Scholarlink

<p>This research presents a systematic investigation into the optimization of process parameters in robotic Metal Inert Gas (MIG) welding to achieve substantial reductions in copper wire consumption and associated costs. The study integrates Response Surface Methodology and Artificial Neural Networks to simulate, model, and refine critical welding parameters, namely welding current (A), welding voltage (V), and wire feed rate (mm/min). Response Surface Methodology is employed to develop localized polynomial models that elucidate the nonlinear interactions between process variables and wire consumption rate, enabling the identification of optimal parameter settings with minimal experimental trials. The predictive capability and robustness of the optimized model are further validated through an Artificial Neural Network developed in MATLAB, ensuring high accuracy in process simulation. Experimental results confirm a 22.38% reduction in wire consumption, translating into a 23.83% decrease in copper wire costs, without compromising weld quality. Beyond the immediate economic benefits, the optimized parameters significantly enhance the operational efficiency of catalytic converter welding, offering a scalable framework for sustainable manufacturing practices in the automotive sector. The combined use of statistical optimization and intelligent modeling underscores the potential for integrating data-driven approaches into advanced manufacturing for cost-efficient, high-quality production.</p> Graphical abstract <p></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

Effect of welding parameters on reducing copper wire consumption costs in robotic welding using response surface methodology and artificial neural networks

  • 期刊论文

  • 发表日期：

    2025年9月11日

查看全文
  • Janarthanam Vijayanand,
  • Vaddi Seshagiri Rao,
  • K. M. B. Karthikeyan,
  • J. Hemanandh,
  • Praveen Barmavatu

摘要

This research presents a systematic investigation into the optimization of process parameters in robotic Metal Inert Gas (MIG) welding to achieve substantial reductions in copper wire consumption and associated costs. The study integrates Response Surface Methodology and Artificial Neural Networks to simulate, model, and refine critical welding parameters, namely welding current (A), welding voltage (V), and wire feed rate (mm/min). Response Surface Methodology is employed to develop localized polynomial models that elucidate the nonlinear interactions between process variables and wire consumption rate, enabling the identification of optimal parameter settings with minimal experimental trials. The predictive capability and robustness of the optimized model are further validated through an Artificial Neural Network developed in MATLAB, ensuring high accuracy in process simulation. Experimental results confirm a 22.38% reduction in wire consumption, translating into a 23.83% decrease in copper wire costs, without compromising weld quality. Beyond the immediate economic benefits, the optimized parameters significantly enhance the operational efficiency of catalytic converter welding, offering a scalable framework for sustainable manufacturing practices in the automotive sector. The combined use of statistical optimization and intelligent modeling underscores the potential for integrating data-driven approaches into advanced manufacturing for cost-efficient, high-quality production.

Graphical abstract