LAUSR

Abstract In this paper, a novel envelope-constraint-based neural controller is addressed for air-breathing hypersonic vehicles with unknown non-affine formulations. The vehicle dynamics is decomposed into velocity subsystem and altitude subsystem that are further described as unknown non-affine expressions. Neural networks are employed to approximate the unknown terms of each subsystem based on Mean Value Theorem, which rejects system uncertainties and non-affine characteristics. New learning algorithms are developed for adaptive parameters to stabilize the closed-loop system and reduce online parameters, yielding a low computational load design. The remarkable advantages are that complicated design procedure of back-stepping is avoided, and desired transient performance is guaranteed for velocity and altitude tracking errors via envelope constraint. The stability of closed-loop control system is proved utilizing Barrier Lyapunov functions. Finally, numerical simulation results are presented to validate the effectiveness of the proposed control approach.