基于特征用户选择的停电范围判定方法研究  

Research on Outage Range Determination Method Based on Feature User Selection

作　　者：李蓓 胡筱曼 林文浩 熊力 LI Bei;HU Xiaoman;LIN Wenhao;XIONG Li(Guangdong Zhongshan Power Supply Bureau,China Southern Power Grid,Zhongshan 528400,China)

机构地区：[1]南方电网广东中山供电局,广东中山528400

出　　处：《自动化仪表》2023年第10期85-89,94,共6页Process Automation Instrumentation

基　　金：南方电网公司科技基金资助项目(032000KK52190153)。

摘　　要：台区停电范围判定是停电管理的核心内容。现有方法需改造台区内所有电表才能实现停电范围判定,成本较高。基于特征用户选择方法,提出了一种新的台区停电范围判定方法。该方法仅需改造少量电表即可实现停电范围判定。基于K-means++聚类算法对用户进行网格划分,通过三个无监督聚类评价指标对聚类结果进行评价,选取最优聚类参数,以实现用户最优分组。计算同簇各用户间停电相关性系数,选择停电相关性系数高的用户作为特征用户,加装具有超级电容器的智能电表,以实现用户停电的主动告警及台区停电范围判定。试验结果表明,所提方法能减少设备配置、较为准确地判定停电范围,并避免无效投资。该研究可实现数据驱动型主动服务,为低压配网主动抢修提供技术支撑。The determination of outage range in a station area is the core content of outage management.Existing methods need to modify all meters in the station area to realize outage range determination,which is costly.Based on the feature user selection method,a new outage range determination method in the station area is proposed.The method only needs to modify a small number of meters to realize the outage range determination.Based on the K-means++clustering algorithm,the users are divided into grids,and the clustering results are evaluated by three unsupervised clustering evaluation indexes,and the optimal clustering parameters are selected to realize the optimal grouping of users.The outage correlation coefficient between users in the same cluster is calculated,the user with high outage correlation coefficient is selected as the feature user,and smart meters with supercapacitors are installed to realize the active alarm of outage of users and the determination of outage range in the station area.The test results show that the proposed method can reduce equipment configuration,determine the outage range more accurately,and avoid ineffective investment.The research can realize data-driven active service,and provide technical support for the active repair of low-voltage distribution network.

关 键 词：配电网 智能电表 特征用户 停电相关性系数 停电范围 聚类 告警 

分 类 号：TH-89[机械工程]