An Improved Model-Free Finite Control Set Predictive Power Control for PWM Rectifiers

The performance of conventional finite control set predictive control (FCSPC) is deteriorated when model parameters are not accurately known. To improve the robustness against model mismatches, model-free FCSPC (MF-FCSPC), which relies on only sensored information, has been proposed in the existing literature. However, the prior MF-FCSPC suffers from random power spikes and higher current distortions. In this paper, a MF-FCSPC based on an enhanced mechanism for storing and updating current difference is proposed, which shows strong robustness as that of the prior MF-FCSPC but still maintaining good performance as like a well-tuned FCSPC. To verify the effectiveness of the developed method, several simulation and experimental tests were carried out on a two-level PWM rectifier platform. The obtained results confirm the feasibility and superiority of the proposed MF-FCSPC when compared with the conventional FCSPC and prior MF-FCSPC.