基于加权灰色关联分析与模糊C-均值聚类的柴油机磨损模式识别  

Wear pattern recognition of diesel engine based on weighted grey relevancy analysis and fuzzy C-mean valued clustering

作　　者：张晓丽[1] 李鑫[1] 郭智春[1] 

机构地区：[1]大连理工大学精密与特种加工教育部重点实验室,辽宁大连116024

出　　处：《机械设计》2008年第10期46-48,共3页Journal of Machine Design

基　　金：国家863主题资助项目(2002AA412020);辽宁省教育厅高校科研资助项目(2006B031)

摘　　要：将加权灰色关联分析与模糊C-均值聚类(fuzzy center-means clustering)算法相结合应用到柴油机磨损模式评价体系中,以Bezdek准则和模糊熵值为度量指标加强全局搜索能力并确定了加权指数m值,通过聚类中心和归一化的标准向量建立了磨损模式的模糊分析模型。对实际采集的20个样本进行模糊聚类分析,并与贴近度算法所得结果相比较,证明加权模糊C-均值聚类算法对柴油机磨损模式的识别是准确、有效的。Let the combination of weighted grey relevancy analysis and fuzzy C - mean valued clustering algorithm be applied to the wear pattern evaluation system of diesel engine, taking the Bezdek criterion and the fuzzy entropy value as the measuring target to enhance the search ability in overall situation and determined the weighted exponent value m, the wear patterned fuzzy analysis model was established by means of the clustering centre and the normalized standard vector. To carrying out fuzzy clustering analysis on 20 actually collected specimens and compared with the result obtained from the closeness algorithm, it was proved that the weighted fuzzy C - mean valued clustering algorithm was accurate and effective on the wear patterned recognition of diesel engines.

关 键 词：C-均值聚类 磨损模式 模糊聚类 模式识别 

分 类 号：TP39[自动化与计算机技术—计算机应用技术] O159[自动化与计算机技术—计算机科学与技术]