Scholarlink

<p>To advance carbon neutrality and green manufacturing, the characteristic analysis of carbon emission (CE) in manufacturing is a growing concern while obtaining ideal performance, particularly for processes with low carbon efficiency. However, the complexity and coupling of CE in typical processes pose challenges for accurate characterization and accounting. Therefore, this work proposes the CE meta-block assisted methodology for CE characteristic analysis and accounting, which is applied to reveal the correlation between CE and processing quality in laser butt welding (LBW). Firstly, the definition of the CE meta-block depending on CE rates, duration, and CE coefficients is developed. The variable and invariable CE blocks consist of meta-blocks. Secondly, taking LBW as the case study, the CE characteristic analysis and accounting are conducted based on the built monitoring system to support the system state diagnosis and process optimization. The complex CE characteristics are summarized as multi-system composition, multi-variable involvement, multi-state operation, multi-energy consumption, and multi-source carbon emissions. Thirdly, the correlation between process parameters and welding results is constructed via an ensemble of metamodels (EM) with the optimized weight coefficients. Fourthly, the Pareto solutions with high carbon efficiency in the process parameter optimization are solved by the differential evolution-genetic algorithm and validated experimentally. The effects of cooling rate on the microstructure variations of the welding bead is analyzed. Additionally, the explainable SHAP method is employed to elucidate and interpret the predictions of EM and the effects of process parameters. Results demonstrate that the carbon efficiency is influenced notably by welding speed and laser power, and the average carbon efficiency of the optimal process parameters increases by 26.2% compared to the average value of the 25 designed experiments. Compared with individual metamodels, the EM is superior in accuracy and robustness, and reduces the mean relative error of prediction by 50.3%. The presented meta-block assisted methodology facilitates the CE accounting and carbon efficiency analysis, providing a reliable basis for sustainable manufacturing practices.</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

Carbon Emission Meta-block Assisted Process Optimization for Tensile Properties and Carbon Efficiency in Laser Butt Welding

  • 期刊论文

  • 发表日期：

    2026年1月28日

查看全文
  • Jianzhao Wu,
  • Huajun Cao,
  • Congbo Li,
  • Yang Xie,
  • Jiahao Sun,
  • Chaoyong Zhang,
  • Qiuping Yi,
  • Wenhui Liu

摘要

To advance carbon neutrality and green manufacturing, the characteristic analysis of carbon emission (CE) in manufacturing is a growing concern while obtaining ideal performance, particularly for processes with low carbon efficiency. However, the complexity and coupling of CE in typical processes pose challenges for accurate characterization and accounting. Therefore, this work proposes the CE meta-block assisted methodology for CE characteristic analysis and accounting, which is applied to reveal the correlation between CE and processing quality in laser butt welding (LBW). Firstly, the definition of the CE meta-block depending on CE rates, duration, and CE coefficients is developed. The variable and invariable CE blocks consist of meta-blocks. Secondly, taking LBW as the case study, the CE characteristic analysis and accounting are conducted based on the built monitoring system to support the system state diagnosis and process optimization. The complex CE characteristics are summarized as multi-system composition, multi-variable involvement, multi-state operation, multi-energy consumption, and multi-source carbon emissions. Thirdly, the correlation between process parameters and welding results is constructed via an ensemble of metamodels (EM) with the optimized weight coefficients. Fourthly, the Pareto solutions with high carbon efficiency in the process parameter optimization are solved by the differential evolution-genetic algorithm and validated experimentally. The effects of cooling rate on the microstructure variations of the welding bead is analyzed. Additionally, the explainable SHAP method is employed to elucidate and interpret the predictions of EM and the effects of process parameters. Results demonstrate that the carbon efficiency is influenced notably by welding speed and laser power, and the average carbon efficiency of the optimal process parameters increases by 26.2% compared to the average value of the 25 designed experiments. Compared with individual metamodels, the EM is superior in accuracy and robustness, and reduces the mean relative error of prediction by 50.3%. The presented meta-block assisted methodology facilitates the CE accounting and carbon efficiency analysis, providing a reliable basis for sustainable manufacturing practices.