Fault Diagnosis Of Engine Control System Sensors And Implementing Agencies

Fault detection and isolation logic plays a crucial roal in enchancing the safety and reliability of the aircraft propulsion systems. However, achieving the FDI task with high reliability is a challenging problem. For this purpose, the approaches to detect and isolate sensor/actuator fault of aircraft engine is proposed.Firstly, based on the response of the small perturbing of the nonlinear aeroengine model, the fitting method is adopted to get the linear state variable model. Then, A bank of Kalman filters are used to detect and isolate sensor fault, each of Kalman filter is designed based on a specific hypothesis for detecting a specific sensor fault. In order to save the time of the detection, sequential probability ratio test is investigated to deal with residual signals. Thereafter, two different fault detection algorithms- SPRT (Sequential Probability Ratio Test) & WSSR (Weighted Sum of Squared Residual)-that analyze the sensor residual signals are compared. The simulation results show that SPRT method can be used to detect aircraft engine sensor soft failures fast. Lastly, the Kalman filter that satisfies the Doyle-Stein condition is refered to as Robust Kalman filter (RKF). The use of the RKF is very useful in the isolation of sensor and control surface failures as it is insensitive to the latter failures.A model-based approach utilizing a bank of Kalman filters and a Robust Kalman filter is investigated for aircraft engine sensor and actuator FDI faults. The simulation results show that a bank of Kalman of filters can detect and isolate sensor fault fast and accurately, and RKF can successful isolate sensor or actuator fault.