Asynchronous Controller Design And Fault Detection For Nonlinear Markov Jump Systems

In this thesis,the asynchronous control and the asynchronous fault detection for a class of nonlinear Markov jump systems are investigated thoroughly.As a kind of hybrid system,Markov jump systems have displayed strong modeling ability in describing systems with sudden changes in structure and parameters.The abrupt changes of structure and parameters are often caused by the interference of external noise,sudden failure of working devices,and maintenance.In terms of a mathematical expression,the continuous Markov jump systems are usually represented by a differential equation,and the discrete Markov jump systems are characterized by a difference equation.The jump between subsystems follows a certain transition rate(for continuous-time Markov jump systems)or a transition probability(for discrete-time Markov jump systems).Most of the published results about the controller and filter design for Markov jump systems are based on the assumption that we can get the jump mode information from the original system timely and accurately,however,due to time delay,external noise,data packet loss,and temporal chaos,the system's mode often cannot be obtained accurately or even be unavailable,which results in a complex asynchronous phenomenon between the system and controller/filter.Based on this,this thesis introduces a hidden Markov model to detect the original system's mode,and further applies it to the controller and filter design of Markov jump systems.On the other hand,the rapid development of science and technology makes the social industrial level improve increasingly,and then makes the structure of the electromechanical system more and more complex.Although a higher level of automation can improve productivity,it may increase the probability of system failure.Minor failures can cause equipment to shut down,while major failures can have catastrophic consequences.Therefore,the world is in urgent need of technology to enhance the safety of the operation of the modern equipment system,and the emergence of fault detection technology provides an effective approach to improve the availability and viability of the system.Resorting to Lyapunov function,T-S fuzzy,finite-time,event-triggered,linear matrix inequality and other related knowledge,the asynchronous controller design and fault detection for a class of nonlinear Markov jump systems have been studied.Besides,the research objects in this thesis all have practical application backgrounds,including DC motor model,wind turbine model and wheeled mobile manipulator with hybrid joints.The specific research contents are as follows:(1)For a class of discrete-time Markov jump systems with Conic-type nonlinearity,an observer-based finite-time asynchronous controller is designed to make the resultant closed-loop Markov jump systems satisfy an H? disturbance rejection performance.(2)For a class of continuous-time Markov jump systems with Lur'e-type nonlinearity,a finite-time asynchronous output feedback controller is designed to make the resultant closed-loop Markov jump systems satisfy a strict(U,G,V)-?-dissipative performance.(3)For a class of discrete-time Markov jump systems with Conic-type nonlinearity,an asynchronous fault detection observer is designed.A multi-objective algorithm for asynchronous fault detection is proposed based on the H? performance and the Hperformance indexes.(4)For a class of nonlinear Markov jump systems,an asynchronous fault detection filter is designed by using T-S fuzzy technology and event-triggered strategy when the mode detection probability is partially unknown.Finally,this thesis gives a general summary of the asynchronous control and asynchronous fault detection methods and points out the problems to be solved and the specific research direction in the future.