Abstract: Around the accuracy of the case retrieval in the product configuration design, this paper puts forward a new case retrieval scheme in accordance with the conception of correlation function in extenics. According to the customization demand and the description of the product attributes, the attributes are classified to four kinds and the computing method of the attribute distance is presented based on the correlation function in extenics. The attribute weight is calculated by AHP（Analytical Hierarchy Process）. It gives the similarity calculation model and comes up with a specific example to prove the validity of the model.

Key words: Analytical Hierarchy Process（AHP）, similarity, extenics, correlation function

摘要： 围绕产品配置设计中实例检索的准确性问题，依据可拓学中关联函数的概念，提出了一种新的实例检索方案。根据客户定制需求和产品实例对属性的描述方式对属性进行分类，在可拓学关联函数的基础上分别定义它们的相似度的计算方法，根据AHP（Analytical Hierarchy Process）计算属性的权重比，给出客户需求与产品实例的相似度计算模型，并通过一个具体例子证明模型的有效性。

关键词: 层次分析法（AHP）, 相似度, 可拓学, 关联函数