Scholarlink

<p>Efficient surrogate modeling is essential for statistical assessment of electromagnetic compatibility problems in modern electrical systems. Yet, traditional approaches fail due to parameter-dependent resonance shifts that violate smoothness assumptions. This paper develops a&#xa0;methodology combining pole-residue decomposition with clustering in order to handle the ordering inconsistencies inherent in vector fitting algorithms. Applied to the differential mode impedance of a&#xa0;traction motor, the approach demonstrates that clustering resolves the fundamental pole-residue ordering problem, enabling polynomial chaos expansion surrogates to achieve the required accuracy.</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

Pole-residue surrogate modeling for electromagnetic compatibility analysis using clustering

  • 期刊论文

  • 发表日期：

    2026年2月3日

查看全文
  • Andreas Klaedtke,
  • Moritz von Tresckow,
  • Christian Bergfried,
  • Yvonne Spaeck-Leigsnering,
  • Frieder Schuller,
  • Marcel Waldhof,
  • Hermann Aichele,
  • Herbert De Gersem

摘要

Efficient surrogate modeling is essential for statistical assessment of electromagnetic compatibility problems in modern electrical systems. Yet, traditional approaches fail due to parameter-dependent resonance shifts that violate smoothness assumptions. This paper develops a methodology combining pole-residue decomposition with clustering in order to handle the ordering inconsistencies inherent in vector fitting algorithms. Applied to the differential mode impedance of a traction motor, the approach demonstrates that clustering resolves the fundamental pole-residue ordering problem, enabling polynomial chaos expansion surrogates to achieve the required accuracy.