Finite-Control-Set Model Predictive Current Closed-Loop Control Based on Prediction Error Compensation for PMSM

Finite-control-set model predictive control (FCS-MPC) for permanent magnet synchronous motors (PMSMs) has attracted attention due to its better theoretical performance. However, as motor operating conditions change, motor parameter mismatch can lead to intolerable prediction errors which significantly deteriorate stator current harmonics and torque ripples. To solve this issue, a finitecontrol-set model predictive current closed-loop control strategy is proposed. First, based on the analysis of the prediction equations, the voltage-independent and voltage-dependent parts of the prediction errors are separated. Secondly, according to the different features of prediction errors caused by zero and non-zero vectors, the decoupling of the two parts of prediction error is realized. And PI controllers are introduced to observe the two different types of DC components respectively to make the observation more stable and accurate. Thirdly, feedback compensation is performed to modify the prediction equations. With the design of model predictive current closed-loop control, the prediction error quickly converges to the minimum. Finally, the experimental outcomes prove the effectiveness of this strategy. © 2024, Electromagnetics Academy. All rights reserved.