基于混沌优化的神经网络预测控制研究

wo new neural network predictive controllers (NNPC) are proposed which combining neural network identifying, COA and the concept of predictive control. The Logistic-map-based COA (LCOA) and Tent-map-based COA (TCOA) are respectively as online optimization. It can avoid calculating the complicated gradient and the inverse matrix in the nonlinear predictive control. As the probability density function of the chaotic sequence for Tent map is the uniform function, the TCOA has the outstanding advantages and higher iterative speed. The simulation studies show the effective performance of the proposed controllers, and also verify the good searching ability of Tent map.