Adaptive Control For Switched Systems Based On Closed-loop Reference Model

As an important class of hybrid systems,switched systems are composed of a family of subsystems and a logical rule that decides the switching mechanism between the various subsystems.Switched systems are widely used in practical engineering control field in recent years such as complex network communication control system,intelligent control system for traffic control,automatic control system used in industrial process.There are many control systems with uncertainties in real world and the traditional control methods cannot guarantee desired performance because of the existence of uncertainties.Thus,studying the stability of switched systems with uncertainties is of great theoretical value and practical significance.This thesis studies the model reference adaptive control problem of a class of switched system with uncertainties based on common Lyapunov function method and average dwell time method,and different switched controllers and adaptive laws are designed respectively in the case of measurable states and immeasurable states situations.The main contents of this thesis are as follows:(1)Based on the common Lyapunov function method and average dwell time method,the stability of the switched error system with parameter uncertainty is studied,and sufficient conditions and theorems for the existence of the corresponding stability are given respectively in the case of measurable states and immeasurable states situations.The sufficient conditions for the asymptotic stability of the switching error system are given in the common Lyapunov function method.In the average dwell time method,because of the uncertainty of parameters in the controlled system,the switched error system ultimately has global practical stability rather than the asymptotic stability,which is that the state error is not converging to 0 but converges to a ball,the practical stability conditions of the same switching error system are also given in terms of the state of measurable states and immeasurable states situations.(2)Compared with the classical open-loop reference model,the reference model of this thesis is a kind of closed-loop reference model,which has a feedback item based on output error.This increases the design margin of the system,and can improve the dynamic performance of the whole switched system by adjusting the feedback coefficient.The state of the controlled system in the practical control systems is difficult to observe so different switched controllers and adaptive laws are designed when the state of the controlled system is measured or unmeasured.When the state of the controlled system is not measured,the state of the reference model is used instead of the state of the controlled system to design controllers and adaptive laws in this thesis as not the same as the direct method and indirect method.Each reference model and controller is individually designed not in common which adds the flexibility of the design.(3)For the main theorems presented in this thesis,numerical simulation examples are given to verify the correctness and validity of the corresponding conclusions by analyzing and comparing the simulation data and images.