Tracking Control for The UAV With Disturbance Observer and RBFNN under Actuator Fault

This paper investigates the fault-tolerant tracking control for the unmanned aerial vehicle (UAV) under disturbance. The model of UAV is converted to the one that is convenient for fault-tolerant analysis by twice matrix coordinate transformation first. Then, the RBFNN is employed to approximate the unknown function induced by the actuator fault. To avoid the singularity of the actuator fault-dependent parameter when the actuator is fault free and decouple the tracking error and disturbance estimation error, based on the above-transformed model, a disturbance observer with a decoupling term is designed. Thereafter, combined with the RBFNN and the designed disturbance observer, the control input is designed, and sufficient conditions of stability for the closed-loop system are derived by designing the actuator fault-dependent Lyapunov function. Finally, the comparative simulations show the feasibility and performance of the designed control strategy.