基于改进FCM聚类分析方法的电力负荷特性研究  

Research on power load characteristics based on improved FCM clustering analysis method

作　　者：李富强 朱晨烜(指导)[1] 骆利勤 LI Fuqiang;ZHU Chenxuan;LUO Liqin(School of Electrical Engineering,Shanghai Dianji University,Shanghai 201306,China)

机构地区：[1]上海电机学院电气学院,上海201306

出　　处：《上海电机学院学报》2023年第5期293-298,共6页Journal of Shanghai Dianji University

摘　　要：针对模糊C-均值(FCM)的聚类结果受初始隶属度矩阵和聚类中心的选择影响,容易遇到局部极值等问题,提出了一种改进的FCM聚类分析方法。首先,考虑到电力负荷曲线的规律性与时序性特点,结合欧氏距离与皮尔逊相关系数,对局部密度进行了优化。然后,利用FCM聚类分析方法对聚类质心更新迭代,以获得更高精度的聚类质心,从而实现了电力负荷曲线的精确聚类。最后,以电网实际负荷数据作为仿真实验样本,通过与K-means和传统FCM方法进行比较,验证了改进的FCM聚类分析方法对电力负荷曲线聚类分析的有效性和优越性。To solve the problem that the clustering results of fuzzy C-means(FCM)are affected by the selection of the initial membership matrix and cluster center,and easily trap into local extremum,etc.,an improved FCM clustering method is proposed.First,considering the regularity and timing characteristics of the power load curve,the local density is optimized by combining the Euclidean distance and Pearson correlation coefficient.Then,the FCM cluster analysis method is utilized to iteratively update the cluster centroids to achieve higher accurate cluster centroids,thus achieving accurate clustering of power load curves.Finally,the real load data of the power grid is used as a simulation experiment sample,and K-means and traditional FCM methods are compared to verify the effectiveness and superiority of the improved FCM cluster analysis method for clustering analysis of power load curves.

关 键 词：聚类分析方法 电力负荷 聚类质心 

分 类 号：TM715[电气工程—电力系统及自动化]