Robust adaptive neural tracking control for a class of switched affine nonlinear systems

In this paper, the adaptive tracking control problem for a class of switched affine nonlinear systems is investigated. We employ RBF neural networks (RBF NNs) to approximate unknown nonlinear functions. Due to the existence of approximation errors of the neural networks and external disturbance, we, respectively, utilize sliding mode method and H∞H∞ method as the robust controller to enhance system robustness and maintain boundedness. In addition, admissible switching laws are constructed and the weights of RBF NNs updated laws are chosen by switched Lyapunov function approach. With the two proposed methods, we can both prove that the resulting closed-loop switched system is robustly stable and uniformly ultimately bounded (UUB), and the output tracking errors converge to 0. Finally, we give a simulation example to demonstrate the proposed methods and do a comparative analysis.