Fuzzy observer for fault detection and reconstruction of unknown input fuzzy models

Abstract

This paper addresses the design of a fuzzy observer for unknown input Takagi-Sugeno (T?S) fuzzy model. The main contribution of this paper is the development of a robust fuzzy observer in the presence of disturbances. Based on Lyapunov function, it is shown how to determine the observers gains in Linear Matrix Inequalities (LMI) terms. The proposed observer structure allows to estimate simultaneously and systematically the unknown inputs and state variables. The designed T?S observer is used for detection and reconstruction of faults which can affect a non-linear model and can be applied directly for fault detection and isolation of actuator faults. The validity of the proposed methodology is illustrated by estimating the state and faults of an automatic steering vehicle.