基于电力交易用户最优特征子集的售电套餐推荐方法  

Recommendation Method of Electricity Selling Packages Based on Optimal Feature Subset of Electricity Trading Users

作　　者：曲朝阳[1] 冯荣强 曲楠 刘耀伟 颜佳 曲翀 QU Zhaoyang;FENG Rongqiang;QU Nan;LIU Yaowei;YAN Jia;QU Chong(School of Information Engineering,Northeast Electric Power University,Jilin 132012,Jilin Province,China;Maintenance Company of State Grid Jiangsu Electric Power Company,Nanjing 210000,Jiangsu Province,China;,State Grid Jilin Electric Power Company,Jilin 132000,Jilin Province,China;Fushun Power Supply Company of State Grid Liaoning Electric Power Supply Co.,Ltd.,Fushun 113001,Liaoning Province,China)

机构地区：[1]东北电力大学信息工程学院,吉林省吉林市132012 [2]国网江苏省电力公司检修分公司,江苏省南京市210000 [3]国网吉林省电力有限公司,吉林省吉林市132000 [4]国网辽宁省电力有限公司抚川页供电公司,辽宁省抚顺市113001

出　　处：《电网技术》2018年第10期3298-3304,共7页Power System Technology

基　　金：国家自然科学基金重点项目(51437003);吉林省科技发展计划重点项目(20180201092GX);吉林省科技发展计划项目(20160623004TC)~~

摘　　要：针对电力市场用户群庞大,交易过程中售电套餐选择困难的问题,提出一种基于电力交易用户最优特征子集的售电套餐推荐方法。首先,定义了电力交易用户最优特征子集,并设计基于加权递增项目覆盖率最优子集的发现算法,合理地从海量交易用户中筛选出最优用户特征子集。然后,提出一种基于属性相关的售电套餐相似性计算方法,通过聚类和确定套餐属性权重的方式计算套餐相似度,得到套餐项目的相似度矩阵。最后,基于最优特征子集和相似度矩阵实现了售电套餐的精准推荐。实验验证表明了电力交易用户最优特征子集的有效性和所提推荐算法的准确性。Aiming at huge user groups in electricity market and difficulty in selection of electricity selling packages in trading process, a recommendation method of electricity selling packages based on optimal feature subset of electricity trading users was proposed. Firstly, the optimal feature subset of electricity trading users was defined, and optimal subset discovery algorithm based on coverage rate with weighting increase was designed. The optimal user feature subset could be screened out reasonably from mass trading users. Next, a similar computing method of electricity selling packages based on attribute correlation was proposed in form of clustering and confirming package attribute weights. Package similarity was calculated to obtain similarity matrix of package projects. At last, accurate recommendation of electricity selling packages based on the optimal feature subset and similarity matrix was realized. Experiment confirmed effectiveness of the electricity trading users＇ optimal feature subset and accuracy of the recommendation algorithm.

关 键 词：电力市场 售电套餐推荐 最优特征子集 加权递增 项目覆盖率 属性权重 

分 类 号：TM73[电气工程—电力系统及自动化]