Scholarlink

<p>Robust design optimization is essential for improving the performance of electromagnetic devices under uncertainties and tolerances. This paper presents a multi-disciplinary robust design optimization method for a high speed permanent magnetic synchronous motor (HSPMSM) featuring an amorphous alloy stator core and Halbach permanent magnet array. The finite element method (FEM) and thermal network modeling are used to calculate the motor performance and establish the dataset, and the radial basis function (RBF) model and extreme gradient boost (XGBoost) regressor has been compared in terms of speed, accuracy, and complexity to construct a high accuracy surrogate model, reducing the computational burden of FEM. The robust optimization under the design for six sigma (DFSS) based on the surrogate model integrating the electromagnetic, thermal, and rotor stress characteristic performance is to achieve a solution with superior performance and robustness. The optimal design result demonstrates high output power and operation efficiency, the effectiveness of the proposed method and its low computational cost are verified.</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

Robust-Oriented Multi-Discipline Design Optimization of High Speed Permanent Magnetic Synchronous Motor Based on RBF Surrogate Model

  • 期刊论文

  • 发表日期：

    2025年6月6日

查看全文
  • Chengcheng Liu,
  • Kexin Ren,
  • Youhua Wang,
  • Zhigang Zhao,
  • Fei Zhao

摘要

Robust design optimization is essential for improving the performance of electromagnetic devices under uncertainties and tolerances. This paper presents a multi-disciplinary robust design optimization method for a high speed permanent magnetic synchronous motor (HSPMSM) featuring an amorphous alloy stator core and Halbach permanent magnet array. The finite element method (FEM) and thermal network modeling are used to calculate the motor performance and establish the dataset, and the radial basis function (RBF) model and extreme gradient boost (XGBoost) regressor has been compared in terms of speed, accuracy, and complexity to construct a high accuracy surrogate model, reducing the computational burden of FEM. The robust optimization under the design for six sigma (DFSS) based on the surrogate model integrating the electromagnetic, thermal, and rotor stress characteristic performance is to achieve a solution with superior performance and robustness. The optimal design result demonstrates high output power and operation efficiency, the effectiveness of the proposed method and its low computational cost are verified.