Medium-frequency transformer (MFT) as one core component of power electronic traction transformer exhibits distinct performance characteristics and design requirements compared to conventional transformer due to its operation at medium-frequency ranges and direct connection to power electronic converters, necessitating higher optimization standards. This paper analyzes the frequency-dependent variation of loss coefficients in the loss separation method based on experimental data of nanocrystalline core losses. A variable-coefficient loss calculation model is subsequently proposed, establishing an accurate computational approach for core loss calculation method. Furthermore, the Dowell model is modified using the area equivalence principle to achieve precise calculation of Litz wire losses. Then, the multiple strategies assisted non-dominated sorting dung beetle optimizer (MS-NSDBO) is proposed. Finally, taking loss and volume as object, the MFT is optimally designed using MS-NSDBO.

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

Optimization of Medium-Frequency Transformer in Power Electronic Traction Transformer

  • Jiahui Zhang,
  • Yimeng Guo,
  • Ziyan Ren,
  • Dianhai Zhang

摘要

Medium-frequency transformer (MFT) as one core component of power electronic traction transformer exhibits distinct performance characteristics and design requirements compared to conventional transformer due to its operation at medium-frequency ranges and direct connection to power electronic converters, necessitating higher optimization standards. This paper analyzes the frequency-dependent variation of loss coefficients in the loss separation method based on experimental data of nanocrystalline core losses. A variable-coefficient loss calculation model is subsequently proposed, establishing an accurate computational approach for core loss calculation method. Furthermore, the Dowell model is modified using the area equivalence principle to achieve precise calculation of Litz wire losses. Then, the multiple strategies assisted non-dominated sorting dung beetle optimizer (MS-NSDBO) is proposed. Finally, taking loss and volume as object, the MFT is optimally designed using MS-NSDBO.