Design of an adaptive fuzzy neural network controller for a kind of the chaotic systems

An adaptive fuzzy neural network controller for a kind of the chaotic systems is designed based on RBF neural network. Firstly, the fuzzy system structured by RBF neural network is used to approximate the non-linear dynamic system function in high-precision. The parameter linearization technique of Taylor series expansion is employed to do partial linearization of membership function for RBF neural network. Then a controller is designed in order to tune membership function´s parameters and connection weights simultaneously. And the controller is with the advantages of on-line optimizing and fast convergence. Finally, the simulation results for the chaotic system illuminate that the proposed controller can reach more favorable tracking performance with characteristic signal and smaller tracking error.