Synthetic Vectors-Based Predictive Control of Dual Three-Phase PMSMs for Current Harmonics Mitigation Considering Average Deception Effect

This article proposes a novel model predictive control (MPC) scheme for dual three-phase permanent magnet synchronous machines (DTP-PMSMs) to further suppress the x-y current harmonics caused by unmodeled disturbances such as inverter nonlinearity and back-EMF harmonics, in which the average deception effect is taken into account to avoid larger x-y current spikes. Instead of following the conventional virtual vectors (VVs) to synthesize a zero-average voltage in the x-y subspace, the rotating nonzero voltage vector is dynamically modulated in a control period by scheduling the dwell times of the three adjacent large vectors and zero vectors to realize the closed-loop current control in the x-y subspace. In the implementation of synthetic vectors, to ensure that large vectors with a higher ratio of alpha-beta/x-y voltage are applied directly to the inverter, the compared registers A and B of the PWM module are configured to solve the state conflicts issue in the switching sequence. Since the x-y unmodeled disturbances have considerable periodic components, a resonant extended state observer (RESO) is employed to provide a more accurate estimate within a limited bandwidth, resulting in more targeted suppression of the fifth and seventh harmonics. Experimental results validate the superiority of the proposed strategy.