163 Nonlinear System Multi-Step Predictive Control Based Neural Network Model and Genetic Algorithm

A nonlinear multi-step predictive control strategy using Radial Basis Function Neural Network (RBFNN) as multi-step predictive model for nonlinear complicated industrialized process with time delay, slow time variety and highly disturbance is proposed in the paper. The modified elitist preserved genetic algorithm is used to obtain the online nonlinear optimization. Simulation results demonstrate that the strategy has good robustness and the resisting time variety ability.