Augmented System Observer-based Robust Fault-estimation Scheme for Nonlinear Systems

Abstract

For the problem of previous fault-estimation studies that the so-called observer-matching condition should be satisfied, as well as the difficulty to estimate actuator and sensor faults simultaneously, an augmented system fault-estimation observer is proposed for a sort of nonlinear systems subject to process and sensor disturbances. The essential idea is to construct an augmented system using a filter and to further create an augmented state vector by considering the system states and simultaneous actuator and sensor faults. The nonlinearity, process, and sensor disturbances and multiple-fault problem of the actuator and sensor are all taken into account. Then, an augmented observer is designed for the augmented system by attenuating the disturbances and simultaneously estimating the system states and actuator and sensor faults. The asymptotic stability is confirmed via the Lyapunov stability theory, and the linear matrix inequality optimization technique is further utilized to solve the designed observer. Numerical simulations of a single-link flexible-joint robot are performed to illustrate the effectiveness of the proposed robust fault-estimation scheme.