半月刊

ISSN 1000-1026

CN 32-1180/TP

+高级检索 English
基于欧氏-动态时间弯曲距离与熵权法的负荷曲线聚类方法
作者:
作者单位:

1.国网湖南省省电力公司;2.湖南大学电气与信息工程学院

作者简介:

通讯作者:

基金项目:

国家重点研发计划项目(2017YFB0903403) ;湖南省电力公司重点计划项目(5216A5180018)


Load Curve Clustering Method Based on Euclidean-Dynamic Time Warping Distance and Entropy
Author:
Affiliation:

1.State Grid Hunan Electric Power Corporation;2.College of Electrical and Information Engineering,Hunan University

Fund Project:

National Key Research and Development Project of China (2017YFB0903403) ;State Grid Hunan Electric Power Company Science and Technology Project (5216A5180018)

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
    摘要:

    为了改善目前负荷建模中聚类方法相似度衡量不准确及聚类结果质量较差的问题,本文综合运用k-means及熵权法原理,提出一种基于欧氏距离与动态时间弯曲距离(Dynamic Time Warping, DTW)的日负荷曲线聚类方法。首先采用欧氏距离与动态时间弯曲距离分别衡量日负荷曲线的整体分布特性、局部动态特性与整体动态特性;然后引入熵权法自适应配置此三种特性的权重系数;最后采用k-means聚类算法,以本文所提相似度衡量方法为依据,对用电日负荷曲线进行聚类。算例对某省区电网典型用户的日负荷曲线展开聚类分析,结果表明本文方法所选相似度衡量指标合理,且在聚类质量、鲁棒性等方面具有一定的优越性,可以真实反映该地区的用户用电特性,满足在线负荷建模的应用需求。

    Abstract:

    In order to improve the accuracy of similarity measurement and the clustering quality of clustering algorithms in load modeling, this paper proposed a daily load curve clustering algorithm based on Euclidean distance and dynamic time warping (DTW) by k-means and entropy weight. Firstly, Euclidean distance and dynamic time warping distance are adopted to measure the overall distribution characteristics, local dynamic characteristics and overall dynamic characteristics of the daily load curves; secondly, entropy weight method is introduced to configure the weight coefficients of these three characteristics adaptively; finally, k-means clustering algorithm is used to cluster the daily load curves based on the similarity measurement method proposed in this paper. This paper analyzed the daily load curves of typical consumers in a province. The results of simulation show that the similarity measurement method adopted in the algorithm proposed in this paper is reasonable, and the algorithm has some advantages in clustering quality and robustness, which can truly reflect the power consumption characteristics of consumers in this area and meet the application requirements of online load modeling.

    参考文献
    相似文献
    引证文献
引用本文

宋军英,崔益伟,李欣然,等.基于欧氏-动态时间弯曲距离与熵权法的负荷曲线聚类方法[J/OL].电力系统自动化,http://doi.org/10.7500/AEPS20191016008.
Song Junying,CuiYiwei,LiXinran,et al.Load Curve Clustering Method Based on Euclidean-Dynamic Time Warping Distance and Entropy[J/OL].Automation of Electric Power Systems,http://doi.org/10.7500/AEPS20191016008.

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2019-10-16
  • 最后修改日期:2020-05-08
  • 录用日期:2020-02-07
  • 在线发布日期:
  • 出版日期: