An Improved Model-Free Predictive Current Control for PMSM Drives Based on Current Circle Tracking Under Low-Speed Conditions

The model predictive current control (MPCC) method for permanent magnet synchronous motor (PMSM) has been widely applied due to its quick current response. However, the performance of MPCC relies on the real-time model accuracy deeply. Meanwhile, the large stator current harmonic under low-speed conditions is another defect of MPCC. To solve these problems, an improved model-free predictive current control method based on online data-driven, and a novel modulation method based on current increment synthesis are proposed in this paper. Firstly, the unknown polynomial related to PMSM parameters (regarding resistance, inductance, and flux-linkage) is defined as a lumped term in the MPCC model. The stator current and voltage measurements in the latest 15 sampling periods are registered as a database by the rolling update mechanism, and used to identify the unknown lumped term. Secondly, according to the identified result, the current increments corresponding to each active voltage in the next sampling period are updated online. Finally, the duty cycle of the inverter at the next sampling period is calculated based on the minimization of the predicted current increment tracking error. The effectiveness and real-time implementation of the proposed method are verified by experiments under different working conditions.