Hybrid Vector Model Predictive Control for Open-Winding PMSM Drives

To improve the control performance of open-winding permanent magnet synchronous motor (OW-PMSM) and suppress zero-sequence current (ZSC), this article proposes a hybrid vector model predictive control (MPC) for OW-PMSM drives. The fast-voltage-vector selection strategy is used to select one optimal nonzero voltage vector (VV) to act on inverter1 (INV1) for each control period, and then, inverter2 (INV2) is controlled with a novel four-segment-mode-vector (FSMV) control strategy to operate the dual inverter separately. In this strategy, the two zero VVs,
$u'_{0}$
and
$u'_{7}$
, are applied at the starting and ending of each control period alternately, while two adjacent nonzero VVs,
$u'_{x}$
and
$u'_{y}$
, with minimum switching times, are applied in the middle of the control period to generate PWM pulses with a switching frequency fixed at half of the control frequency. Furthermore, the durations of two zero VVs (
$u'_{0}$
and
$u'_{7}$
) are optimized to suppress the ZSC based on the zero-axis current deadbeat (DB) control principle. Finally, the experimental findings suggest that the proposed strategy ensures the low-switching frequency of INV1 and the fixed switching frequency of INV2 to obtain good control performance.