摘要
随着大规模分布式电源的接入,采用集中式控制的传统配电网面临通信延时、计算量大与控制设备过多等问题,而基于集群划分的分布式发电群调群控技术能有效解决上述问题。而现有集群划分方法在集群划分指标与集群划分算法上均存在一定不足,因此提出一种考虑集群规模的分布式发电集群划分方法。首先,提出考虑电气距离、集群功率平衡以及集群规模的综合性集群划分指标体系,在保证集群结构强度的基础上使集群具有一定电压调节能力。其次,采用嵌入莱维飞行优策略的灰狼优化算法,对K-means算法进行改进,并将其应用于集群划分。最后,以某地实际35 kV/10 kV配电网验证了所提方法的可行性与有效性,为分布式发电集群划分提供参考。
With the access of large-scale distributed power supply,the traditional centralized-control distribution network encounters issues including communication delays,heavy computational loads,and an overabundance of control equipment.However,the distributed generation group control technology,which is based on cluster partitioning,can effectively address these challenges.Existing cluster partitioning methods exhibit deficiencies in both cluster partitioning index and cluster partitioning algorithm,so we propose a new cluster partitioning method considering cluster size.Firstly,we propose a comprehensive index system for cluster partitioning,which considers electrical distance,cluster power balance and cluster scale.This ensures that the cluster maintains a certain voltage regulation ability and structural strength of the cluster.Secondly,we use the grey wolf optimization algorithm embedded with Levy flight optimization strategy to improve the K-means algorithm and apply it to cluster partitioning.Finally,we verify the feasibility and effectiveness of the proposed method by the actual 35 kV/10 kV distribution network.This serves as a valuable reference for the partitioning of distributed generation clusters.
作者
尉同正
杜红卫
夏栋
韩韬
吴雪琼
徐政
WEI Tongzheng;DU Hongwei;XIA Dong;HAN Tao;WU Xueqiong;XU Zheng(NARI Group Corporation(State Grid Electric Power Research Institute),Nanjing 211106,China;NARI Technology Co.,Ltd,Nanjing 211106,China;College of Engineering,University of Georgia,Athens 30602,USA)
出处
《华北电力大学学报（自然科学版）》
CAS
北大核心
2024年第6期49-57,共9页
Journal of North China Electric Power University：Natural Science Edition
基金
国家重点研发计划(2020YEF0200400).
关键词
灰狼优化算法
莱维飞行
K-MEANS算法
分布式电源
集群划分
gray wolf optimization algorithm
Levy flight
K-means algorithm
distributed generation
cluster partition
