Virtual Voltage Vectors Based Model Predictive Control for Symmetrical Six-Phase Induction Machines

Based on the spatial layout of their windings, six-phase induction machines are classified into two categories: symmetrical and asymmetrical. Symmetrical six-phase induction machines (S6PIM) present significant advantages in terms of control, namely the fact that certain voltage vectors (VV) offer high torque/flux generation while simultaneously providing null injection of x-y components. Model predictive control (MPC) reveals as the most promising control approach for these complex machines. Nonetheless, adoption of traditional MPC to S6PIM has significant drawbacks, such as high computation burden and the use of a single switching state over the entire sampling period (Ts), resulting in poor efficiency and deterioration of power and torque quality. To overcome such bottlenecks, this paper proposes deadbeat model predictive control using virtual voltage vectors (DBMPC-VVV). Thanks to the deadbeat feature, only four control actions are evaluated per Ts, resulting in a reduced computational burden. Besides, virtual voltage vectors (VVV) minimize harmonic content and enable improved stator currents tracking. The proposed method's effectiveness is demonstrated by experimental data.