There is a large computational burden in the traditional optimal duty cycle model predictive control due to the six times evaluations in each cycle. After constant analysis of the selection process of the optimal voltage vectors, an improved SVM-based optimal duty cycle MPCC method is proposed in this paper. Compared with the traditional method, this method simplifies evaluation process from six times to three times in each cycle which optimizes half of the computation while ensuring the performance of the system. Simulation results show that the effect of proposed method is similar to the traditional one, and it requires less computation.

model predictive control;current control;optimal duty cycle control;space vector modulation;algorithm optimization