Adaptive dynamic RBF neural controller design for a class of nonlinear systems

In this paper, an adaptive DRBF neural control (ADNC) system which is composed of a neural controller and a smooth compensator is proposed. The neural controller utilizes a dynamic radial basis function (DRBF) network to online mimic an ideal controller and the smooth compensator is designed to eliminate the effect of the approximation error between the ideal controller and neural controller. The DRBF network can self-organizing its network structure. All the controller parameters of the proposed ADNC system are online tuned in the Lyapunov sense, thus the stability analytic shows the system output can exponentially converge to a small neighborhood of the trajectory command. Finally, the proposed ADNC system is applied to a chaotic system and a DC motor. Simulation and experimental results verify that a favorable tracking performance and no chattering phenomena can be achieved by the proposed ADNC system.