高电压技术2017，Vol.43Issue(11)：3690-3697,8.DOI:10.13336/j.1003-6520.hve.20171031026
基于大数据挖掘技术的输变电设备故障诊断方法
Fault Diagnosis Method of Transmission and Transformation Equipment Based on Big Data Mining Technology
摘要
Abstract
The traditional faulty diagnosis method of power transmission and transformation equipment has the disadvantages of being susceptible to experts' subjectivity and model's ossification.In this paper,a new method of equipment fault diagnosis based on big data mining was proposed.Key technologies of this method were introduced,including clustering algorithm of fault patterns,analysis of relevance among status parameters and fault diagnosis based on correlation matrix.The fault cases of an operation oil immersed transformer bushing in recent 10 years were used as big data mining object.The k-means clustering algorithm together with silhouette coefficient could be used to classify fault pattern.Combination of Apriori association algorithm and Tanimoto coefficient could characterize the strength of the relationship between statuses.Fault diagnosis matrix built by Pearson correlation coefficient could precisely evaluate the fault patterns,which was consistent with actual maintenance results.The results of this study show that the inherent law of the recorded data could be obtained based on big data mining,and an adaptive and more accurate device fault diagnosis could be achieved.关键词
大数据分析/故障诊断/相关性/k-means聚类算法/轮廓系数/Tanimoto系数/Apriori关联算法Key words
big data analysis, fault diagnosis, k-means clustering algorithm/silhouette coefficient/Tanimoto coefficient/Apriori association引用本文复制引用
胡军,尹立群,李振,郭丽娟,段炼,张玉波..基于大数据挖掘技术的输变电设备故障诊断方法[J].高电压技术,2017,43(11):3690-3697,8.基金项目
国家自然科学基金(51429701) （51429701）
南方电网公司科技项目(GX2014-2-0025).Project supported by National Natural Science Foundation of China (51429701),Science and Technology Project of CSG (GX2014-2-0025). （GX2014-2-0025）
