Abstract: To improve the reusability of module base, a cluster analysis method of module base was proposed. The definitions of module base, module reusability, dissimilarity of module case, dissimilarity matrix and cluster center were introduced, and the calculation methods of the dissimilarity, minimum spanning tree and cluster center were also presented. The redundant module cases were eliminated through cluster analysis of module base. A bucket of wheel loader's module base was used to demonstrate the rationality and effectiveness of proposed method. If the threshold of dissimilarity was limited to 0　03, 11 kinds of redundant cases could be eliminated from module base while the initial number was 39, and the reuse degree was raised to 39.2%. In this case, the maximum error between redundant case and the substitute case was less than 4%.

Key words: module base, reuse degree, dissimilarity, minimum spanning tree, duster analysis, bucket, product design

摘要： 为提高产品模块库的重用度，提出了一种模块库的聚类分析方法。给出模块库、模块重用度、模块实例相异度、相异矩阵和聚类中心的定义，以及相异度、最小生成树和聚类中心的计算方法。对模块库进行聚类分析来消除冗余的模块实例，从而提高了模块的重用度。以轮式装载机的铲斗模块库为例说明了该方法的合理性和有效性，在阈值为0　03时消除了模块库原来39种实例中11种冗余的模块实例，提高重用度39.2%，此时冗余模块与替代模块之间最大的误差小于4%。

关键词: 模块库, 重用度, 相异度, 最小生成树, 聚类分析, 铲斗, 产品设计