东北电力技术2025,Vol.46Issue(1) :1-5.

基于改进K-means聚类算法的分布式储能集群划分方法

Distributed Energy Storage Cluster Partitioning Method Based on Improved K-means Clustering Algorithm

刘春雨
东北电力技术2025,Vol.46Issue(1) :1-5.

基于改进K-means聚类算法的分布式储能集群划分方法

Distributed Energy Storage Cluster Partitioning Method Based on Improved K-means Clustering Algorithm

刘春雨1
扫码查看

作者信息

  • 1. 上海联联睿科能源科技有限公司,上海 200063
  • 折叠

摘要

随着规模化分布式电源及储能的接入,配电网的功率返送、节点过电压等问题愈加显著,对电网规划、运行监视和调度控制等造成一定影响,也不利于储能大范围发展.为此提出一种适用于规模化分布式储能的集群划分方法,基于功率节点电压灵敏度的电气距离模块度指标,对经典K-means算法进行改进,设计节点指数法、肘部法则优化初始聚类中心选择和集群数确定.以IEEE33 系统算例进行验证,结果表明所提集群划分方法具有较强的电气耦合性、准确性和运算效率.

Abstract

With the access of large-scale distributed power supply and energy storage,the problems of power backflow and node over-voltage in the distribution network are more significant,which have a certain impact on the grid planning,operation monitoring and scheduling control,and is not conducive to the large-scale development of energy storage.Therefore,it proposes a cluster partitioning method suitable for large-scale distributed energy storage,and an electrical distance modularity index based on voltage sensitivity of power nodes.It improves the classical K-means algorithm,and designs the node index method and elbow rule to optimize the selection of initial cluster centers and the determination of cluster number.It verifies the algorithm by an example of IEEE33 system.The re-sults show that the proposed cluster partitioning method has strong electrical coupling,accuracy and operation efficiency.

关键词

分布式储能/集群划分/配电网结构/K-means聚类算法/划分指标

Key words

distributed energy storage/cluster partitioning/distribution network structure/K-means clustering algorithm/dividing in-dicators

引用本文复制引用

出版年

2025
东北电力技术
东北电网有限公司,辽宁省电力有限公司

东北电力技术

影响因子:0.744
ISSN:1004-7913
段落导航相关论文