Research on Real-Time Fault Detection of UAV Flight Control System Sensor Based on Riemannian Manifold

Aiming at the difficulty of real-time fault diagnosis of sensors in flight control system of unmanned aerial vehicle (UAV), a fault diagnosis method based on Pearson correlation coefficient and Riemannian manifold is proposed. Firstly, a correlation coefficient matrix can be calculated from the data of a sensor in flight control system to quantitatively represent the correlation characteristics of each channel data in a sensor. Then, the similarity degree between the real-time correlation coefficient matrix and the reference matrix is calculated in the Riemann space, and the working state of the sensor is judged by whether the similarity degree exceeds the limit value. Finally, in order to verify the effectiveness of the method, a semi-physical simulation platform is built, and simulation results show that the method is efficient and robust, which can quickly detect the real-time faults of the sensors in UAV flight control system, and has the potential for real flight verification.