Robust Fault and Disturbance Reconstruction for Continuous-Time Markovian Jump Systems

This paper presents the robust actuator fault and disturbance reconstruction scheme for a kind of continuous-time Markovian jump systems (MJSs). A novel observer with the output derivative signal is proposed to realize the estimation for system states, actuator faults and disturbances simultaneously. By using the disturbance decoupled method, the influence of disturbances is completely eliminated from estimation error dynamics, which can facilitate the stability analysis and improve estimation accuracy. Moreover, the proposed method does not impose known boundedness assumptions on disturbances, thereby enlarging its applicability compared with the existing schemes. Based on the Lyapunov stability theory and the robust H∞ method, sufficient conditions ensuring robust stochastic stability of the resulting error systems including system states, actuator faults and disturbances are established in the form of strict linear matrix inequalities (LMIs), by which observer gains are also solved. Finally, the simulation study on an electronic circuit model is presented to demonstrate the effectiveness of the proposed multivariate estimation approach.