Scholarlink

<p>The adoption of flat-wire motors in new energy vehicle traction systems is accelerating rapidly, with the integration of highly conductive graphene-copper (Gr/Cu) composites emerging as a key future development direction for motor technology. The complex calculation of AC copper losses has constrained motor design optimization, prompting the development of an innovative approach combining proper orthogonal decomposition with neural networks for generalized winding magnetic field solutions. This method enables rapid magnetic field determination at any operating point within Gr/Cu flat-wire motor operating conditions. Results show the proposed method achieves three orders of magnitude faster computation speed compared to finite element methods, with maximum magnetic field errors below 0.007T within the generalization range. Prototype testing has validated the method’s effectiveness, demonstrating maximum relative experimental errors not exceeding 9.6%.</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

A Fast Calculation Method for AC Losses Based on Graphene Copper Electrical Machine

  • 期刊论文

  • 发表日期：

    2025年10月6日

查看全文
  • Jiaqiang Li,
  • Ruilin Pei,
  • Yuejun An

摘要

The adoption of flat-wire motors in new energy vehicle traction systems is accelerating rapidly, with the integration of highly conductive graphene-copper (Gr/Cu) composites emerging as a key future development direction for motor technology. The complex calculation of AC copper losses has constrained motor design optimization, prompting the development of an innovative approach combining proper orthogonal decomposition with neural networks for generalized winding magnetic field solutions. This method enables rapid magnetic field determination at any operating point within Gr/Cu flat-wire motor operating conditions. Results show the proposed method achieves three orders of magnitude faster computation speed compared to finite element methods, with maximum magnetic field errors below 0.007T within the generalization range. Prototype testing has validated the method’s effectiveness, demonstrating maximum relative experimental errors not exceeding 9.6%.