Sensor fault detection algorithm of flight control system under modeling uncertainty and noise disturbance

摘要

Sensor fault detection method based on UIF (Unknown Input Filter) and minimum variance is proposed to address the real-time fault detection problem of the critical flight sensor and precise positioning the sensor fault in this literature. For aircraft flight control systems that include modeling uncertainties, the UIF method is given and the greatest decoupling between the residuals and the modeling uncertainties is achieved. Through the gain matrix optimizing, the minimum variance estimation error is obtained. The detection function and residual are generated to achieve sensor fault detection and position the fault location. Simulation example of a certain aircraft with system modeling uncertainty and disturbance, the sensor fault detection is verified. For multiple sensor fault, the detection time is determined by senor fault and for only pitot tube fault, the detection time is about 2.5s, in addition, when only angular rate gyro fault, the fault detection is almost one sample time.