸ - 0.2Writing out the first term on the left-hand side of this equation, we have

A p Q q B pŽ 11 Ž C Žp. A p q B pŽ Ž D Žp C p Ž

A Žp. P A p q11 Ž B Žp C p Ž and the closed-loop stability condition can be expressed as

E p q EŽ TŽp.- 0or

This last condition holds if and only if

We will now assume that the parameter-dependent plant can be described

by a fuzzy T-S model using r model rules In this case, the

parameter-depen-dent plant can be described by the equation

D Žp s Ýh iŽp Di

is1 r

i

J Ýh iŽp D . c

is1

Ž The matrix E p then becomes

lizable ®ia a DPDC controller 12.25 ᎐ 12.28 if the LMI conditions 12.16 and

Ž12.35 are feasible with LMI ®ariables Q , P ,11 11 Ai jk, Bi j, Cjk, and Dj.The controller is gi®en by

Ž

Reduction of LMI Equations 12.34 can be simplified by permuting dices To this end, we first note that the controller equations can be rewrittenas

1

21

Ž Our aim is to show that the stability conditions 12.34 can also be written

E p q EŽ TŽp.- 0or

globally quadratically stabilizable ®ia a DPDC controller 12.44 ᎐ 12.47 if the

following LMIs are feasible with LMI ®ariables Q , P ,11 11 Ai, Ai j, Ai jk, Bi,

B , C, C , and D:

B cŽp s Ýh iŽp B ,. c Ž12.62.

is1 r

Ž The matrix E p then becomes

T

E p q E pŽ Ž - 0or

i

B cŽp s Ýh iŽp B ,. c Ž12.74.

is1 r

i

C cŽp s Ýh iŽp C ,. c Ž12.75.

is1

LMI conditions 12.16 , 12.85 ,and 12.86 are feasible with LMI ®ariables Q ,11

P , T ,11 i j Ai j, Bi, Ci, and D.The controller is gi®en by

Linear Parameterization: Common B Assume that B s B s1 2 ⭈⭈⭈ s

Dynamic Part: Rule i

i

u s Ýh iŽp C x q D y.. c c c Ž12.96.

is1

The controller parameters are

C, that is, C s C s1 2 ⭈⭈⭈ s C s C in system 2.3 and 2.4 can be handled r

analogous to the Common B case Consider

As in the previous case, a commonC matrix case can always be obtained

by augmenting the outputs of the system with integrators and using theaugmented states as a new set of outputs

In this case, a linear parameterization dynamic controller takes the ing form:

follow-Dynamic Part: Rule i

The controller can be written as

where P , P , Q , and Q11 12 11 12 satisfy the constraint P Q q P Q11 11 12 T12sI.

Linear Parameterization: Common B and Common C Consider the case

In this case, a linear parameterization dynamic controller takes the ing form:

follow-Dynamic Part: Rule i

Fig 12.1 The ball and beam system.

12.3 EXAMPLE

In this section, we consider a ball-and-beam system which is commonly used

as an illustrative application of various control schemes The system is shown

in Figure 12.1 To begin with, we represent the original model exactly using aT-S model via sector nonlinearity

The beam is made to rotate in a vertical plane by applying a torque at thecenter of rotation and the ball is free to roll along the beam Assume no

00

g x sŽ

01

There are two nonlinearities in 12.130 , the x x1 4 term and the sinx3

term As we know, most nonlinearity can be bounded by sector In this

example, assume x g y3 ␲r2 ␲r2 and x x g yd d This is the re-1 4

gion that we assume the system will operate within It follows that

1 y sinŽx rx3. 3

M12Žx3.s

1 y 2r␲and

Since the ball-and-beam system is a common B and common C case as

discussed in Section 12.2.3.3, we will apply DPDC with linear tion for the system The simulation result is shown in Figure 12.2 The systemparameters for simulation are chosen as B s 0.7143, G s 9.81, d s 5, and

the initial condition is 1, 0, 0.0564, 0

Fig 12.2 Response of Ball and Beam using DPDC with linear parameterization.

Remark 40 From the simulation results, we know that x and x x do not3 1 4

exceed the bound limit assumed in the modeling A more systematic proach is to incorporate the constraints as performance specifications in thecontroller design This issue is addressed in the next chapter

ap-BIBLIOGRAPHY

1 J Li, D Niemann, H O Wang, and K Tanaka, ‘‘Multiobjective Dynamic Feedback Control of Takagi-Sugeno Model via LMIs’’ Proc 4th Joint Conference

of Information Sciences, Durham, Vol 1, Oct 1998, pp 159᎐162.

2 J Li, D Niemann, H O Wang, and K Tanaka, ‘‘Parallel Distributed tion for Takagi-Sugeno Fuzzy Models: Multiobjective Controller Design,’’

Compensa-Proc 1999 American Control Conference, San Diego, June 1999, pp.1832᎐1836.

3 D Niemann, J Li, H O Wang, and K Tanaka, ‘‘Parallel Distributed tion for Takagi-Sugeno Fuzzy Models: New Stability Conditions and Dynamic Feedback Designs,’’ Proc 1999 International Federation of Automatic Control

Compensa-ŽIFAC World Congress, Beijing, July 1999, pp 207 ᎐212.

4 J Li, H O Wang, D Niemann, and K Tanaka, ‘‘Synthesis of Gain-Scheduled Controller for a Class of LPV Systems,’’Proc 38th IEEE Conference on Decision and Control, Phoenix, Dec 1999, pp 2314᎐2319.

5 J Li, H O Wang, D Niemann, and K Tanaka, ‘‘Dynamic Parallel Distributed Compensation for Takagi-Sugeno Fuzzy Systems: An LMI Approach,’’ Inform.

Sci., Vol 123, pp 201᎐221 2000

6 P Apkarian, P Gahinet, and G Becker, ‘‘ Self-Scheduled H Control of Linear⬁

Parameter Varying Systems: A Design Example,’’ Automatica, Vol 31, No 9,

Control,’’Int J Robust Nonlinear Control, Vol 4, No 4, pp 421᎐428 1994

11 M Johansson and A Rantzer, ‘‘On the Computation of Piecewise Quadratic Lyapunov Function,’’ in Proc 36th CDC, 1997, pp 3515᎐3520.

12 R Langari and M Tomizuka, ‘‘Analysis and Synthesis of Fuzzy Linguistic Control Systems,’’ in Proc 1990 ASME Winter Annual Meet., 1990, pp 35᎐42.

13 J Li, H O Wang, and K Tanaka, ‘‘ Stable Fuzzy Control of the Benchmark Nonlinear Control Problem: A System-Theoretic Approach,’’ in Joint Conf of Information Science, 1997, pp 263᎐266.

14 C Scherer, P Gahinet, and M Chilali, ‘‘Multiobjective Output-Feedback Control via LMI Optimization,’’ IEEE Trans Automatic Control, Vol 42, No 7, pp.

896 ᎐911 1997

15 J Shamma and M Athans, ‘‘Analysis of Nonlinear Gain Scheduled Control

Systems,’’IEEE Trans Automatic Control, Vol 35, pp 898᎐907 1990

16 T Takagi and M Sugeno, ‘‘Fuzzy Identification of Systems and Its Applications

to Modeling and Control,’’ IEEE Trans Syst Man and Cybernet., Vol 15, pp.

Systems,’’Fuzzy Sets Syst., Vol 45, No 2, pp 135᎐156 1992

19 H O Wang, K Tanaka, and M F Griffin,‘‘ Parallel Distributed Compensation of Nonlinear Systems by Takagi-Sugeno Fuzzy Model,’’ in Proc FUZZ-IEEErIFES

24 S Boyd, L E Ghaoui, E Feron, and V Balakrishnan, Linear Matrix Inequalities

in Systems and Control Theory, SIAM, Philadelphia, PA, 1994.

25 P Gahinet, A Nemirovski, A J Laub, and M Chilali, LMI Control Toolbox,

Math Works, 1995.

26 L X Wang, Adapti®e Fuzzy Systems and Control: Design and Stability Analysis,

Prentice-Hall, Englewood Cliffs, NJ, 1993.