用户节电的大数据分析及应用  

Big Data Analysis Research of Power Saving in Consumer Side

作　　者：陈海文 王守相[1] 梁栋[2] 苏运 CHEN Haiwen;WANG Shouxiang;LIANG Dong;SU Yun(Key Laboratory of Smart Grid of Ministry of Education (Tianjin University),Nankai District,Tianjin 300072,China;State Key Laboratory of Reliability and Intelligence of Electrical Equipment (Hebei University of Technology),Beichen District,Tianjin 300401,China;Shanghai Municipal Electric Power Company of State Grid,Pudong New District,Shanghai 200122,China)

机构地区：[1]智能电网教育部重点实验室(天津大学),天津市南开区300072 [2]电工装备可靠性与智能化国家重点实验室(河北工业大学),天津市北辰区300401 [3]国网上海市电力公司,上海市浦东新区200122

出　　处：《电网技术》2019年第4期1345-1353,共9页Power System Technology

基　　金：国家高技术研究发展计划(863计划)(2015AA050203)~~

摘　　要：为解决用户在开展节电工作时面临的用电数据不透明、缺少节电指导依据等问题,提出大数据背景下基于数据挖掘的用户节电通用分析方法,并在大数据平台上予以并行化实现,设计了直观的可视化展示形式。首先依据用户用电特性通过高维聚类实现了用电群体细分,然后融合电力、气象、经济等多维度数据开展节电分析,基于用户能效综合评估确定群体内节电标杆并量化用户节电潜力,接着通过多源数据关联分析获得用户节电策略,最后,通过SparkR在大数据平台上实现了节电算法业务的并行化,基于JavaWeb MVC框架实现了分析结果的可视化展示。实际应用效果表明,所提出的节电大数据分析方法,能有效关联多源数据,实现对海量用户数据的高效分析。To solve the problems existing in user-side power saving, such as data opacity and lack of guidance, a general power-saving analysis method is put forward based on data mining under the background of big data. Then it is implemented on a big data platform in a distributed manner and a direct visualization form is designed. Firstly, electricity groups are subdivided with high-dimensional clustering method according to the characteristics of power consuming. Then power, weather, economy and other multi-dimensional data are combined to carry out power saving analysis. The benchmark power consumer is acquired based on comprehensive evaluation of energy efficiency within a certain group, so the power-saving potential can be calculated. A power-saving strategy is obtained through multi-source data correlation analysis. Finally, the power-saving algorithm and business are implemented on a big data platform. Based on JavaWeb MVC framework, the analysis results are visualized directly. Practical application results show that the proposed method can effectively correlate multi-source data and realize efficient analysis of huge amount of user data.

关 键 词：大数据 节电分析 分布式存储 并行化计算 数据可视化 

分 类 号：TM721[电气工程—电力系统及自动化]