摘要
本文采用大数据分析技术,针对违约窃电行为、表计电量采集数据频繁缺失等原因导致的四类线损进行分析挖掘,构建了台区线损综合治理排查模型。应用层次聚类分析法对四类线损问题关键变量进行识别,应用皮尔逊积距相关系数法和神经网络算法对疑似用户和疑似问题进行判断。将该模型部署于移动作业应用平台,实现对线损疑似问题的现场核查,完成线损治理工作的闭环管理。
This paper adopts big data analysis technology to analyze and mine four types of line losses caused by power theft,frequent lack of metered electricity collection data and other reasons to build a model of comprehensive line loss management and investigation in supply area.Hierarchical clustering analysis is used to identify key variables of four types of line losses,and the Pearson product-moment correlation coefficient and neural network algorithm are used to judge the suspected users and problems.The model is deployed on a platform for mobile operation to investigate on-site suspected line losses and achieves closed-loop management of line loss management.
作者
陈未
潘越
符煌莹
沈华胄
岑颖蓓
CHEN Wei;PAN Yue;FU Huangying;SHEN Huazhou;CEN Yingbei(State Grid Xiangshan Power Supply Company,Xiangshan Zhejiang 315700,China)
出处
《浙江电力》
2021年第5期60-65,共6页
Zhejiang Electric Power
基金
国网浙江省电力有限公司营销专项资金(6211XT20002X)。
关键词
线损
大数据
层次积累分析法
皮尔逊积距相关系数法
神经网络算法
line loss
big data
hierarchical clustering analysis
Pearson product-moment correlation coefficient
neural network algorithm
