Fault Detection Observer Design for Nonlinear Discrete-Time Systems with Measurement Outliers

Abstract

This paper focuses on the fault detection (FD) observer design issue for a category of nonlinear discrete-time Takagi-Sugeno (T-S) fuzzy systems, in which measurement outliers are considered. A new FD observer is constructed to decrease the affect on anomalous measurement signals which exist in the considered system based on the scheme of saturation function. Then, sufficient conditions are provided to ensure the error system is stable and satisfies performance requirements. Measurement outliers may exist in the system and produce the false alarm, contrasted to the present works, the observer proposed in the article can avert the false alarm. At last, an example is provided to identify the superiority of the technique proposed in this paper.