基于大数据平台的中压配电网精准损耗分析  

Accurate Loss Analysis of Medium Voltage Distribution Network Based on Big Data Platform

作　　者：黄园芳 谭涛 郑世明 徐达艺 王晓明 王星华[2] HUANG Yuanfang;TAN Tao;ZHENG Shiming;XU Dayi;WANG Xiaoming;WANG Xinghua(Zhanjiang Power Supply Bureau of Guangdong Power Grid Co.,Ltd.,Zhanjiang,Guangdong 524005,China;College of Automation,Guangdong University of Technology,Guangzhou,Guangdong 510006,China)

机构地区：[1]广东电网有限责任公司湛江供电局,广东湛江524005 [2]广东工业大学自动化学院,广东广州510006

出　　处：《广东电力》2021年第9期11-17,共7页Guangdong Electric Power

基　　金：广东电网有限责任公司科技项目(030800KK52180032);国家自然科学基金项目(61903091)。

摘　　要：大数据平台为中低压配电网线损精细化分析提供了新的方法和基础,但数据缺失会影响计算精度和新方法的实用性。为此,提出基于大数据平台的前推回代精细化损耗分析框架;其次,针对部分缺失和全部缺失2种情况,提出基于粒子群优化的模糊C均值聚类的负荷曲线聚类方法,并利用聚类中心和隶属度来填充缺失值;最后利用实际数据进行仿真验证。仿真结果表明:数据充足时前推回代法可以得到精度更高的计算结果,改进后的模糊C均值算法能更有效地拟合真实数据,有效解决数据缺失的问题。The big data platform provides a new method and basis for the refined analysis of the line loss of the medium and low voltage distribution network,but the lack of data seriously affects the calculation accuracy and the practicability of the method.For this reason,this paper proposes a forward-backward refined loss analysis framework based on a big data platform.Aiming at two cases of partial and all missing,it proposes a load curve clustering method based on fuzzy C-means clustering of particle swarm optimization,which uses the clustering center and the degree of membership to fill in the missing values.Finally,the actual data is used for simulation verification.The simulation results show that when the data is sufficient,the forward-backward substitution method can obtain the calculation results with higher precision,and the improved fuzzy C-means algorithm can more effectively fit the real data and solve the problem of data missing.

关 键 词：电力大数据 分布式延迟 进化计算 模糊聚类 负荷数据填充 前推回代法 

分 类 号：TM73[电气工程—电力系统及自动化] TM727