基于交叉域分析的能源大数据中心数据质量评估方法  

Data Quality Evaluation Method of Energy Big Data Center Based on Cross Domain Analysis

作　　者：徐厚东[1] 李赋欣[1] 刘建华 林茂 XU Houdong;LI Fuxin;LIU Jianhua;LIN Mao(State Grid Sichuan Electric Power Company,Chengdu,Sichuan 610095,China;State Grid Ziyang Electric Power Supply Company,Ziyang,Sichuan 641300,China)

机构地区：[1]国网四川省电力公司,四川成都610095 [2]国网资阳供电公司,四川资阳641300

出　　处：《广东电力》2022年第12期41-45,共5页Guangdong Electric Power

摘　　要：为解决能源大数据中心多源异构数据质量评估难的问题,提出一种基于交叉域分析的能源大数据中心数据质量评估方法。首先,通过主成分分析法提取能源大数据中心的电、水、煤、气、油等数据;其次,通过模糊均值聚类方法对各类能源数据进行聚类,形成能源典型数据特征,并建立数据特征评估指标集;最后,通过能源数据交叉域分析来评估能源大数据中心数据质量。应用所提方法评估某市能源大数据中心的数据质量,评估准确率达99.32%,应用结果验证了该评估方法的有效性。In order to solve the problem of difficult evaluation on multi-source heterogeneous data quality in the energy big data center,this paper proposes a data quality evaluation method based on cross domain analysis.This method firstly extracts the data of electricity,water,coal,gas,and oil in the energy center by using principal component analysis.Secondly it uses the fuzzy mean clustering method to cluster various energy data so as to form typical data characteristics of energy and establish the data characteristic evaluation set.indicator set.On this basis,the energy center data quality is evaluated by means of energy data cross domain analysis.Finally,the method is applied in a city s energy big data center for verification.The evaluation accuracy rate reaches 99.32%,which verifies the effectiveness of the proposed evaluation method.

关 键 词：交叉域分析 能源大数据中心 数据质量评估 主成分分析法 近邻聚类 

分 类 号：TM73[电气工程—电力系统及自动化]