Scribd
Upload
49 views10 pages

Applied Nonlunear Control: Fundamentals of Lyapunov Theory

This document discusses techniques for analyzing nonlinear control systems using Lyapunov's direct method. It begins by introducing basic assumptions of Lyapunov theory and noting that while Lyapunov functions can systematically describe stable linear systems, there is no general way to find them for nonlinear systems. The document then examines three methods: analyzing linear time-invariant systems using quadratic Lyapunov functions, Krasovskii's method which uses the Jacobian matrix, and the variable gradient method which searches for a gradient-based Lyapunov function. It also discusses using Lyapunov functions to analyze transient performance and convergence rates for linear systems.

Uploaded by

Jinmyung Kim
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
49 views10 pages

Applied Nonlunear Control: Fundamentals of Lyapunov Theory

This document discusses techniques for analyzing nonlinear control systems using Lyapunov's direct method. It begins by introducing basic assumptions of Lyapunov theory and noting that while Lyapunov functions can systematically describe stable linear systems, there is no general way to find them for nonlinear systems. The document then examines three methods: analyzing linear time-invariant systems using quadratic Lyapunov functions, Krasovskii's method which uses the Jacobian matrix, and the variable gradient method which searches for a gradient-based Lyapunov function. It also discusses using Lyapunov functions to analyze transient performance and convergence rates for linear systems.

Uploaded by

Jinmyung Kim
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 10

APPLIED NONLUNEAR

CONTROL
Chapter3
Fundamentals of Lyapunov Theory

2010 . 10 . 22
3.5 System Analysis Based on Lyapunov’s Direct Method

Basic assumption
 Lyapunov functions is somehow know?
 No general way of finding Lyapunov functions for nonlinear
systems.(use experience, intuition, physical insight)
 Lyapunov functions can be systematically found to describe
stable ‘linear system’.
 We discuss a number of techniques which can search of
Lyapunov functions.
 Transient performance analysis(3.5.5)
3.5 System Analysis Based on Lyapunov’s Direct Method

3.5.1 Lyapunov Analysis of Linear Time-Invariant Systems


  We should be able to describe linear systems in the
Lyapunov Formalism.

 Review some basic results on matrix algebra.


 Symmetric, Skew- Symmetric, Positive Definite Matrices
 Definition 3.10
 일단 정사각 nxn 행렬 M 을 가정하자 . Symmetric 란 이고 , Skew- Symmetric 이면
이다 . 즉 , 전자는 처럼 대각 대칭이고 후자는 와 같이 대각 대칭성분의 부호가 반
대이다 .
 뿐만 아니라 행렬 M 은 Symmetric 와 Skew- Symmetric 로 표현이 가능하다 .
 , 유일한 표현 방법이다 .
3.5 System Analysis Based on Lyapunov’s Direct Method

3.5.1 Lyapunov Analysis of Linear Time-Invariant Systems



 Quadratic
  function associate with a Skew- Symmetric matrix is always 0
 임의의 nx1 행렬 x 와 Skew- Symmetric 행렬 M 의 경우 이며 scalar 값을 갖는다 .
(1xn nxn nx1 = 1x1) 이를 통해 (3.16) 이 성립한다 .
 즉 , 행렬이 Skew- Symmetric 이기 위한 필요충분 조건이 (3.16) 인 셈이다 .
 Basis vector 로 증명

 Definition 3.11
 라면 M 은 Positive Definite(p.d.) 이다 .
 즉 , 가 p.d. 함수라면 M 또한 p.d. 이다 .
3.5 System Analysis Based on Lyapunov’s Direct Method

3.5.1 Lyapunov Analysis of Linear Time-Invariant Systems


3.5 System Analysis Based on Lyapunov’s Direct Method

3.5.1 Lyapunov Analysis of Linear Time-Invariant Systems


 Lyapunov

  Functions For Linear Time Invariant Systems
 Linear system
 consider a quadratic Lyapunov function candidate

 P, Q satisfies the conditions of the Lyapunov Functions

 Example 3.17
 Q 가 p.d. 조건만으로는 시스템이 안정한지 장담할 수 없다 . 왜 ? P 때문에 !

1. p.d. Q 를 선택하고
2. Lyapunov Functions 에 대해서 P 를 구하고
3. 구한 P 가 p.d. 인지 확인한다 .
 Teorem 3.6
 LTI system is strictly stable any symmetric p.d. Q uniqu matrix P solution of the Lyapunove equation(3.19) be
symmetric positive definite
 Example 3.18
 Q 를 정하고 (=I) 만족하는 P 를 찾는다 .
3.5 System Analysis Based on Lyapunov’s Direct Method

3.5.2 Krasovskii’s Method



 Theorem3.7(Krasovskii)
 
 Definition3.1, with equilibrium point of interest being the origin. Let denote the Jacobian matrix of the
system
 If the matrix is negative definite in a neighborhood , then equilibrium point at the origin is asymptotically
stable. A Lyapunov function for the system is
 If is the entire state space and, in addition, , then the equilibrium point is globally asymptotically
stable.
 어떤 입력 x 에 대한 출력 y 의 관계는 함수 f 의 관계를 가진다고 하자 . 특정 입력 x1 에 대한 출력이 y1 이라
고 할 때 , 이 x1 지점에 대한 함수 f 의 미분값으로 f 가 선형일 때 x2 에 대한 출력 y2 를 계산할 수 있다 . 이
러한 특성을 이용한 것이 자코비안 (Jacobian) 이며 , 이를 행렬로 표현한 것이 자코비안 행렬 (Jacobain Matrix) 이
다.
 Jacobian 은 함수의 변화율을 나타내기 때문에 비선형 시스템과 같이 특정 구간마다 함수의 변화율이 다른 시스
템에서는 사용하기 어렵지만 함수의 어느 부분을 선형화 (Linearization) 해서 시스템을 근사화 하고자 할 때 많이
사용된다 . 이럴 경우에는 함수의 변화율 자체가 작아야 그 근사값이 신뢰도가 높아진다 .

 Example 3.19
3.5 System Analysis Based on Lyapunov’s Direct Method

3.5.3 The Variable Gradient Method



   함수를 시행착오에 의존하지 않고 작성하는 방법 중 한가지
V
 즉 , 시스템이 주어진 상태 공간 내에서 안정하다고 하면 그 시스템에 대해서 V 함수가 존재하고 V 가
존재하면 그 그레디언트 (Gradient) 도 또한 존재한다 . 따라서 이 그레디언트가 주어지면 와 를 모두 계
산할 수 있지 않을까 ?
3.5 System Analysis Based on Lyapunov’s Direct Method

3.5.5 Performance Analysis



 A
  Simple Convergence Lemma
 If a real function satisfies the inequality

Where is a real number. Then

 Estimating Convergence Rates For Linear Systems


You might also like