Scholarlink

<p>Longitudinal cracks are common surface defects of continuous casting slabs. Reducing the risk of longitudinal cracks through reasonable control and optimizing continuous casting process parameters is crucial for improving slab quality and caster productivity. In this study, an intelligent optimization framework of process parameters for longitudinal cracks of continuous casting slabs integrating machine learning and an improved particle swarm optimization algorithm is designed to achieve the goal of reducing the risk of longitudinal cracks for specific steel compositions. A nonlinear mapping function is defined by constructing an optimal classification prediction model for longitudinal cracks based on process parameters to assess the tendency of crack occurrence. The dataset employed for training and testing comprises 104 longitudinal crack samples and 312 normal samples, and the developed optimal classification prediction model achieves a prediction accuracy of 94&#xa0;pct. On this basis, the improved particle swarm optimization algorithm is used to search for the optimal solution of the nonlinear mapping function, thereby realizing the intelligent optimization of the process parameters for longitudinal cracks. Results from continuous casting experiments demonstrate that the proposed intelligent optimization framework for continuous casting process parameters can reduce the probability of longitudinal crack occurrence in medium-carbon steel from 11.3 to 1.3&#xa0;pct. The framework effectively lowers the risk of longitudinal cracks and significantly improves slab surface quality.</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

Intelligent Optimization Method to Reduce Longitudinal Cracks of Continuous Casting Slabs

  • 期刊论文

  • 发表日期：

    2025年7月23日

查看全文
  • Qican Wang,
  • Liandong Zhang,
  • Zijian Wei,
  • Yonghui Cheng,
  • Man Yao,
  • Xudong Wang

摘要

Longitudinal cracks are common surface defects of continuous casting slabs. Reducing the risk of longitudinal cracks through reasonable control and optimizing continuous casting process parameters is crucial for improving slab quality and caster productivity. In this study, an intelligent optimization framework of process parameters for longitudinal cracks of continuous casting slabs integrating machine learning and an improved particle swarm optimization algorithm is designed to achieve the goal of reducing the risk of longitudinal cracks for specific steel compositions. A nonlinear mapping function is defined by constructing an optimal classification prediction model for longitudinal cracks based on process parameters to assess the tendency of crack occurrence. The dataset employed for training and testing comprises 104 longitudinal crack samples and 312 normal samples, and the developed optimal classification prediction model achieves a prediction accuracy of 94 pct. On this basis, the improved particle swarm optimization algorithm is used to search for the optimal solution of the nonlinear mapping function, thereby realizing the intelligent optimization of the process parameters for longitudinal cracks. Results from continuous casting experiments demonstrate that the proposed intelligent optimization framework for continuous casting process parameters can reduce the probability of longitudinal crack occurrence in medium-carbon steel from 11.3 to 1.3 pct. The framework effectively lowers the risk of longitudinal cracks and significantly improves slab surface quality.