摘要
为落实国家电网公司营销计量大数据分析应用工作安排,实现营销计量业务数据的价值挖掘,该文实现了一种基于数据挖掘方法的智能电能表故障研究方案,通过收集各信息化系统有关电能表的检测及故障数据,建立智能电能表故障库。通过常规的下钻统计初步获取各维度下电能表的故障分布情况,并通过APRIORI关联分析算法和k-means聚类算法,分析电能表的家族性缺陷,为供应商评价及厂家性能改进提供依据。建立异常故障预警功能,及时发现电能表的异常故障,并通过故障树的形式快速定位电能表故障,实现故障的快速定位及维修。
In order to implement the national grid company’s marketing metering data analysis and application work arrangement,the value mining of marketing metering business data is realized. This paper implements an intelligent watt-hour meter fault study scheme based on data mining meth ods,establish intelligent watt-hour meter fault library by collecting the information system about the detection of watt-hour meter and failure data; Obtained through conventional drilling statistics preliminary fault distribution of each dimension of watt-hour meter and through the APRIORI correlation analysis algorithm and k-means clustering algorithm,analysis of familial defects of watt-hour meter,provide the basis for supplier evaluation and manufacturer performance improvements. The abnormal failure of the electric energy meter was found in a timely manner,and the fault of the power meter was quickly identified through fault tree,as the result,the fault was quickly located and repaired.
作者
朱东升
李天阳
徐石明
ZHU Dong-sheng;LI Tian-yang;XU Shi-ming(NARI Group Corporation(State Grid Electric Power Research Institute),Nanjing 210003,China)
出处
《自动化与仪表》
2018年第5期14-18,共5页
Automation & Instrumentation
基金
国家电网公司科技资助项目(5216A01600VZ)