Adaptive neural network control of hypersonic flight vehicle with actuator constraints

An adaptive neural network controller based on the back-stepping is developed for a generic hypersonic flight vehicle. The controller addresses two main problems, including model uncertainty and input saturations. First, the longitudinal dynamic model is transformed into an altitude subsystem and a velocity subsystem with the strict feedback form. Then, the combination of the adaptive neural network controller via the back-stepping method and command filter is utilized to track the altitude and velocity command. The stability analysis of the closed-loop system is proved based on Lyapunov’s stability theorem. Simulation results display that the proposed controller is robust in terms of parametric uncertainty and meets the performance requirements with input saturation.