摘要
窃电的多样性与反窃电成本的逐步提高,给供电企业的反窃电工作带来了巨大的挑战。介绍了一种综合贝叶斯分类算法与聚类算法的分析研究方法,通过对与窃电相关的事件进行特征值赋予,并根据关联关系编制了一套窃电概率值评价体系,实现了对窃电嫌疑用户进行窃电概率评估。该方法可精准定位窃电现场,提升窃电用户查处的效率,改善社会用电环境。
The diversity of electricity stealing and the gradual improvement of the anti electricity stealing cost have brought great challenge to the power supply enterprises anti stealing work.This paper introduces an analysis and research method of comprehensive Bias classification algorithm and clustering algorithm.By assigning characteristics to the related incidents of stealing electricity,works out a set of evaluation system of stealing electricity probability value based on the relationship,and realizes the probability of stealing electricity for suspected users.The method can accurately locate the field of electricity stealing,can improve the efficiency of investigation for the users of electricity stealing and improve the environment of social electricity consumption.
作者
赵俊鹏
吕孟玉
张洋瑞
张冰玉
李晶
段子荷
ZHAO Junpeng;LV Mengyu;ZHANG Yangrui;ZHANG Bingyu;LI Jing;DUAN Zihe(State Grid Hebei Electric Power Research Institute,Shijiazhuang 050021,China;State Grid Hebei Electric Power Corporation Cangzhou Power Supply Branch,He Bei Cangzhou 061000,China)
出处
《河北电力技术》
2019年第1期24-25,29,共3页
Hebei Electric Power
关键词
用电信息采集系统
聚类算法
数据权值
风险评价体系
数据推送
贝叶斯分类方法
electricity information collection system
clustering algorithm
data weight
risk evaluation system
data push
Bias classification method