设备健康状态灰色聚类评估的一种改进方法  

An Improved Grey Clustering Assessment for Equipment Health

作　　者：吴波[1] 夏良华[1] 郑百源 杜军乐[1] 吴豫杰 

机构地区：[1]军械工程学院装备指挥与管理系 [2]75134部队保障处 [3]防空兵学院弹炮一体系

出　　处：《计算机测量与控制》2013年第10期2756-2758,共3页Computer Measurement &Control

基　　金：总装某部科研资助项目(2010SY4308002)

摘　　要：灰色聚类评估方法是按照最大灰色聚类系数原则来判定聚类对象属于某一灰类的,在评估设备健康状态的应用中,往往会出现灰色聚类系数无显著性差异的情况,这时就无法判定设备健康状态属于何种灰类;因此,提出了一种灰色聚类评估的改进方法;首先计算设备健康状态的归一化聚类系数,然后给出一个较为合理的聚类系数无显著性差异的定义,接着利用灰色关联分析的思想对灰色聚类评估模型进行改进;最后,通过对设备健康状态评估的示例分析,证实了灰色聚类评估改进方法的可行性和有效性。The method of grey clustering assessment is often used to classify the clustering object according to the principle of maximum clustering coefficient. There are not distinguished differences between grey clustering coefficients which often come forth in the application of cquipment health assessment, then the principle of maximum clustering coefficient is non effective. An improved grey clustering method is put forward based on the present theories on grey clustering. Firstly, the normalized clustering coefficient of the equipment health is formula ted. Secondly, a reasonable definition of clustering coefficient with non distinguished difference is presented. Then, the general grey clustc ring model is improved using grey correlation analysis. Lastly, the feasibility and validity of the improved grey clustering method is illustrated through the case study of equipment health assessment.

关 键 词：设备健康状态 灰色聚类评估 聚类系数 无显著性差异 灰色关联度 

分 类 号：TP206[自动化与计算机技术—检测技术与自动化装置]