电力系统及其自动化学报2025，Vol.37Issue(7)：31-37,7.DOI:10.19635/j.cnki.csu-epsa.001597
电力系统频率自适应谱聚类空间分群方法
Spatial Clustering Method for Frequency Adaptive Spectral Clustering of Power System
摘要
Abstract
With the advancement of the construction of new power systems under the goal of"Double-carbon",the fre-quency spatial distribution characteristics of wide-area systems become more and more significant.Aimed at the defi-ciencies of dimensionality reduction and data adaptation in the clustering method for frequency spatial distribution char-acteristics based on the measurement data,a frequency spatial clustering method based on adaptive spectral clustering is proposed.First,a grid frequency measurement space is constructed,and the data window is adaptively determined.Second,the Pearson correlation coefficient is introduced to measure the similarity of frequency response at each fre-quency measurement node of the grid at the inertial response stage after perturbation.Third,based on the spectral clus-tering theory,a frequency space clustering method for power grid is proposed to portray the frequency space distribution phenomenon after the power grid disturbance.Finally,the effectiveness and feasibility of the frequency space clustering method is verified based on an improved IEEE 39-node system and a real system in one region.The results of examples show that the proposed clustering method can adaptively determine the time range of data measurement,and the cluster-ing results can effectively reflect the frequency spatial distribution characteristics.关键词
惯量/频率空间分布/频率分群/自适应/谱聚类Key words
inertia/frequency spatial distribution/frequency clustering/adaptation/spectral clustering分类
信息技术与安全科学引用本文复制引用
谢岩,唐晓骏,申家锴,马世英,苏志达,任健淼..电力系统频率自适应谱聚类空间分群方法[J].电力系统及其自动化学报,2025,37(7):31-37,7.基金项目
国家电网公司科技项目(5100-202255300A-2-0-QZ). （5100-202255300A-2-0-QZ）
