Robust approximation-based adaptive control of multiple state delayed nonlinear systems with unmodeled dynamics

Abstract

This paper addresses the problem of tracking control for a class of uncertain nonstrict-feedback nonlinear systems subject to multiple state time-varying delays and unmodeled dynamics. To overcome the design difficulty in system dynamical uncertainties, radial basis function neural networks are employed to approximate the black-box functions. Novel con- tinuous functions that deal with whole states uncertainties are introduced in each step of the adaptive backstepping to make the controller design feasible. The robust problem caused by unmodeled dynamics when constructing a stable controller is solved by employing an auxiliary signal to regulate its boundedness. A novel Lyapunov-Krasovskii functional is developed to compensate for the delayed nonlinearity without requiring the priori knowledge of its upper bound functions. Based on the proposed robust adaptive neural controller, all the closed-loop signals are semiglobal uniformly ultimately bounded with good tracking performance.