关键词: 永磁同步电机, 无模型预测电流控制, 改进型扩张状态观测器, 非参数模型电感辨识, 参数鲁棒性

Abstract: In response to the problem of decreased current control performance and increased torque ripple in model-free predictive current control strategies of permanent magnet synchronous motor based on extended state observers when the motor inductance parameters were mismatched, a model-free predictive current control strategy based on an improved extended state observer was proposed. A non parametric model inductance identification algorithm was added to the extended state observer. The algorithm used the d-axis disturbance estimation to calculate the motor identification inductance, which was fed back to
the current and disturbance estimation of the extended state observer at the next moment, eliminating the adverse effect of inductance mismatch on traditional extended state observers. The identified inductance was used for calculating the output voltage command of model-free predictive current control, further enhancing the parameter robustness of model-free predictive current control. Based on the control strategy, a motor torque control system was constructed for experiments. The results showed that the control strategy had strong parameter robustness and good current control performance. It can effectively reduce motor torque ripple in case of motor inductance mismatch, and has high engineering application value.

Key words: permanent magnet synchronous motor, model-free predictive current control ( MFPCC ), improved extended state observer( ESO), non parametric model inductance identification, parameter robustness