Sliding mode control based on the modified fuzzy RBF for uncertain chaotic systems

This paper presents a novel method to controlling uncertain chaotic systems by means of sliding mode control based on fuzzy radial basis function neural network(FRBF). The proposed method combines the advantages of weights direct-determination, sliding mode control compensator and neural network. The neural network with five layer is constructed. For this neural network, the activation is sigmoid membership function, and the optimal weights received by weights direct-determination. The identification of chaotic system is first inferred by the modified FRBF in order to realize the nonlinear mapping between input and output, and its approximation ability to any chaotic system is perfective. Then the sliding mode compensation control is implemented by using the FRBF model. The performance of simulation results show the scheme is effective and feasible for uncertain chaotic system, and the robustness is provided on the parametric and extern disturbance.