Neural control of hypersonic flight vehicle model via time-scale decomposition with throttle setting constraint

Abstract

Considering the use of digital computers and samplers in the control circuitry, this paper describes the controller design in discrete time for the longitudinal dynamics of a generic hypersonic flight vehicle (HFV) with Neural Network (NN). Motivated by time-scale decomposition, the states are decomposed into slow dynamics of velocity, altitude and fast dynamics of attitude angles. By command transformation, the reference command for γ−θp−q subsystem is derived from h−γ subsystem. Furthermore, to simplify the backstepping design, we propose the controller for γ−θp−q subsystem from prediction function without virtual controller. For the velocity subsystem, the throttle setting constraint is considered and new NN adaption law is designed by auxiliary error dynamics. The uniformly ultimately boundedness (UUB) of the system is proved by Lyapunov stability method. Simulation results show the effectiveness of the proposed algorithm.