Fault Detection of UAV Fault Based on a SFUKF

Abstract

Abstract The UAV system is a typical closed-loop control system. Its good robustness can inhibit the fault signal, which poses certain difficulties for the detection of early or small amplitude faults. In this paper, a nonlinear longitudinal control system model of a class of UAVs is established, and a fault detection method based on the suboptimal fading unscented Kalman filter (SFUKF) is designed. Aiming at the common failure of actuators and sensors of the drone, this paper proves that the method realizes the fault detection of the airspeed tube blockage and the elevator part failure by simulation.