Unknown Input Observer-Based Robust Sensor Fault Estimation in Discrete-Time Takagi-Sugeno Systems

Abstract

In this paper, a robust fault estimation technique is proposed to estimate the states and sensor faults simultaneously in a discrete-time Takagi-Sugeno (TS) fuzzy system with uncertainties. Initially, an augmented system is obtained from fault dynamics model and from the original system. Then an robust TS fuzzy unknown input observer (UIO) is designed for the augmented system with the aim of performing decoupling of the process disturbances. The UIO design is achieved by an $H_{\infty}$ technique, which is developed to maintain the UIO stable and to reduce the disturbances that cannot be decoupled. The proposed method is validated through a simulation example.