Optimal Voltage-Vector-Sequence-Based Model Predictive Control With Minimum Current Harmonic for Low-Switching-Frequency PMSM Drives

This paper proposes an optimal voltage- vector-sequence-based model predictive control (MPC) method for permanent magnet synchronous motor (PMSM) drives that exploits the properties of optimized pulse patterns (OPPs). Specifically, by analyzing OPPs from a space vector perspective, the optimal voltage vector sequence and the corresponding duration are determined in an efficient way, bringing about satisfying steady-state performance. Moreover, the switching pulses are modified in real time to make the current track the optimal current trajectory that is easily obtained according to the harmonic model of the machine, thus achieving fast dynamic responses. First, the reference voltage vector is calculated and the candidate solutions are obtained based on the allowable switching frequency and the fundamental frequency. Second, the optimal solution is selected by evaluating the designed cost function, which expresses the weighted total harmonic distortion (WTHD) of phase voltage in a closed form. Finally, the optimal current is calculated and the switching instants are modified. To verify the validity of the proposed method, experiments are carried out on a 2.5 kW PMSM drive. At nominal speed and rated load, the proposed method reduces the current THD by 54% and 30% when compared with space- vector-modulation-based MPC (SVM-MPC) and finite-control-set MPC (FCS-MPC) at 625 Hz switching frequency, achieving the steady-state performance nearly as good as that of the OPPs-based control method.