Research On Fault Estimation Of Time-Delay Systems Based On T-S Fuzzy Model

Due to the large scale and high complexity of modern control systems,people have put forward stricter requirements on the safety and reliability of control systems,and fault-tolerant control can ensure the system's requirements for safety and reliability,so fault-tolerant control has become a research hot spot.At the same time,as the basis of fault-tolerant control,fault diagnosis technology has also attracted much attention.Fault detection,fault isolation and fault estimation in the field of fault diagnosis have become hot research topics.Especially fault estimation,it has dual functions of detection and isolation.This is also an important reason for studying the problem of fault estimation in this thesis.This thesis mainly studies the fault estimation problem of Takagi-Sugeno(T-S)fuzzy model control system affected by time delay and external disturbance.By constructing fuzzy learning observers and adaptive observers and using optimization techniques,accurate estimation of system states and faults is achieved.The main work can be summarized as follows:1.For a class of T-S fuzzy systems with continuous time-varying delays,actuator failures and external disturbances constrained by norms,a robust fault estimation problem is investigated in this work.First,an improved fuzzy learning observer(Fuzzy Learning Observer,FLO)is constructed,and then the H? optimization technique is applied to achieve simultaneous estimation of system state and actuator failure.Based on the Lyapunov method,some sufficient and sufficient conditions for the existence of FLO are given.These sufficient conditions are given by a set of linear matrix inequalities.The fault estimation obtained by FLO can well describe the size and shape of faults.Finally,the simulation results of the dynamic model show the effectiveness of the proposed method.2.The problem of fault estimation for T-S fuzzy generalized time-delay systems affected by actuator failure,sensor failure and external disturbances is studied.An adaptive fuzzy observer is given to realize the simultaneous estimation of system state,actuator failure and sensor failure.Based on Lyapunov's method,system stability analysis and design conditions are transformed into a set of linear matrix inequality.Based on this,a set of linear matrix inequality is solved to calculate the observer gain,and it is used to estimate the system state,actuator failure and sensor failure.Finally,an example is given to verify the effectiveness of the proposed method.