Model predictive control method to reduce common-mode voltage for permanent-magnet synchronous machine drives

All the eight voltage vectors of a two-level voltage source inverter are treated as candidate voltages applied to the cost function in the conventional model predictive control (MPC) in order to minimize the error between reference and practical stator current. Due to the usage of zero vector voltage, there are large common-mode voltage (CMV) which will deteriorates the operation performance and reliability of the drive system. A MPC- based CMV reduction method was presented to reduce the peak value of the CMV at the expense of large current ripple. Then, a two-vector-based MPC method utilizing two selected nonzero voltage vectors at every sampling period was presented to achieve satisfactory current ripple performance, while it significantly increased the computational burden and system complexity. This paper investigate the CMV for permanent magnet synchronous machine drives and proposes a novel two-vector-based MPC scheme to reduce the CMV without deterioration in the current ripple performance. Furthermore, a simple and efficient duty ratio determination method of the two selected vectors is designed in order to avoid significantly increasing the computational burden. The simulation results will be presented to confirm that the proposed method can limit the CMV with ± V
_dc
/6 while preserving the low current ripple performance and fast dynamic response of the conventional MPC method.