摘要
基于深入掌握电力用户群体特征,提供精准的电力服务,就创建了基于大数据的电力用户地址库。首先,分析售电体积累用户社会属性和用电行为等大数据,创建用户之间相似度权重模型;之后,将用户-标签作为基础的二元用户网络相似群体识别方法,并且利用群体分析得到群属性与典型负荷特征,对新入网用户用电行为进行预测。最后,对方法分析测试。通过测试结果表示,利用分析标签数据,方便发现群体中的重要用户,以此能够为互联网售电体开展个性化用电服务与增殖服务推荐提供支持,从而提高电力服务质量。
Based on in⁃depth understanding of the characteristics of power user groups and providing accurate power services,a power user address database based on big data is created.First,the paper analyzes the big data such as social attributes and electricity consumption behavior of the accumulated volume of electricity sales,and creates the similarity weight model between users;then,the binary user network similar group identification method based on user tag is used,and the group attributes and typical load characteristics are obtained by group analysis,and the electricity consumption behavior of new users is predicted.Finally,the method is analyzed and tested.According to the test results,it is convenient to find the important users in the group by analyzing the tag data,so as to provide support for Internet power sellers to carry out personalized power consumption service and proliferation service recommendation,so as to improve the quality of power service.
作者
陈超
谢辉
杨劲峰
CHEN Chao;XIE Hui;YANG Jin⁃feng(China Southern Power Grid Co.,Ltd.,Guangzhou 510623,China)
出处
《电子设计工程》
2020年第14期154-157,162,共5页
Electronic Design Engineering
关键词
大数据
电力用户
地址库
群体识别
big data
power users
address base
group identification