Fault Detection and Isolation for Nonlinear System via ESO

Abstract

In this paper, a fault detection and isolation system for nonlinear systems is presented. Fault detection and isolation is accomplished by using extended state observer (ESO) and fuzzy logic system. The major of observer-based fault detection methods rely on the accurate mathematical model of the system, but in the real world an accurate model of the system may not be available. The ESO is different from conventional observers, it does not require an accurate model of the system, it provides vital information for fault detection with only partial information of the plant. The ESO has ability to augment unknown dynamics combined with unknown external disturbance as extended state and estimate it in real time by using given input-output data. This paper presents a new sensor fault detection and isolation (FDI) via ESO and fuzzy logic system. A two-tank system is used as a case study. The simulation results confirm the simplicity and effectiveness of the proposed FDI technique.