A novel neural prescribed performance control design for constrained flexible hypersonic vehicles with compensation system

In this study, a novel prescribed performance function and auxiliary system are investigated for the constrained flexible hypersonic vehicles. The non-affine model of hypersonic vehicles is decomposed into velocity and altitude dynamic systems. For the velocity one, the adaptive neural prescribed controller is devised to guarantee that the tracking error converges to an arbitrary small compact. Then, the altitude dynamics subsystem is delivered as a pure feedback non-affine pattern, which gets the controller out of the dependence on precise affine model. On this premise, the exploited prescribed performance mechanism can not only avoid obtaining the initial sign of tracking error in advance but also can be employed in solving the problem that the conventional prescribed performance function could not cover the tracking error when the altitude subsystem is input constraint. And a Nussbaum-type function is employed to handle the control direction. Meanwhile, the controller utilizes a new auxiliary compensation algorithm to pursue the stable situation for the input restricted system. In order to cut down the computational burden, neural networks employ the min-learning parameter theory to approximate uncertainties in the subsystems. Finally, the simulation results convey that the presented approach has a superior dynamic and steady-state performance.