基于数据挖掘的电采暖电量预测及应用  

Prediction and Application of Electric Heating Electricity Based on Data Mining

作　　者：陈广宇[1] 袁绍军[1] 夏革非[1] 王宏亮 张华东[1] 陈东洋 CHEN Guangyu;YUAN Shaojun;XIA Gefei;WANG Hongliang;ZHANG Huadong;CHEN Dongyang(Chengde Power Supply Company,State Grid Jibei Electric Power Co.,Ltd.,Chengde,Hebei Province,067400 China)

机构地区：[1]国网冀北电力有限公司承德供电公司,河北承德067400

出　　处：《科技资讯》2023年第23期78-82,共5页Science & Technology Information

摘　　要：为简便实现电采暖电量预测,提出了基于数据挖掘的电量预测方法。该方法对电采暖电量和众多影响因素进行灰色关联分析,筛选出与电量有强关联的因素;利用线性回归算法构建了采暖用户预测模型,根据平均采暖户和权重系数实现对电采暖电量准确快速的预测,并创建了预测分析小工具,实现了电量的可视化预测分析。对河北某地的电量实测数据进行验证,与双向LSTM网络预测结果对比,该预测方法效果较好,同时模型的参数少且计算时间短,简化了电采暖电量预测方法。In order to realize the forecasting of electric heating electricity easily,an electricity consumption forecasting method based on data mining is proposed.This method carries out the gray correlation analysis of electric heating electricity and many influence factors,and screens out factors that are strongly correlated with electricity consumption,and it uses the linear regression algorithm to build a heating user prediction model,realizes the accurate and rapid prediction of electric heating electricity according to the average heating rate and weight coefficient,and creates small predictive analysis tools to realize the visual forecasting analysis of electricity.This method is verified according to the measurement data of electricity in a certain place in Hebei Province,and compared with the prediction results of the the two-way LSTM network,it has better prediction effect with few model parameters and short calculation time,so it can simplify the prediction method of electric heating electricity.

关 键 词：电采暖 灰色关联分析 电量预测 数据挖掘 

分 类 号：TM715[电气工程—电力系统及自动化]