Scholarlink

<p>To address the challenges associated with high experimental costs and low efficiency in optimizing TIG welding process parameters under external magnetic fields, this study proposes a multi-objective optimization framework that integrates numerical simulation with machine learning. A three-dimensional multi-physics model was developed using COMSOL Multiphysics, generating a dataset of 201 samples to train an XGBoost model for predicting the characteristics of the molten pool temperature field. Comparative evaluations against a BPNN model confirmed the superior predictive accuracy of the XGBoost model. Further, by incorporating a welding quality weighting formula and implementing the non-dominated sorting genetic algorithm II (NSGA-II), multi-objective optimization of the process parameters was conducted, leading to a set of optimal solutions. The findings demonstrate that the proposed methodology substantially improves the accuracy and efficiency of welding process design.</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

Numerical simulation of TIG welding under applied magnetic field and optimization of welding process parameter

  • 期刊论文

  • 发表日期：

    2026年2月9日

查看全文
  • Li Wang,
  • Cheng Peng,
  • Lei Zhang,
  • Xiao Jie Guo,
  • Zhi Wu Ke,
  • Fu Xin Lu

摘要

To address the challenges associated with high experimental costs and low efficiency in optimizing TIG welding process parameters under external magnetic fields, this study proposes a multi-objective optimization framework that integrates numerical simulation with machine learning. A three-dimensional multi-physics model was developed using COMSOL Multiphysics, generating a dataset of 201 samples to train an XGBoost model for predicting the characteristics of the molten pool temperature field. Comparative evaluations against a BPNN model confirmed the superior predictive accuracy of the XGBoost model. Further, by incorporating a welding quality weighting formula and implementing the non-dominated sorting genetic algorithm II (NSGA-II), multi-objective optimization of the process parameters was conducted, leading to a set of optimal solutions. The findings demonstrate that the proposed methodology substantially improves the accuracy and efficiency of welding process design.