摘要
为掌握代购电用户月用电量的影响因素,提出一种代购电用户的画像方法。从周日类型、节假日、气温、电价敏感性维度构建特征指标,分别将卡方值和提升度指标用于周日类型和节假日敏感性分析,克服了传统灵敏度指标难以分析分类指标灵敏度的问题;进而通过指标的标签化和归类,获得代购电用户分类的多角度描绘。上海某地的实例应用验证了所提方法的可行性,可进一步用于代购电量的分类预测方法设计。
In order to grasp the factors affecting the monthly electricity consumption of electricity purchasing on behalf of customers,a profiling method for electricity purchasing service was proposed.Constructing characteristic indicators from the sensitivity dimensions of Sunday type,holidays,temperature,and electricity price,respectively using chi square values and improvement degree indicators for Sunday type and holiday sensitivity analysis,the problem of traditional sensitivity indicators being difficultly used to analyze the sensitivity of classification indicators was overcome;furthermore,through the labeling and classification of indicators,a multi perspective description of the classification of purchasing electricity users can be obtained.The feasibility of the proposed method was verified through an example application at a certain location in Shanghai,and the portrait results showed the classification of purchasing electricity users from the perspective of influencing factors of monthly electricity consumption.The result can be further used for the design of classification and prediction methods for purchasing electricity.
作者
韩跃峻
张洋
陆斌
杜漾腾
辛洁晴
Han Yuejun;Zhang Yang;Lu Bin;Du Yangteng;Xin Jieqing(Shibei Electricity Supply Company,SMEPC,Shanghai 200072,China;Dept.of EE,Shanghai JiaoTong University,Shanghai 200240,China)
出处
《电气自动化》
2023年第6期7-9,共3页
Electrical Automation
基金
国网上海市电力公司项目资助(520914220005)。
关键词
代购电
用户画像
电量预测
分类变量
关联分析
电力市场
electricity purchasing service
user portrait
electricity forecasting
categorical variables
association analysis
power market
