基于电压时空聚合曲线的乡镇配电台区户变关系识别方法  

Identification Method for Relationship Between Residential Electricity Meters and Distribution Transformers in Township Distribution Station Area Based on Voltage Spatial-temporal Aggregation Curves

作　　者：唐冬来 李强 陈永东 张捷 刘友波[3] 宋卫平 TANG Donglai;LI Qiang;CHEN Yongdong;ZHANG Jie;LIU Youbo;SONG Weiping(Aostar Information Technology Co.,Ltd.,Chengdu 610074,China;State Grid Information&Telecommunication Group Co.,Ltd.,Beijing 102211,China;School of Electrical Engineering,Sichuan University,Chengdu 610065,China)

机构地区：[1]四川中电启明星信息技术有限公司,四川省成都市610074 [2]国网信息通信产业集团有限公司,北京市102211 [3]四川大学电气工程学院,四川省成都市610065

出　　处：《电力系统自动化》2023年第19期86-96,共11页Automation of Electric Power Systems

基　　金：四川省科技计划资助项目(2021GFW0021)

摘　　要：户变关系是指用户电表与配电变压器之间的连接关系,是智能微电网的数据基础。乡镇配电台区供电半径大,受台区分布式能源并网影响,用户电表的电压变化规律与配电台区内多个电源电压变化相关,严重影响了电压相关性户变识别的准确率。因此,考虑分布式能源并网、电源到用户电表低压线路长度的影响因素,采用电压时空聚合方法计算用户电表理论电压波动曲线,并将其与用户电表的真实电压波动曲线进行比较,从而准确识别户变关系。所提方法在中国某地区乡镇示范配电台区进行了验证,户变关系识别准确率为99.8%。结果表明,所提方法较自适应分段聚合近似、关联卷积、密度拓扑聚类等主流的户变关系识别方法具有更好的识别效果。The user-transformer relationship refers to the connection relationship between residential electricity meters and distribution transformers,which is the data foundation of smart microgrid.The power supply radius of the township distribution station area is large,and due to the influence of the distributed energy grid integration in the station area,the voltage variation pattern of the residential electricity meter is related to the voltage changes of multiple power sources in the distribution station area,which seriously affects the accuracy of voltage related user-transformer identification.Therefore,considering the factors affecting the grid connection of distributed energy and the length of the low-voltage line from the power supply to the residential electricity meter,the voltage spatiotemporal aggregation method is used to calculate the theoretical voltage fluctuation curve of residential electricity meters.And it is compared with the actual voltage fluctuation curve of residential electricity meters to identify accurate user-transformer relationships.The proposed method is validated in a demonstration distribution area of a certain township in China,and the accuracy of the user-transformer relationship recognition is 99.8%.The results show that the proposed method has better recognition performance than mainstream methods such as adaptive segmentation aggregate approximation,correlation convolution,and density topological clustering.

关 键 词：户变关系 时空聚合 行波测距 用户电表 渐进迭代逼近方法 乡镇配电台区 

分 类 号：TM73[电气工程—电力系统及自动化]