Robust Fault Diagnosis for Discrete-Time Switched System with Unknown State Delays Subject to Component Faults

Abstract

This paper investigates the problem of fault detection and diagnosis in a class of discrete-time switched systems with unknown state delays against component faults. First, a polytopic learning observers(LOs) is proposed to reconstruct both the system state and the loss of actuator effectiveness. The proposed LOs technique is also extended to the system with unknown time delays and external disturbance. Then the sufficient conditions for the convergence of LOs and the uniformly ultimately boundedness of the switched system are proved using the switched Lyapunov function approach and H_inf  techniques. Finally, three kinds of fault signals are provided to verify the effectiveness of the developed techniques. The simulation results show that the reconstructed error of the fault signal is maintained within 1.9% after the convergence process, while the implicated error of the reconstructed signal for normal components can be limited to less than 2% in the most serious case.