电力客户准确定位与立体画像多维构建研究

Research on Power Customer Accurate Positioning and Multidimensional Construction of Three-dimensional Portrait

针对传统供电企业电网用户画像构建存在信息覆盖少、准确率低的问题,提出基于电力大数据的电网用户立体画像构建方法,通过IV值特征筛选法甄选电力用户的特征指标,应用多尺度逐层分析法提取指标特征量,根据聚类中心距离进行信息分析,构建电力用户立体画像特征域的分类结构模型,提升画像特征定位精度,实现准确营销和个性化服务。仿真结果表明电力用户立体画像构建方式能有效提高电力用户多维特征分类精度,有利于向电力用户提供个性化服务。

Aiming at the problems of less information coverage and low accuracy in the construction of power grid user portraits in traditional power supply enterprises,a method of constructing power grid user stereo portraits based on power big data is proposed.The IV value feature screening method is used to select the feature indexes of power users.The multi-scale layer-by-layer analysis method is used to extract the feature quantity of the index,and the information analysis is carried out according to the distance of the cluster center.The feature domain of the power user stereo portrait is used to improve the positioning accuracy of the portrait features and realize accurate marketing and personalized services.The simulation results show that the proposed stereoscopic portrait construction method of power users can effectively improve the multidimensional feature classification accuracy of power users,and it is beneficial to provide personalized service to power users.