基于C-均值聚类算法的列车故障诊断系统  

A Train Fault Diagnosis System Based on C-Means Clustering

作　　者：石华 毕越宽 

机构地区：[1]唐山轨道客车有限责任公司产品技术研究中心,河北唐山063035 [2]唐山轨道客车有限责任公司转向架新厂,河北唐山063035

出　　处：《机车电传动》2014年第1期94-96,共3页Electric Drive for Locomotives

摘　　要：介绍了2动1拖内燃电传动动车组网络拓扑结构、列车故障诊断系统,故障分类等级及应用C-均值聚类算法准确识别列车故障评估方法,阐述列车诊断系统收集并处理故障信息后,根据聚类核心及数据样本形成平方和聚类准则,对数据样本进行决策分析并输出诊断结果。人机接口设备可存储并及时显示列车故障。列车故障诊断系统增强了动车组运行的可靠性,减少维护人员的工作量,有助于列车长期稳定运行。Network topological structure of DEMUs with two motor cars and one trailer, fault diagnosis system, fault classification and fault identification by C-Means Clustering Algorithm were introduced. The fault diagnosis system collected and processed information, then, according to the clustering core and the data sample, it produced the sum of square clustering criteria, last, it made decision analysis and output diagnostic results. Human machine interface could show the DEMUs fault and record it in files. The fault diagnosis system improved the reliability of the DEMUs and reduced the work of maintenance staff, promoting the train long-term stable running.

关 键 词：故障分类 诊断 C-均值聚类 网络拓扑 内燃电传动 内燃动车组 

分 类 号：U266.1[机械工程—车辆工程]