作者：杨婧, 辛明勇, 欧家祥, 王俊融, 宋强

作者单位：贵州电网有限责任公司电力科学研究院, 贵州 贵阳, 550000

关键词：配电网;线损;关联分析;大数据

摘要：

针对当前配电网输电线路损耗异常无法溯源且定位难的问题，基于计量自动化系统采集的数据，通过对站、线、变、户基础数据的治理，采用自动最优聚类算法对用户用电行为分类，采用随机森林建立各类线损之间的关联关系模型，构建配电网线路损耗和台区损耗分析与定位方法，并开发基于线损异常精确定位的计量自动化运维平台。通过对贵州省某供电局辖区2 516个用户的数据进行分析和实验验证，该文所提出的线损分析与定位方法能对配电网线损异常进行溯源和精确定位。

Distribution network line loss location and evaluation method study based on big data
YANG Jing, XIN Mingyong, OU Jiaxiang, WANG Junrong, SONG Qiang
Guizhou Power Grid Corp Electric Power Science Research Institute, Guiyang 550000, China
Abstract: Aiming at the problem that the transmission line loss of the distribution network is not traceable and difficult to locate, the article is based on the data collected by the measurement automation system. Through the management of the station, line, transformer and household basic data, the automatic optimal clustering algorithm is used for the user. The classification of electrical behaviors is based on random forests to establish the correlation model between various types of line losses. The method of analyzing and locating the line loss and the area loss of the distribution network is constructed, and the automatic operation and maintenance platform based on the line loss anomaly positioning is developed. Through the analysis and experimental verification of the data of 2 516 users in a power supply bureau in Guizhou Province, the line loss analysis and location method proposed can trace and accurately locate the line loss anomaly of the distribution network.
Keywords: distribution network;line loss;correlation analysis;big data
2019, 45(7):19-24  收稿日期: 2019-04-01;收到修改稿日期: 2019-05-14
基金项目:
作者简介: 杨婧(1988-),女,湖南邵阳市人,工程师,硕士,研究方向为计量自动化及电网节能降损
参考文献
[1] CHATTERJEE S, ARCHANA V, SURESH K, et al. Detection of non-technical losses using advanced metering infrastructure and deep recurrent neural networks[C]//IEEE International Conference on Environment and Electrical Engineering, 2017.
[2] 杨柳, 罗璇, 肖宝辉, 等. 配电网线损在线监测系统与计算分析研究[J]. 自动化与仪器仪表, 2019(1):25-28
[3] 杨悦辉. 计量自动化系统在计量管理工作中的应用[J]. 计量与测试技术, 2013(4):16-37
[4] 刘飞轮. 供电企业电力营销中的电能计量自动化系统应用分析[J]. 硅谷, 2013, 6(20):95, 91
[5] 汪司珂, 汪应春, 郭雨, 等. 同期线损故障排查关键技术研究与应用[J]. 仪表技术, 2017(12):11-14
[6] LI H, LIANG X, LU R, et al. EPPDR:An efficient privacy-preserving demand response scheme with adaptive key evolution in smart grid[J]. IEEE Transactions on Parallel and Distributed Systems, 2014, 25(8):2053-2064
[7] 徐茹枝, 王宇飞. 粒子群优化的支持向量回归机计算配电网理论线损方法[J]. 电力自动化设备, 2012(5):86-89
[8] HUANG S, WU Q, CHENG L, et al. Optimal reconfiguration-based dynamic tariff for congestion management and line loss reduction in distribution networks[J]. IEEE Transactions on Smart Grid, 2016, 7(3):1295-1303
[9] 高志芳, 邹香香, 汪祺. 智能计量监控在反窃电工作中的应用[J]. 决策与信息, 2014(36):181
[10] NOURAI A, KOGAN V I, SCHAFER C M. Load leveling reduces t&d line losses[J]. IEEE Transactions on Power Delivery, 2008, 23(4):2168-2173
[11] DANGAR B, JOSHI S K. Notice of violation of IEEE publication principles electricity theft detection techniques for metered power consumer in GUVNL, GUJARAT, INDIA[C]//Clemson University Power Systems Conference (PSC), 2015
[12] 康宁宁, 李川, 曾虎, 等. 采用FCM聚类与改进SVR模型的窃电行为检测[J]. 电子测量与仪器学报, 2017, 31(12):2023-2029
[13] 何喜玲, 韩婷, 杜佳, 等. 变电所电量对比系统分析[J]. 现代电子技术, 2016, 39(12):41-44, 47
[14] 翟金凤, 孙立博, 鲁凯, 等. 基于抽样和两级CBF的长流识别算法[J]. 中国测试, 2018, 44(7):105-109
[15] 刘耀杰, 刘独玉. 基于不平衡数据集的改进随机森林算法研究[J]. 计算机技术与发展, 2019(6):1-7