Nonlinear model predictive control based on fuzzy wavelet neural network and chaos optimization

In this paper a combined controller is proposed for nonlinear dynamical systems. The controller is constructed by a fuzzy wavelet network and nonlinear model predictive control. Chaotic optimization, which is fast and robust, is applied to generate optimized controlled input in nonlinear model predictive control. The ability of the fuzzy wavelet neural network and the proposed controller is shown by simulation. It is illustrated that the proposed method is able to increase the speed of tracking in addition to having very little steady state error. Using chaotic optimization makes the controller robust in the presence of noise.