摘要
基于成组技术 ,提出了用零件分类编码的哈明距离函数来表达零件相似性。在此基础上 ,利用领域知识和概念性聚类的方法 ,将零件的属性分为有序属性和无序属性来表示其语义距离 ,并结合模糊表示方法对属性值进行规格化处理 ,从而根据零件之间距离的远近建立零件聚类树 ,为零件之间相似性建立层次结构 ,奠定了自适应的动态分类和工艺制作的算法基础。
Based on the GT technology, it introduces a clustering method with Hamming distance function. Using the domain knowledge and conceptual clustering, it shows the semantic distance to represent the categorical and numerical features of parts, and the fuzzy method to normalize. Then, a parts cluster tree is established, and a hierachical structure is constructed for dynamic classification and design of the process plan.
