Neural-approximation-based robust adaptive control of flexible air-breathing hypersonic vehicles with parametric uncertainties and control input constraints

In this paper, a neural-approximation-based robust adaptive control methodology is proposed for a constrained flexible air-breathing hypersonic vehicle (FAHV) subject to parametric uncertainties. To reduce the computational costs, only two radial basis function neural networks (RBFNNs) are applied to approximate the lumped unknown nonlinearities of the velocity subsystem and the altitude subsystem, while guaranteeing the exploited controller with satisfactory robustness against system uncertainties. Furthermore, a minimal-learning parameter (MLP) approach is employed to update the norm rather than the elements of RBFNNs' weight vectors, which yields a low computational load design. By constructing a novel auxiliary system to compensate the desired control laws, the effects of magnitude constraints on actuators are tackled. The Lyapunov synthesis proves that the closed-loop uniformly ultimately bounded stability can be achieved even when the physical limitations on actuators are in effect. Finally, simulation results are presented to verify the efficacy of the addressed control strategy in the presence of uncertain parameters, external disturbances and control input constraints.