基于层次K-means聚类算法的配电网电压分区控制被引量：1

Voltage Zoning Control of Distribution Network Based on Hierarchical K-means Clustering Algorithm

下载PDF

导出

摘要可再生能源大量接入配电网,其出力的不确定性和波动性增大了配电网系统无功电压控制的难度。本文提出一种配电网无功电压分区控制方法,通过将系统划分为多个控制区域,利用无功源的无功出力实现电压分区控制。配电网分区策略是利用层次聚类为基础的Kmeans聚类算法,以轮廓系数和Calinski-Harabasz为分区评价指标,各分区分布式电源接入点通过灵敏度系数确定主导节点。以系统网损最小、电压偏差和电压稳定性指标建立无功优化模型,权重系数采用主观赋值法,通过改进粒子群算法(IPSO)对模型求解寻得最优解。分区控制后,网损减少9.7184 kW,电压偏差减小0.0561 p.u.。最后,通过改进IEEE33节点配电网系统进行仿真验证,结果表明本方法能有效控制无功电压安全运行。 A large number of renewable energy sources are connected to the distribution network,while the uncertainty and volatility of their output increases the difficulty of reactive voltage control in the distribution network system.A reactive power divider control method for distribution networks is proposed in the present study,which divides the system into multiple control areas and utilizes the reactive power output of the reactive power source to achieve divider control.The hierarchical clustering based K-means clustering algorithm is adopted as the distribution network partitioning strategy,using the profile coefficients and Calinski-Harabasz as the partitioning evaluation indexes,and the dominant nodes are determined by the sensitivity coefficients for the access points of distributed power sources in each partition.A reactive power optimization model is established based on the indicators of minimum system network loss,voltage deviation,and voltage stability,whose weight coefficients are subjectively assigned,and the optimal solution for the model is obtained by using the improved Particle Swarm Optimization(IPSO)algorithm.With zoning control,the network loss is reduced by 9.7184 kW and the voltage deviation is reduced by 0.0561 p.u..Finally,the improved IEEE33 node distribution system is simulated to verify that this method can effectively carry out reactive power voltage control to ensure safe operation.

作者张令雄杨毅强徐万唐青林 ZHANG Lingxiong;YANG Yiqiang;XU Wan;TANG Qinglin(School of Automation and Information Engineering,Sichuan University of Science&Engineering,Yibin 644000,China;Intelligent Perception and Control Key Laboratory of Sichuan Province,Yibin 644000,China)

机构地区四川轻化工大学自动化与信息工程学院智能感知与控制四川省重点实验室

出处《四川轻化工大学学报(自然科学版)》 2025年第4期92-101,共10页 Journal of Sichuan University of Science & Engineering(Natural Science Edition)

基金四川省科技厅项目(2022YFS0518,2022ZHCG0035) 自贡市科技局项目(2020YGJC16)。

关键词配电网分区控制层次聚类灵敏度系数优化模型 distribution network zoning control hierarchical clustering sensitivity coefficients optimization models

分类号 TM7 [电气工程—电力系统及自动化]