A novel funnel control mechanism for hypersonic flight vehicles under input constraints

Abstract

This article proposes a novel adaptive neural network funnel control scheme based on auxiliary system, which is used to solve the problem that the conventional funnel control strategies have difficulty in solving the control singularity problem caused by transiently increasing tracking errors under input constraints. To begin, the longitudinal motion model of hypersonic flight vehicle is split into a velocity subsystem and an altitude subsystem, with the velocity subsystem being controlled by a proportional integral funnel controller and the altitude subsystem being controlled by an adaptive neural network funnel controller. An adaptive neural network funnel controller is designed for altitude subsystem, which is turned into a pure feedback form. Next, the auxiliary system is used to compensate the control law and tracking error, the neural network is used to estimate the system uncertainty, and the funnel control is used to constrain the tracking error to meet the steady-state and dynamic requirements of hypersonic flight vehicle system. The stability analysis proves the stability of funnel controller. The robustness of our newly proposed method is demonstrated by the simulation results.