基于电力大数据的居民用电行为分析与用户画像构建  

Analysis of residential electricity consumption behavior and construction of user profile based on power big data

作　　者：王淇 程江洲[1] 张川[3] WANG Qi;CHENG Jiangzhou;ZHANG Chuan(Faculty of Electrical and New Energy,Three Gorges University,Yichang 443000,China;State Grid Hongshan Power Supply Center of Wuhan Power Supply Company,Wuhan 430070,China;State Grid Yichang Power Supply Company Transmission and Inspection Branch,Yichang 443000,China)

机构地区：[1]三峡大学电气与新能源学院,湖北宜昌443000 [2]国网武汉供电公司洪山供电中心,湖北武汉430070 [3]国网宜昌供电公司输电运检分公司,湖北宜昌443000

出　　处：《电子设计工程》2025年第6期57-62,共6页Electronic Design Engineering

摘　　要：为了精准分析居民用电行为模式,优化电力资源配置与服务质量,提出基于电力大数据的居民用电行为分析与用户画像构建方法。利用电力大数据技术获取居民用电行为数据,计算居民用电行为的直观描述和比值描述指标,根据居民用电行为生成用户行为标签,通过多个行为标签的融合,得出用户画像的构建结果。通过性能测试实验可知,与传统方法相比,优化设计方法的居民用电量和负荷率的分析误差分别减小0.33 kW·h和3.0%,且构建用户画像的偏离度明显降低。In order to accurately analyze residential electricity consumption behavior patterns,optimize power resource allocation and service quality,a method for analyzing residential electricity consumption behavior and constructing user profiles based on power big data is proposed.Using power big data technology to obtain residential electricity consumption behavior data,calculating intuitive descriptions and ratio description indicators of residential electricity consumption behavior,generating user behavior labels based on residential electricity consumption behavior,and constructing user profiles through the fusion of multiple behavior labels.The conclusion drawn from performance testing experiments is that compared with traditional methods,the optimization design method reduces the analysis errors of residential electricity consumption and load rate by 0.33 kW·h and 3.0%,respectively,and significantly reduces the deviation of constructing user profiles.

关 键 词：电力大数据 居民用电 行为分析 用户画像 画像构建 

分 类 号：TN249[电子电信—物理电子学]