Back-Stepping Fault-Tolerant Control Method of Morphing Aircraft Based on Fixed-Time Neural Network Observer

Abstract

Morphing aircraft hold the advantages of superior performance compared with their conventional counterparts, but face the challenge of robust attitude tracking control given failures and uncertainties coupled with external disturbances. Therefore, this paper proposes a back-stepping fault-tolerant control method based on fixed-time neural network observer. The longitudinal dynamics model of morphing aircraft is built with augmented actuator failure modes. The equations of motion subsequently are reduced separately into velocity and altitude subsystems. Furthermore, the altitude and velocity tracking controllers are designed using the back-stepping method with the fixed-time neural network observer. The neural network is used to estimate the unknown dynamics, whereas the disturbance observer ensures that the error converges within a fixed time independent of initial system states. The convergence of disturbance observation errors and stability of the control law were proved by applying the Lyapunov theorem. Comparative simulation results showed that the maximum altitude and velocity tracking errors generated by the method proposed in this paper were reduced by 95.4% and 88%, respectively, compared with those of the fixed-time observer method when the actuator failed. The robustness of the closed-loop system is improved significantly with the proposed control method.