基于改进ID3决策树的停电敏感用户辨识方法  

Power failure sensitive user identification method for power failure based on an improved ID3 decision tree algorithm

作　　者：陈丽光 何绍洋 俞晓峰 钟永城 张中超 CHEN Li-guang;HE Shao-yang;YU Xiao-feng;ZHONG Yong-cheng;ZHANG Zhong-chao(Guangdong Power Grid Co.,Ltd.,Heyuan Power Supply Bureau,Heyuan 517000,Guangdong Province,China)

机构地区：[1]广东电网有限责任公司河源供电局,广东河源517000

出　　处：《信息技术》2020年第5期49-53,共5页Information Technology

基　　金：中国南方电网有限责任公司科技项目(03160020180-30304DN00002)。

摘　　要：停电敏感用户辨识是用户用电信息管理的重要内容,对提升供电服务水平具有重要作用。近年来随着我国电力监管日趋严格,电网企业不断加大对客户停电的管理力度。停电敏感用户的辨识也成为电网企业研究的重点。针对传统的辨识方法对实际应用中样本属性缺失考虑不足的实际问题,提出了改进ID3决策树分类算法。以该算法为核心设计了停电敏感用户辨识方法,实现了对海量用户的自动辨识。最后,基于地区电网的实际数据构造的算例表明,与传统忽略缺省项的方法相比,该方法能有效提升辨识覆盖率和准确率,对推动电网服务水平提升具有促进作用。Power failure sensitive user identification is an important part of user electricity information management and plays an important role in improving power supply service level.In recent years,with the increasingly strict supervision of electricity power in China,power grid enterprises have been strengthe-ning the users’ power failure management.Therefore,there is an urgent need for power failure sensitive user identification method for power grid.An improved ID3 decision tree classification algorithm is proposed to solve the problem that the traditional identification method fails to consider the lack of sample attributes in practical application.Based on this algorithm,a power failure sensitive user identification method is designed,which realizes the automatic identification of massive users.Finally,a case study based on the actual data of a regional power grid shows that,compared with the traditional method of ignoring default terms,this method can effectively improve the identification coverage and accuracy and has a promoting effect on the improvement of power supply service.

关 键 词：停电敏感用户 供电服务 决策树算法 属性缺失 

分 类 号：TM743[电气工程—电力系统及自动化]