Hybrid adaptive CMAC sliding mode controller design for unknown nonlinear system

Abstract

A new hybrid adaptive cerebeller model articulation controller (CMAC) sliding mode control system is developed for unknown nonlinear systems. The hybrid adaptive CMAC sliding mode controller (HACSMC) uses the direct and indirect adaptive CMAC controllers to perform the equivalent control of sliding mode control (SMC) and a weighting factor is adopted to sum together the control efforts from the direct and indirect adaptive CMAC controller. The CMAC network is used to approximate the switching control law of SMC. A supervisory controller is appended to the HACSMC to guarantee the states staying in the boundary layer. All adaptive laws of the control system are derived based on Lyapunov stability theorem, so that the stability of the system can be guaranteed. Finally, the proposed control system is applied to inverted pendulum system. The simulation results show that the HACSMC can not only make control system have good tracking performance and strong robustness but also have more flexibility during the design process.