An Improved Model-Free Predictive Current Control for PMSM Under Low-Speed Condition

The conventional finite-control set model predictive current control (MPCC) for surface-mounted permanent magnet synchronous motor (SPMSM) is been widely used due to its quick current response and ability of stator current behavior prediction. However, the stator current predicted the accuracy of MPCC could be decreased when the motor parameters are variated. Moreover, the single-switch state modulation of MPCC leads to outstanding current harmonics under low-speed steady-state conditions. To overcome the model dependency of MPCC, an improved model-free predictive current control (MFPCC) method based on data-driven is proposed in this article. First, the stator current prediction model in MPCC under low-speed conditions is analyzed theoretically. The prediction model is approximately simplified based on the analyzed results. Then, the sum terms related to the motor parameters in the simplified model could be calculated online by the stator current and voltage measurement data. To decrease the stator current harmonics under low-speed conditions, a dual-voltage vector modulation method based on vector projection is proposed. This modulation could reduce the modulation error in stator voltage amplitude by introducing the zero-voltage vector. Finally, the control effectiveness and real-time implementation of the proposed method are verified by experiments under different low-speed conditions.