Adaptive neural network-based control of uncertain nonlinear systems with time-varying full-state constraints and input constraint

Abstract

This paper investigates the problem of state-constraint adaptive neural network-based tracking control for a class of nonlinear systems with input saturation constraint. The considered systems are with uncertain nonlinearities which are not required to be globally Lipschitz or be with a prior knowledge of the structure. To facilitate the stability analysis, radial basis function neural networks (RBFNNs) are first utilized to approximate the unknown nonlinear terms. The constraint problem of input saturation often appears in the control system. To solve above issue, a novel adaptive control scheme is proposed with the help of an augmented function with auxiliary control signal, which ensures that all the closed-loop signals are semi-globally uniformly ultimately bounded. On the other hand, to guarantee better transient performance under input saturation, an improved barrier Lyapunov function with time-varying barriers is developed, which makes the tracking errors preserve within the specified constraint bounds. Simulation results are given to demonstrate the effectiveness of the proposed approach.