Scholarlink

<p>In order to achieve high-performance casting production, it is urgent to accurately and efficiently determine the squeeze casting process parameters. Therefore, this paper proposes a squeeze casting process parameter design method based on two-stage intelligent integrated optimization and finite element (FEM). Firstly, a prediction model is constructed, combining the BP neural network with the bidirectional long short-term memory network to establish the mapping relationship between process parameters and casting properties. Subsequently, the improved love evolutionary algorithm (LLEA) is used to optimize the above (BB) model. In order to deal with multiple output indicators, the entropy weight method is introduced to convert it into a single objective function. On this basis, LLEA is used again to search for the optimal combination of process parameters. The effectiveness of the proposed method is verified by a case study, and its improvement effect in shrinkage defect control is verified by finite element analysis. The results show that the LLEA–BB framework outperforms traditional algorithms in both prediction accuracy and optimization efficiency, with an accuracy of 98.82%. The optimal parameter combination obtained improves the casting performance and reduces the shrinkage rate to 4.033%.</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

Squeeze casting process parameter design based on two-stage intelligent integrated optimization and FEM

  • 期刊论文

  • 发表日期：

    2026年1月20日

查看全文
  • Junjie Shen,
  • Deqiang He,
  • Haimeng Sun,
  • Zhenzhen Jin,
  • Jiegao Ma

摘要

In order to achieve high-performance casting production, it is urgent to accurately and efficiently determine the squeeze casting process parameters. Therefore, this paper proposes a squeeze casting process parameter design method based on two-stage intelligent integrated optimization and finite element (FEM). Firstly, a prediction model is constructed, combining the BP neural network with the bidirectional long short-term memory network to establish the mapping relationship between process parameters and casting properties. Subsequently, the improved love evolutionary algorithm (LLEA) is used to optimize the above (BB) model. In order to deal with multiple output indicators, the entropy weight method is introduced to convert it into a single objective function. On this basis, LLEA is used again to search for the optimal combination of process parameters. The effectiveness of the proposed method is verified by a case study, and its improvement effect in shrinkage defect control is verified by finite element analysis. The results show that the LLEA–BB framework outperforms traditional algorithms in both prediction accuracy and optimization efficiency, with an accuracy of 98.82%. The optimal parameter combination obtained improves the casting performance and reduces the shrinkage rate to 4.033%.