摘要
该文分析了接入节点的光伏电源与负荷的相关特性,选取节点净负荷、光伏电源有功和无功调节容量为划分指标,并考虑节点之间的电气距离特性,通过节点相似性计算公式量化光伏电源节点间运行特性的相似性,利用Fast Unfolding聚类算法进行光伏电源集群的划分。在划分的过程中引入复杂网络模块度函数的概念,通过优化模块度函数,无需人为设定集群的数目即可得到合理的集群划分结果。
Based on the analysis of correlation characteristics between PV and load in the distribution network,this paper select net load,active and reactive power regulation capacity of PV as indicators,and the electrical distance characteristics between nodes are also considered.The similarity between PV nodes is quantified by the node similarity calculation formula,and the Fast Unfolding clustering algorithm is used to partitioning the PV clusters.Modularity function of complex network is introduced in the partitioning process.With the aid of optimizing of the modularity function,reasonable clustering results can be obtained without artificially setting the number of clusters.
作者
王磊
张帆
寇凌峰
徐毅虎
侯小刚
Wang Lei;Zhang Fan;Kou Lingfeng;Xu Yihu;Hou Xiaogang(Anhui Provincial Engineering Laboratory for Renewable Energy and Energy Saving(Hefei University of Technology),Hefei 230009,China;China Electric Power Research Institute,Beijing 100192,China)
出处
《太阳能学报》
EI
CAS
CSCD
北大核心
2021年第10期29-34,共6页
Acta Energiae Solaris Sinica
基金
国家重点研发计划智能电网技术与装备重点专项(2016YFB0900600)
国家电网公司科技项目(52094017000W)。
关键词
大规模系统
聚类分析
聚类算法
光伏电源
模块度函数
large scale system
clustering analysis
clustering algorithm
photovoltaic generation
modularity function
