Scholarlink

<p>Module partition is a fundamental task in the modular design of complex products. With the increasing adoption of graph-theoretic representations, community detection has become a common strategy for deriving module structures from product networks. However, the intricate coupling among complex product components and the existence of bus components conflicting with the partitioning logic often lead to unreasonable partitioning schemes. To mitigate this issue, an iterative module partition method is proposed by integrating structural entropy with the Louvain algorithm, where modularity is used as the optimization objective. Initial modules are first obtained via the Louvain algorithm. Bus components are then characterized quantitatively by evaluating the changes in modularity and structural entropy when a component is separated from its original community and treated as an independent module. Through iterative identification and adjustment, bus components unsuitable for internal modules can be detected, and the resulting partition scheme derived from community detection becomes more rational. A motorcycle engine case study is used to demonstrate the effectiveness and feasibility of the proposed method.</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

Module partition and bus components identification for complex products based on structural entropy and Louvain algorithm

  • 期刊论文

  • 发表日期：

    2026年4月10日

查看全文
  • Zhichao Wang,
  • Hui Deng,
  • Shaopeng Xia,
  • Qi Wang,
  • Wei Guo

摘要

Module partition is a fundamental task in the modular design of complex products. With the increasing adoption of graph-theoretic representations, community detection has become a common strategy for deriving module structures from product networks. However, the intricate coupling among complex product components and the existence of bus components conflicting with the partitioning logic often lead to unreasonable partitioning schemes. To mitigate this issue, an iterative module partition method is proposed by integrating structural entropy with the Louvain algorithm, where modularity is used as the optimization objective. Initial modules are first obtained via the Louvain algorithm. Bus components are then characterized quantitatively by evaluating the changes in modularity and structural entropy when a component is separated from its original community and treated as an independent module. Through iterative identification and adjustment, bus components unsuitable for internal modules can be detected, and the resulting partition scheme derived from community detection becomes more rational. A motorcycle engine case study is used to demonstrate the effectiveness and feasibility of the proposed method.