Multilayer neural tracking control with periodically switching method of uncertain switched nonlinear systems

In this paper, a robust tracking control scheme with periodically switching method is presented for a class of switched nonlinear systems which is in the presence of unknown dead-zone input and external disturbances. A neural tracking controller based on multilayer neural networks (MNNs) which are utilized to approximate unknown functions and uncertain nonlinear terms is designed to enhance robustness and maintain boundedness. Adaptive neural updated laws are based on switched Lyapunov function approach, and a periodically switching signal is constructed. It is proved that with the proposed control scheme, the resulting closed-loop switched system is robustly stable such that the satisfactory tracking control performance is well achieved. A simulation example is provided to illustrate the effectiveness and the feasibility of the proposed approach.