Enhanced Predictive Current Control of PMSM Drive With Virtual Voltage Space Vectors

Abstract

Model predictive current control (MPCC) is commonly used method for the control of permanent magnet synchronous motor (PMSM) drive. MPCC is simple to implement and faster to execute on digital platform. The classical cost function in MPCC does not contain any weighting factor thereby resulting in a simpler cost function for the selection of optimal voltage space vector (VSV). Conventional MPCC (C-MPCC) applied to PMSM results in higher torque ripples. The ripples in the torque and stator flux can be reduced by reducing the duration of application of active VSV. This article proposes an MPCC based on virtual voltage space vector (VVSV) for reducing the torque ripples and improve the THD in stator current without affecting dynamic performance. Duty ratio for VVSV is chosen proportional to commanded speed to reduce change in stator current thereby reducing ripple in torque stator flux. The VVSVs have magnitude near to developed back-EMF space vector. Proposed MPCC is compared with C-MPCC and two other MPCC based on VVSV. Effectiveness of proposed MPCC is successfully verified after comparing the results of experimentation at different speed and load conditions. Proposed MPCC fails to reduce ripple in torque and stator flux above base speed and proposed scheme collapses into C-MPCC.