Model Predictive Control for PMSM Based on the Elimination of Current Prediction Errors

Predictive control is recognized as a promising control method. Model predictive control (MPC), as one of its branches, has been widely applied in the field of power electronics and motor drives due to the advantage of simple structure and satisfactory dynamic performance. However, MPC is a strongly motor-parameters-dependent method. In practical applications, the challenge exists that motor-parameters may change, bringing the error between actual prediction currents and ideal prediction values and thus directly affecting the control performance of permanent magnet synchronous motor (PMSM) system. To eliminate the current prediction errors and enhance the control performance of PMSM system, an improved MPC based on the elimination of current prediction errors (MPC-CPE) method is proposed in this paper. First, the conventional MPC method and the influence of parameter mismatch are introduced. Then, MPC-CPE method is put forward where the current prediction errors corresponding to the eight basic voltage vectors are calculated and an updated cost function is used to choose the optimal one. In this way, the bad influence caused by parameter mismatch can be eliminated effectively. It is worth noting that the calculation process in MPC-CPE method is relatively simple. Finally, experiments are finished to prove that MPC-CPE method can significantly circumvent the problem of parameter mismatch and enhance the control performance of PMSM system on a 1.25-kW experimental platform.