An Improved MPC With Reduced CMV and Current Distortion for PMSM Drives Under Variable DC-Bus Voltage Condition in Electric Vehicles
Jiayao Li, Wensheng Song, Hao Yue, Na Sun, Chenwei Ma, Rong Feng
Abstract
The conventional reduced-common-mode-voltage model predictive control (RCMV-MPC) faces the challenges of low control accuracy and high switching frequency at low speeds. To reduce CMV, improve steady-state performance and lower switching frequency, an improved MPC is proposed for permanent magnet synchronous motor (PMSM) drives with variable dc-bus voltage in electric vehicles (EVs). First, the virtual vectors with low CMV and switching frequency are defined, the optimal virtual vector and its duty cycle are determined. Then, the limitation of designed virtual vectors with fixed dc-bus voltage is illustrated. Furthermore, based on the analysis of the effect of dc-bus voltage on current distortion, the optimal reference dc-bus voltage is adopted to minimize the current distortion. At last, the dc-bus voltage is adjusted to track the reference value by dc–dc converter. An experimental comparison of the conventional methods and the proposed scheme is carried out. The results have verified that the proposed scheme can reduce CMV, current harmonics and switching frequency, especially in low-speed regions. As dc–dc converter has become more popular in motor drives of EVs, the proposed scheme could have a promising application perspective.