State Estimation And Fault Detection For A Class Of Uncertain Switched Systems

Switched systems are distinctive subclass of hybrid systems composed of a family of continuous-time or discrete-time subsystems with a criterion that rules the switching among them.Since many physical processes switched nature,and many real-word applications resort to switching strategy to improve the control performance,switched systems have received a great attention during the recent decades.This paper mainly discusses the issues of asymptotic and finite time state estimation for a switched linear system with unknown inputs where the active modes of the switched system are unknown.Two different methods which estimate the state and identify the active mode for switched systems are compared to show the advantage of finite time observer.After the state estimation of the switched system,the fault detection is also carried out to ensure that the stability of the system can be improved immediately when the fault occurs.First,both state and output transformations are used to deal with unknown inputs,and a reduced-order switched system that removes the influence of unknown inputs is constructed for the state estimation.Second,for each reduced-order subsystem,the corresponding asymptotic and finite time observers are designed and two different active mode identification methods are proposed by minimizing the output residual norm.Then,based on the asymptotic observer,the state estimation of a reduced-order decoupled system can be obtained and based on the finite time observer,the state estimation of a reduced-order decoupled system can be obtained by choosing any small time-delay parameters in finite time.Finally,a MATLAB simulation example is given to verify the reliability by two different methods.For the fault detection of the switched system,first,we assume the fault occurs and then we can use the fault detector to detect it.Finally,the feasibility and effectiveness of the fault detector are verified by MATLAB simulation.At the end of the article,this paper summarizes the work that has been completed and looks forward to the next work.