Actuator fault detection and performance recovery with Kalman filter-based adaptive observer

Abstract

A novel Kalman filter-based adaptive observer for the sampled-data nonlinear time-varying system is proposed in this paper. With the high gain property of Kalman filter, it is applicable to a large variation of unknown parameters, which can be estimated optimally. Then a method of actuator fault detection is proposed. With the estimated faults, one can use the proposed input compensation method to solve actuator faults. Additionally, the optimal linearization technique is used to obtain the locally optimal linear model for a nonlinear system at each sampled state, so that the actuator fault detection and performance recovery of a sampled-data nonlinear time-varying system is accomplished. In this paper, we also introduce a prediction-based digital redesign method to develop the corresponding sampled-data controller.