Fault detection and isolation problem: Sliding mode fuzzy observers and neural networks

Abstract

In this paper results of the application of a hybrid Fault Detection and Isolation scheme are presented. A Takagi-Sugeno fuzzy model is used to describe the system and a type of sliding mode observers are designed to estimate the system state vector; from this, the diagnostic signal-residual is generated by the comparison of measured and estimated output. Neural Networks are proposed in order to solve the fault isolation problem based on signal-residual. The faulted component is identified from the active signal-residuals by means of the application of the presented technique based on neural networks. This paper shows an application of the fault diagnosis technique, which was satisfactorily tested in a two-tank hydraulic system.