Novel Sensor Fault Detection and Isolation for an Unmanned Aerial Vehicle

To avoid adverse impacts on humans and the environment and upon the system itself, early detection and isolation of faults and failures in the flight control system are crucial. This paper focuses on designing a novel detection and isolation (FDI) of sensor faults in quadrotors. The traditional nonlinear fault detection techniques including integral chain differentiator and high gain observer are incorporated to realize the accurate reconstruction of the sensor faults. However, the proposed approach based on the improved high gain observer is employed that offers fast detection of sensor faults with improved detection accuracy and lower gain requirement than the traditional observer. Various simulation results validate the superiority of the proposed FDI method and prove its efficacy under the influence of different fault scenarios.