Design of a smooth adaptive CMAC neural controller for a chaotic dynamic system

In the conventional adaptive CMAC neural controller design, the compensator is usually designed in a sliding-mode control form, thus it will occur the chattering phenomena in control effort. To tackle this problem of chattering phenomena, this paper proposes a smooth adaptive CMAC neural control (SACNC) system for a chaotic dynamic system. The proposed SACNC system is composed of a neural controller and a saturation compensator. The parameter adaptive algorithms of SACNC are derived based on Lyapunov stability theory, so the asymptotic stability can be achieved. Finally, simulation results show the proposed SACNC system can achieve favorable tracking performance without any chattering phenomena.