Aircraft Engine Sensor Fault Detection, Isolaton and Reconfiguration Using Fuzzy Logic Control Based Unknown Input Observer

Abstract

In this paper, a model-based robust fault diagnosis approach 'Unknown Input Observer (UIO)', 'Generalized Observer Scheme (GOS)' and 'a Fuzzy Logic Control Mechanism' will be studied respectively for fault detection, isolation (FDI) and reconfiguration of sensor on aircraft's engine. UIO is defined as the linear observers which are completely independent of the immeasurable disturbances. The aim of this method to make the state estimation error decoupled from the unknown inputs (disturbances). UIO's reduce or almost eliminate the effect of unknown disturbances in a system. Generalized Observer Scheme is defined as the observers which uses all residuals and all outputs except one output. It describes the location of the fault. In our study, Fuzzy Logic Control Mechanism controls the system due to the rules that depend on the residual information of each sensor.