Fault Diagnosis by an Observer-Based Fuzzy Decision System

This paper deals with the problem of nonlinear system fault detection by using extended state observer (ESO) and fuzzy logic. Most of observer-based fault diagnosis methods require a nominal mathematical model of the system. However, the extended state observer can provide vital information for fault detection with only partial information of the plant, which can not be easily accomplished with any existing observer-based methods. In this paper, the ESO can be applied to generate signatures, called residuals, which are indicative of the presence of faults in the system, and then, the generation of residuals can be evaluated by fuzzy decision system to make fault diagnosis decisions. Another advantage of this observer is its simplicity in tuning only a single parameter, which is a trade-off between the tracking speed and the sensitivity to measurement noise. A strongly coupled three-tank nonlinear dynamic system is chosen as a case study. The simulation results indicate that the presented ESO-based fuzzy decision system effectively detected faults for nonlinear system.