Many existing aircraft engine diagnostic methods are based on linearized engine models. However, the dynamics of aircraft engines are highly nonlinear and rapidly changing. Future engine health management designs will benefit from new methods that are directly based on intrinsic nonlinearities of the engine dynamics. In this paper, a fault detection and isolation (FDI) method is developed for aircraft engines by utilizing nonlinear adaptive estimation and nonlinear observer techniques. Engine sensor faults, actuator faults and component faults are considered under one unified nonlinear framework. The fault diagnosis architecture consists of a fault detection estimator and a bank of nonlinear fault isolation estimators. The fault detection estimator is used for detecting the occurrence of a fault, while the bank of fault isolation estimators is employed to determine the particular fault type or location after fault detection. Each isolation estimator is designed based on the functional structure of a particular fault type under consideration. Specifically, adaptive estimation techniques are used for designing the isolation estimators for engine component faults and actuator faults, while nonlinear observer techniques are used for designing the isolation estimators for sensor faults. The FDI architecture has been integrated with the Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) engine model developed by NASA researchers in recent years. The engine model is a realistic representation of the nonlinear aero thermal dynamics of a 90,000-pound thrust class turbofan engine with high-bypass ratio and a two-spool configuration. Representative simulation results and comparative studies are shown to verify the effectiveness of the nonlinear FDI method.
Fault Detection in Inverter-Based Microgrids Utilizing a Nonlinear Observer
