Fault-tolerant control based on disturbance observer for stochastic systems with multiple disturbances

The problem about anti-disturbance and fault-diagnosis is considered for a class of stochastic systems with multiple disturbances and fault. The multiple disturbances include the disturbance with partially-known information generated by exogenous systems and the white noise. The disturbance generated by exogenous systems can not only represent the disturbance with partially-known information, but also describe a stochastic disturbance. The coupling leads to the invalidity of certainty equivalence principle. To solve the difficulty, the composite pole placement and LMI methods are proposed. Firstly, a stochastic disturbance observer is constructed to estimate the disturbance with partially-known information. Then, a stochastic fault diagnosis observer is constructed to estimate the fault, based on which, a composite fault-tolerant control scheme is proposed by combining fault-tolerant control and stochastic control. It can be guaranteed that all the signals in the composite system are asymptotically bounded in mean square under certain conditions. Finally, a simulation example is given to illustrate the correctness and effectiveness of the proposed method. © 2019, Editorial Office of Control and Decision. All right reserved.