Sliding-mode control for nonlinear state-delayed systems using neural-network approximation

Abstract

The sliding-mode control problem is studied for a class of state-delayed systems with mismatched parameter uncertainties, unknown nonlinearities and external disturbances. By integrating neural-network approximation and the Lyapunov theory into the sliding-mode technique, a neural-network-based sliding-mode control scheme is proposed. The major advantage of the present work over traditional sliding-mode designs is the relaxation of the requirement that the unknown nonlinearities are to be bounded. By means of linear matrix inequalities, a sufficient condition for ensuring the asymptotic stability of the sliding-mode dynamics restricted to the defined sliding surface is given. Further, by utilising a neural-network model to approximate the unknown nonlinearity, a sliding-mode control scheme is proposed to guarantee that the system state trajectory is attracted to the designed sliding surface.