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

传统输变电设备故障诊断方法具有易受专家意见主观影响以及模型固化的缺点。为此,提出了基于大数据挖掘技术的设备故障诊断方法,介绍了设备故障模式聚类算法、状态参量相关性分析算法以及基于相关性矩阵的故障诊断方法等基于大数据分析的设备故障诊断关键技术,并采用某电网公司500 kV电压等级油浸式变压器套管近10 a故障记录数据作为数据挖掘案例进行了实证。研究结果表明:k-means聚类算法配合轮廓系数能准确得得出故障分类模式;Apriori关联算法配合Tanimoto系数可用于评估状态参量之间的强弱关系;基于皮尔逊相关系数构造故障诊断矩阵能够准确判断出与实际运行维护试验结果相符合的设备故障模式。因此,基于大数据挖掘技术的输变电设备故障诊断方法能够有效挖掘出设备状态记录数据内在的规律,实现具有数据自适应性的、更加准确的设备故障诊断。

The traditional faulty diagnosis method of power transmission and transformation equipment has the disad- vantages 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 clus- tering 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. Com- bination 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.