A Neural Network Approach for Tracking Control of Uncertain Switched Nonlinear Systems with Unknown Dead-Zone Input

Abstract

This paper is concerned with adaptive neural tracking control problem for uncertain switched nonlinear systems with unknown dead-zone input. Multilayer neural networks (MNNs) are employed to approximate unknown nonlinear functions, and an adaptive neural network controller is introduced to enhance system robustness. With the proposed control scheme, boundedness of all the signals of the closed-loop system is established regardless of the parameter adjustment mechanism, and better tracking control performance can eventually be achieved in view of the universal approximation capability of MNNs. Also, a switching signal is suitably defined using average dwell-time technique. By using a switching control scheme, it is demonstrated that the transient performance and stability can be simultaneously obtained. Finally, a simulation example is given to illustrate the effectiveness and validity of this approach.