电工技术2024,Issue(20) :71-73,76.DOI:10.19768/j.cnki.dgjs.2024.20.019

基于随机森林算法的电网拓扑结构识别

Random Forest Algorithm-based Grid Topology Identification

杨化冰 周雷 史松鹤 吴迪 王玉龙
电工技术2024,Issue(20) :71-73,76.DOI:10.19768/j.cnki.dgjs.2024.20.019

基于随机森林算法的电网拓扑结构识别

Random Forest Algorithm-based Grid Topology Identification

杨化冰 1周雷 1史松鹤 1吴迪 1王玉龙1
扫码查看

作者信息

  • 1. 国网河南省电力有限公司舞阳县供电公司,河南 漯河 462400
  • 折叠

摘要

低压台区的拓扑结构对于配电网络的线损分析、需求响应以及安全运行有着重要意义,针对目前配电网存在的拓扑关系不明朗的问题,提出了基于电压时间序列的随机森林聚类方法.通过对用户侧与配变二次侧的电压相似度机理的分析,利用皮尔逊相似度构造电压时间序列的特征向量集,建立基于随机森林的拓扑结构模型.最后通过仿真,验明该方法具有较高的准确性.

Abstract

The topology of the low-voltage distribution area is of great significance to line loss analysis,demand response,and safe operation of distributing networks.This work proposed a random forest clustering method using voltage time se-ries to solve the problem of unclear topological relationship in distributing networks.Through the analysis of the voltage similarity mechanism between the user side and the secondary side,the Pearson similarity was used to construct the fea-ture vector set of the voltage time series,and random forest-based topological structure model was established.Through simulation analysis,the proposed method was verified highly accurate.

关键词

低压台区/拓扑识别/皮尔逊相关系数/随机森林

Key words

low-voltage distribution area/topology identification/Pearson correlation coefficient/random forest

引用本文复制引用

出版年

2024
电工技术
重庆西南信息有限公司(原科技部西南信息中心)

电工技术

影响因子:0.177
ISSN:1002-1388
段落导航相关论文