Sensor Fault Diagnosis for Unmanned Quadrotor Helicopter via Adaptive Two-Stage Extended Kalman Filter

This paper proposes a new method to address the problem of sensor Fault Detection and Diagnosis (FDD) of an unmanned quadrotor helicopter. Firstly, in order to eliminate the effect of model uncertainties the gyroscopic effects, the kinematic model of the unmanned quadrotor helicopter replacing the traditional dynamic equations is simply presented. Then, through introducing a set of forgetting factors into the general two-stage extended Kalman filter which is capable of estimating the system states and the faults simultaneously, an Adaptive Two-Stage Extended Kalman Filter (ATSEKF) is developed to diagnose the sensor faults more quickly and accurately. Finally, the proposed approach is validated in three different scenarios including without fault, faults occurred separately and simultaneously. The simulation results demonstrate the effectiveness of the proposed FDD method.