给出一个基于模糊c-平均(FCM)算法的零件簇聚类分析的过程模型来描述分析过程;构造了适合于零件簇聚类分析的FCM算法,该方法考虑了零件簇对象特征之间的模糊关系和各零件对象特征聚类中心之间的距离,无需设计权重系数;通过实例进行了聚类分析,并与模糊聚类和k-平均聚类两种方法进行比较,证明该FCM算法是有效的。
A fuzzy c--means(FCM) model of part family clustering was provided. Part family clustering was analyzed by a FCM algorithm. The method considered fuzzy relations of part family object characteristics and distances among clustering centers, it is unnecessary to design power coefficients. The clustering results were compared with that from the fuzzy clustering algorithm and the k--means algorithm through an example that proves the algorithm is validity and corresponding conclusion was presented.