Model Predictive Control With Space Vector Modulation Based on a Voltage Angle for Driving Open-End Winding IPMSM

Abstract

This paper proposes a model predictive control with space vector modulation (MPC-SVM) method that is based on using a voltage angle to reduce ripple components in open-end winding interior permanent magnet synchronous motor (OEW-IPMSM). Conventional model predictive control (CMPC) provides fast transient states by applying an actual voltage vector each switching period. However, due to the limited magnitude of voltage vectors, the CMPC causes large torque ripple and low current quality. The MPC-SVM method proposed herein generates virtual voltage vectors based on the voltage angle and surrounding area. Virtual voltage vectors provide various magnitudes and positions except for areas that are not in use. The optimal reference voltage vector is derived as the result of the cost function operation applied through the space vector modulation (SVM) method, and it improves the steady-state compared to the CMPC. Therefore, torque ripple is reduced and current quality is improved by using fewer virtual voltage vectors, and this results in superior steady-state characteristics. The validity of the proposed MPC-SVM method based on the voltage angle for driving OEW-IPMSM is affirmed through simulation and experimental results in comparison to the CMPC.