Adaptive neural network disturbance observer based nonsingular fast terminal sliding mode control for a constrained flexible air-breathing hypersonic vehicle

In this paper, a novel constrained nonsingular fast terminal sliding mode control scheme based on adaptive neural network disturbance observer is proposed for a flexible air-breathing hypersonic vehicle in the presence of diverse disturbances and actuator constraints. Firstly, velocity and altitude subsystems in the strict feedback formulations are obtained by decomposing the longitudinal dynamics of flexible air-breathing hypersonic vehicle, while uncertainties with regard to flexible effects, aerodynamic parameter uncertainties, modeling errors, and external disturbances are formed as the lumped disturbances which are excellently estimated by the proposed adaptive neural network disturbance observer with the adaptive regulation laws of weight matrices. Then based on the nonsingular fast terminal sliding mode control, the proposed scheme integrated with adaptive neural network disturbance observer is developed to design the controllers with nonsingularity and fast convergent rate in order to provide robust and fast tracking performance of velocity and altitude. Furthermore, to tackle the saturation effects caused by the constraints of actuators, the auxiliary systems constructed in the proposed scheme are conducted to compensate the desired controllers timely. Lyapunov stability analyses prove that the stable tracking errors of velocity and altitude are bounded with the sufficiently small regions around zero even when flexible air-breathing hypersonic vehicle is subject to lumped disturbances and actuator constraints. Finally, the contrastive simulation results demonstrate that the proposed scheme provides the superior tracking performance and the effectiveness in tackling actuator constraints and counteracting lumped disturbances.