可处理精确数据和模糊数据的实例匹配算法  

A Case-Matching Algorithm Suitable for Non-Fuzzy/Fuzzy Data

作　　者：李玲玲[1] 王建楠[1] 赵军[1] 李志刚[1] 

机构地区：[1]河北工业大学电气与自动化学院,天津300130

出　　处：《低压电器》2006年第10期8-11,共4页Low Voltage Apparatus

基　　金：河北省博士基金(05547003D-3);河北省自然科学基金(602069)

摘　　要：基于实例的推理,其关键是要计算作为当前待求解问题的实例(称为靶实例)与已有实例(称为基实例)的相似度。靶实例的数值型特征的值有些是精确的,有些是模糊的,这两类数据可能并存。为了在该情况下进行实例匹配,基于隶属度和贴近度的概念,构造了靶实例与基实例的模糊相似度函数,从而以统一形式同时处理上述两类数据。以继电器产品的设计为例,验证了有关算法的正确性。Case-based reasoning takes the calculation of the similarity between the cases to be solved （casetarget） and the case-base. Some of the numerical eigenvalue of case-target are exact value and some are fuzzy value. In order to match the case in this situation, a fuzzy similarity function between case-target and Case-base was constructed based on the conception of membership degree and closeness degree. It can deal with the two kinds of data above in the same form. Taking the relay products as an example, the validity of this algorithm was proved.

关 键 词：基于实例的推理 实例匹配 数值型特征 模糊相似度 

分 类 号：TM501.4[电气工程—电器]