Algebraic Tuning Guidelines for Model Predictive Torque and Flux Control

Abstract

Finite control set model predictive torque and flux control uses a scalar weight in its cost function to determine the tradeoff between the torque and flux magnitude tracking error terms. This weight strongly influences the current distortions. It will be shown that an optimal weight can be computed algebraically, which minimizes the current distortions. The resulting predictive torque and flux controller achieves current distortions per switching frequency that are very similar to those of a predictive current controller, provided that the corresponding weight on the switching effort is selected appropriately. To this end, a second analytical expression will be derived.