摘要
针对当前负荷建模中存在的负荷时变性问题,提出了基于自适应模糊C均值聚类的电力负荷动特性分类方法。探讨了聚类分析方法在负荷动特性分类中的应用,包括聚类特征向量的选取和分类方法研究两个方面。对原始模糊C均值聚类算法中的聚类数c进行了研究,在原始算法中融入新的聚类有效性函数,对算法进行了改进,改进算法不需要预先选择类的数目作为先验值。通过动模实验数据的负荷分类实例,表明该方法可自动获取最佳分类数,且分类效果要好于原始算法。
For the load time-variant characteristic in load modeling,this paper proposes a classification approach of the power load characteristics based on adaptive fuzzy C-means algorithm.It discusses the applications of clustering analysis in load characteristics classification,including the clustering feature vectors selection and study of the classification means.Based on study of the classification number C in the primal fuzzy C-means clustering algorithm,a new clustering validity function is added into the original fuzzy C-means to improve the original method,and the improved algorithm does not require pre-selected number of classes as a priori.A load classification example result with data based on dynamic simulation shows that the method can automatically obtain the best classification number,and classification is better than the original algorithm.
出处
《电力系统保护与控制》
EI
CSCD
北大核心
2010年第16期111-115,122,共6页
Power System Protection and Control
关键词
电力负荷
模糊C均值算法
自适应
动态特性聚类
负荷建模
power load
fuzzy C-means algorithm
adaptive
dynamic characteristics clustering
load modeling
