Diagnosis And Compensation Of Multiple Chanel Faults In One-dimensional Systems And Two-dimensional Systems

With the increasing of complexity and scale in modern industry,numbers of faults occur in such cases,which threat the system performance and even safety.The safety is very important for system and has attracted much attention from the whole society.Fault diagnosis(FD)can detect the occurrence of fault,judge the location of fault,and estimate the size of fault effectively,and then it can provide efficient information for fault compensation.When faults occur,fault tolerant control(FTC)can guarantee acceptable performance of system,such as stability and reliability.Therefore,fault diagnosis and fault tolerant control have become the focus of academic studies and industrial applications,lots of theories and techniques have appeared and developed quickly.Two-dimensional systems exist in many kinds of industrial fields,whose theory can solve many problems efficiently and easily.In practice,two-dimensional systems face the threat of fault and also have the needs of reliability and safety.On the contrary,due to the complexity of two-dimensional systems,the results of fault diagnosis and fault tolerant control are rare,and many fundamental problems need to be solved.In this study,we investigate the multiple channel faults of one-dimensional systems and two-dimensional systems;the fault tolerant control of multiple intermittent channel faults in one-dimensional systems are considered;at the same time,the existence conditions and designs of observers for two-dimensional multiple channel faults are studied.The main results of this paper can be given as follows:For multiple channel intermittent faults in one-dimensional systems,we bring up a new mathematical model of intermittent fault,compared with the traditional Bernoulli model,the new model can describe the actuator fault and sensor fault simultaneously.In chapter 2,to deal with the fault tolerant control of intermittent faults,the systems subjected to intermittent faults are transformed as Markovian jump systems,then dynamic output feedback controllers are designed to achieve the passive fault tolerant control of intermittent faults.In chapter 3,time delays,parameter uncertainty in one-dimensional nonlinear systems are further considered,when the transition probabilities of Markov jump systems are partly unknown,the passive fault tolerant control of intermittent fault in such systems is investigated.For multiple channel faults in two-dimensional systems,we study the fault estimation and observer design.Two-dimensional time-invariant systems and two-dimensional time-variant systems are analyzed respectively.In chapter 4,for two kinds of two-dimensional system conditions,asymptotic stable observers and uniformly ultimately bounded observes are designed,the rank conditions and linear matrix inequality conditions about the existence of observers are given.Based on two-dimensional singular system theory and Lyapunov stability theory,the existent conditions and concrete procedures are presented to estimate system states and multiple channel faults.In chapter 5,with multiple faults and disturbance,we consider further the unbiased minimum variance estimation of two-dimensional time-variant systems and design relevant observers,based on Kalman filter theory in two-dimensional systems,the existent conditions of unbiased minimum variance estimator are also given.