Study Of Generalized Predictive Control Algorithm Based On Support Vector Machines

On the basis of the study on development of Generalized Predictive Control algorithm, this paper proposes the methods which combine the Support Vector Machine algorithm and Generalized Predictive Control algorithm for the problems in GPC,such as the identification for model parameters and prediction model error compensation issue. There are detailed analysis of the principle of generalized predictive control algorithm, model identification function of support vector machine in this paper. And it applys the function to generalized predictive control algorithm,proposes generalized predictive control and error compensation generalized predictive control based on support vector machine model identification, and does simulation experiments using MATLAB for the specific objects of main steam pressure and long-time delay in thermal power plant. The simulation results show the effectiveness of the method, and it provides a theoretical basis for practical engineering application.