基于多特征符号聚合近似和层次聚类的户变关系识别方法  

Identification Method for User-transformer Relationship Based on Multi-feature Symbolic Aggregate Approximation and Hierarchical Clustering

作　　者：周赣[1] 茅欢 冯燕钧 华济民 曾瑛 ZHOU Gan;MAO Huan;FENG Yanjun;HUA Jimin;ZENG Ying(School of Electrical Engineering,Southeast University,Nanjing 210096,China;State Grid Wuxi Power Supply Company of Jiangsu Electric Power Co.,Ltd.,Wuxi 214000,China;Guangdong Power Grid Co.,Ltd.,Guangzhou 510600,China)

机构地区：[1]东南大学电气工程学院,江苏省南京市210096 [2]国网江苏省电力有限公司无锡供电分公司,江苏省无锡市214000 [3]广东电网有限责任公司,广东省广州市510600

出　　处：《电力系统自动化》2024年第3期133-141,共9页Automation of Electric Power Systems

基　　金：广东省重点领域研发计划资助项目(2020B0101130023)。

摘　　要：针对低压配电台区拓扑档案中可能存在的户变关系异常问题,文中提出了一种基于多特征符号聚合近似(MF-SAX)和层次聚类的户变关系识别方法。首先,运用符号聚合近似表达方法将用户电压时间序列转化为字符串序列,并引入电压波动系数和电压变化趋势两个附加参数对其特征表达进行强化。然后,基于编辑距离生成用户电压曲线相似性矩阵,并结合层次聚类算法实现户变关系的识别。最后,实际算例结果表明,提出的方法相比于现有方法准确率更高,误报更少,能直接应对数据缺失的情况,且具有更高的效率。In view of the possible problem of wrong user-transformer relationships in the topology file of the low-voltage distribution station area,this paper proposes a user-transformer relationship identification method based on multi-feature symbolic aggregate approximation(MF-SAX)and hierarchical clustering.First,the symbolic aggregate approximation expression method is used to convert the user voltage time series into string series,and two additional parameters,i.e.,the voltage fluctuation coefficient and the voltage change trend,are introduced to strengthen its feature expression.Then,the similarity matrix of the user voltage curves is generated based on the edit distance,and the hierarchical clustering algorithm is combined to realize the identification of the user-transformer relationships.Finally,the results of the practical case show that,compared with some existing methods,the proposed method achieves higher accuracy and fewer false alarms,which can directly respond to missing data situations,and has higher efficiency.

关 键 词：低压配电台区 户变关系 层次聚类 拓扑识别 电压曲线相似性 

分 类 号：TM73[电气工程—电力系统及自动化]