电力信息与通信技术2025，Vol.23Issue(6)：39-47,9.DOI:10.16543/j.2095-641x.electric.power.ict.2025.06.06
基于相似性分析和Lasso回归算法的配电网拓扑识别方法
Topology Identification Method for Distribution Network Based on Similarity Analysis and Lasso Regression Algorithm
摘要
Abstract
In response to the problem of inaccurate topology identification caused by multicollinearity in the voltage of distribution network nodes,this paper proposes using multi time cross-sectional node voltage data for topology identification,identify nodes with high similarity,and use Lasso regression algorithm to screen neighboring nodes.Firstly,Pearson algorithm was used to analyze the correlation coefficients between nodes,and it was found that multiple nodes had high correlation with non adjacent nodes.An approximate linear correlation between node voltages was derived.Then,Pearson algorithm,Euclidean distance,and DTW algorithm are used as correlation evaluation indicators to identify nodes with multicollinearity.Next,the main power supply node is taken as the parent node,and the Lasso regression algorithm is used to determine the child nodes.Then,the child nodes are taken as the new parent nodes,and this process is repeated for secondary identification to generate the topology structure.Finally,the feasibility and accuracy of the method were verified through an IEEE33 node example.关键词
配电网/拓扑识别/多重相关性/相似性算法/Lasso回归Key words
distribution network/topology identification/multiple correlations/similarity algorithm/Lasso regression分类
动力与电气工程引用本文复制引用
宋炆学,张腾飞,邹花蕾..基于相似性分析和Lasso回归算法的配电网拓扑识别方法[J].电力信息与通信技术,2025,23(6):39-47,9.基金项目
江苏省自然科学基金项目(SJ223029). （SJ223029）
