Composite dynamic surface control of hypersonic flight dynamics using neural networks

This paper addresses the composite neural tracking control for the longitudinal dynamics of hy- personic flight dynamics. The dynamics is decoupled into velocity subsystem, altitude subsystem, and attitude subsystem. For the altitude subsystem, the reference command of flight path angle is derived for the attitude subsystem. To deal with the system uncertainty and provide efficient neural learning, the composite law for neural weights updating is studied with both tracking error and modeling error. The uniformly ultimate bound- edness stability is guaranteed via Lyapunov approach. Under the dynamic surface control with novel neural design, the neural system converges in a faster mode and better tracking performance is obtained. Simulation results are presented to show the effectiveness of the design.