Support vector machine and genetic algorithm based predictive control for active power filter

A new control method for active power filters using support vector machine (SVM) is presented. In the strategy, SVM is employed to model and predict future harmonic compensating current, it has the advantages of nonexistence of local minima solutions, automatic choice of model complexity and good generalization performance. Based on the model output, Genetic algorithm optimization method is adopted to produce proper value of control vector, this value is adequately modulated by means of a space vector PWM modulator which generate proper gating patterns of the inverter switches to maintain tracking of reference current. The SVM based predictive algorithm is used in internal model control scheme to compensate for process disturbances, measurement noise and modeling errors. The proposed method is applied to the control of a shunt active power filter, simulation results show SVM based predictive controller is more effective and feasible than PI control or digit adaptive control.