A Graphic Weighting Factor Design Method for Finite Control Set Model Predictive Control of Power Converters

Finite control set model predictive control (FCS-MPC) has been widely used in the control of power converters. However, for converters with multiple optimization objectives, the design of the weighting factors in the cost function is a tricky problem. To solve the weighting factor design problem in multiobjective FCS-MPC, a graphic weighing factor design method is proposed in this article. First, the discrete control variables are taken into the prediction model to determine the state boundaries of each control objective under all the system states. Then, the priorities and weighting factors of the control objectives are determined and designed by adjusting the intersection area of these state boundaries, so as to optimize the system performance. Finally, the operating condition feedback is introduced to realize an adaptive adjusting of the weighting factors so that the relationship of the state boundaries between different control objectives will not be affected by the change of system operating condition, and the controller robustness can be guaranteed. Analysis and experiment are carried out based on the parallel three-level dc-dc system. The results show that the proposed method can effectively solve the weighting factor design problem of multiobjective FCS-MPC. It is simple, efficient, and robust.

Automated summary

This article proposes a graphic weighing factor design method to solve the weighting factor design problem in multiobjective FCS-MPC. The method involves incorporating discrete control variables into the prediction model to determine the state boundaries for each control objective under all system states. The priorities and weighting factors of the control objectives are then determined by adjusting the intersection area of these state boundaries, and operating condition feedback is introduced for adaptive adjusting of the weighting factors to ensure controller robustness.