Current Prediction Error Reduction Method of Predictive Current Control for Permanent Magnet Synchronous Motors

Predictive current control (PCC) has been widely investigated in motor drive field. The PCC is a model-based control method and the motor prediction model is quite sensitive to motor parameters which may be inaccurately measured and change under different working conditions. In order to reduce the current prediction error of PCC caused by inductance mismatch, a current prediction error reduction method is proposed for permanent magnet synchronous motor (PMSM) by online inductance correction which is based on the relationship between inductance mismatch and current prediction error. Firstly, a proportional regulator is used to obtain the absolute correction value according to the relationship between inductance error and current prediction error. Secondly, the polarity of correction value is judged according to the fluctuation amplitudes of predicted current and actual current. Finally, an adaptive adjustment method for correction period is proposed to ensure the accuracy of inductance correction value. Experimental tests are carried out to verify the effectiveness of proposed method, and results show that the proposed method can obviously reduce current prediction error and improve control performance in practical applications. In addition, the proposed method is suitable for interior and surface mounted PMSMs.