基于大数据的配电网故障停电指标关联因素建模研究  

Research on Modeling of Correlative Factors of Fault Outage Index in Distribution Network Based on Big Data

作　　者：刘江东 吴志坚 黄振勇 吴予乐 濮实 高洁 Liu Jiangdong;Wu Zhijian;Huang Zhenyong;Wu Yule;Pu Shi;Gao Jie(Yangzhou Power Supply Branch,State Grid Jiangsu Electric Power Co.,Ltd.,Yangzhou Jiangsu 225009,China)

机构地区：[1]国网江苏省电力有限公司扬州供电分公司,江苏扬州225009

出　　处：《电气自动化》2022年第2期45-46,50,共3页Electrical Automation

摘　　要：传统配电网故障停电指标关联因素分析方法存在停电预测耗时长和准确率低等问题。为此,研究基于大数据的配电网故障停电指标关联因素模型。基于大数据建立配电网停电故障风险指标,利用层次分析法计算配电网大数据风险指标总体权重。通过数据预处理生成随机森林分类器,将配电网停电故障风险指标作为分类器输入,以OOB残差均方法实现配电网故障停电指标关联因素重要性排序,分类器输出结果即为预测到的配电网停电故障。试验结果表明:模型的电网故障覆盖率较高,说明所提方法具有理想的准确性。The traditional method of analyzing related factors of fault outage indices in distribution network has the problems of long time-consuming and low accuracy of outage prediction.For this reason,the research on the correlation factor model of the distribution network fault outage indices based on big data.The distribution network outage failure risk indices were set up based on big data,and the analytic hierarchy process was used to calculate the overall weight of the distribution network big data risk index.The random forest classifier was generated by data preprocessing,and the risk index of distribution network fault outage was used as the input of the classifier.The OOB residual mean square method was used to rank the importance of related factors of distribution network fault outage index.The output of the classifier was the predicted distribution network outage fault.The experimental results show that the fault coverage of the model is high,which shows that the proposed method has ideal accuracy.

关 键 词：电网故障 停电指标 关联因素 随机森林 泛化误差 

分 类 号：TM76[电气工程—电力系统及自动化]