电力系统二次设备缺陷数据挖掘与分析方法研究

为了准确挖掘与分析电力系统二次设备缺陷数据，判断电网二次设备运行状态，提出基于缺陷数据挖掘的电力系统二次设备缺陷分析方法。基于层次聚类算法，获取电力系统二次设备缺陷数据挖掘，以原子聚类、原子簇合并、基于层次聚类算法的缺陷数据识别三步骤，将缺陷数据设成基于XGBoost的二次设备缺陷分类模型的输入数据，设置缺陷特征指标集、缺陷级别标签后，有效挖掘电力系统二次设备缺陷数据，然后通过XGBoost模型，实现二次设备缺陷级别识别。研究结果表明，所提方法对二次设备缺陷数据的挖掘结果和实际缺陷数据样本数一致，对二次设备缺陷级别识别后，识别结果的准确率、召回率、F₁值均高达0.99，且识别耗时低于400 ms，具备使用价值。

Research on data mining and analysis method of secondary equipment defects in power system

In order to accurately analyze the defect data of secondary equipment in power system and judge the operation status of secondary equipment in power grid，a defect analysis method of secondary equipment in power system based on defect data mining was proposed.Based on the hierarchical clustering algorithm，the defect data mining of secondary equipment in power system was obtained.The defect data is set as the input data of the secondary equipment defect classification model based on XGBoost by three steps of atomic clustering，atomic cluster merging and defect data recognition based on hierarchical clustering algorithm，after setting defect characteristic index set and defect level label，the defect data of secondary equipment in power system can be effectively mined；secondly，through XGBoost model，the defect level of secondary equipment was identified.The results showed that:the proposed method for secondary equipment defect data mining results are consistent with the actual number of defect data samples，after the secondary equipment defect level recognition，the accuracy，recall rate and F₁ value of the recognition results are as high as 0.99，and the recognition time is less than 400 ms，which has the use value.