Scholarlink

<p>To improve efficiency and reduce electromagnetic noise in permanent magnet synchronous motor (PMSM) drives for new energy vehicles, this study proposes a novel multi-mode cooperative control strategy. The strategy integrates time-shifted SVPWM/DPWM hybrid modulation, harmonic closed-loop suppression, and a deep over-modulation algorithm. Its novelty lies in this integrated approach and the establishment of a dynamic operation mapping model to balance efficiency and NVH performance. Offline parameter calibration identified quadrature- and direct-axis inductances and fitted flux linkage surfaces, building a 2D current MAP database. A dynamic table feedforward control architecture was developed for precise harmonic extraction and active suppression. Using the Nyquist stability criterion and pole distribution analysis, the frequency-domain model of the deep over-modulation closed-loop system was verified to be globally stable within a modulation depth of 1.05. All system poles remain in the left half-plane, ensuring dynamic robustness. Compared to conventional methods, the proposed 24/48-order harmonic suppression achieves maximum noise reductions of 25.55 dB (drive mode) and 30.6 dB (generation mode). The improved over-modulation algorithm extends stable operation to a modulation depth of 1.05, surpassing typical limits. This strategy reduces switching and conduction losses. It achieves a 5.94% overall efficiency improvement while maintaining superior acoustic performance.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Multimodal SVPWM-DPWM Modulation with Harmonic Injection for Efficiency and NVH Optimization in PMSM Drives

  • 期刊论文

  • 发表日期：

    2026年4月7日

查看全文
  • Junying Li,
  • Kai Guo,
  • Weiyuan Sun,
  • Yumei Zhang,
  • Tao Deng

摘要

To improve efficiency and reduce electromagnetic noise in permanent magnet synchronous motor (PMSM) drives for new energy vehicles, this study proposes a novel multi-mode cooperative control strategy. The strategy integrates time-shifted SVPWM/DPWM hybrid modulation, harmonic closed-loop suppression, and a deep over-modulation algorithm. Its novelty lies in this integrated approach and the establishment of a dynamic operation mapping model to balance efficiency and NVH performance. Offline parameter calibration identified quadrature- and direct-axis inductances and fitted flux linkage surfaces, building a 2D current MAP database. A dynamic table feedforward control architecture was developed for precise harmonic extraction and active suppression. Using the Nyquist stability criterion and pole distribution analysis, the frequency-domain model of the deep over-modulation closed-loop system was verified to be globally stable within a modulation depth of 1.05. All system poles remain in the left half-plane, ensuring dynamic robustness. Compared to conventional methods, the proposed 24/48-order harmonic suppression achieves maximum noise reductions of 25.55 dB (drive mode) and 30.6 dB (generation mode). The improved over-modulation algorithm extends stable operation to a modulation depth of 1.05, surpassing typical limits. This strategy reduces switching and conduction losses. It achieves a 5.94% overall efficiency improvement while maintaining superior acoustic performance.