Abstract: The performance of fixed-wing unmanned aerial vehicles (FWUAVs) in complex dynamic environments is negatively affected by external time-varying disturbances, sensor faults, and input/output nonlinear constraints. The sensor faults and output constraints can lead to inaccurate measurement of state information, while input saturation can limit the output capacity of actuators. Unfavorable external time-varying disturbances can also reduce flight control performance. The combined effect of these factors may cause the FWUAVs to lose control. To address these issues, this paper proposes a neural network adaptive control method based on the state observer, the fault observer, the disturbance observer, and an auxiliary system. Firstly, a FWUAV attitude dynamics model considering the combined effects of sensor fault, external disturbance, output and input saturation is established. Secondly, the radial basis function neural networks are integrated to design the state observer, the fault observer, and the disturbance observer to estimate unknown states, actuator fault, and external disturbances, respectively. The outputs of these three observers, the first-order filter and the auxiliary system state variables are combined for the controller design. Meanwhile, the Lyapunov stability theory is used to prove that all signals in the closed-loop system are ultimately uniformly bounded. Finally, simulation results show that the proposed method can ensure the stable flight of FWUAVs under the combined effects of external time-varying disturbances, sensor faults, and input/output saturation.

Key words: Fixed-wing unmanned aerial vehicle, Sensor fault, Output saturation, Input saturation, Disturbance observer