摘要
针对电力负荷数据聚类过程中K均值算法人为指定聚类个数,导致聚类结果陷入局部最小解的问题,提出了基于动态时间归整(DTW)直方图的电力负荷数据聚类方法。利用主成分分析(PCA)法对高维电力负荷数据进行降维;引入直方图法确定负荷数据的初始聚类数目;通过DTW将负荷曲线分为K个类别;在MATLAB仿真平台上验证了该方法的有效性。实验结果表明:本文提出的算法在电力负荷数据聚类分析时减少了运算过程的迭代次数,加快了算法的收敛速度,并且聚类数目达到全局最优解的效果。
Aiming at the problem that the K-means algorithm needs to artificially specify the number of clustering in the process of power load data clustering,which leads to clustering result falling into the local minimum solution,a clustering method for power load data based on dynamic time warping(DTW)histogram is presented.Principal component analysis(PCA)method is used to reduce the dimension of high dimensional power load data.Histogram method is introduced to determine the initial clustering number of load data,the load curves are divided into K categories by DTW.The effectiveness of the method is verified on MATLAB simulation platform.The experimental results show that the proposed algorithm reduces the number of iterations in the operation process,accelerates the convergence speed of the algorithm,and achieves the effect of global optimal solution in the clustering analysis of power load data.
作者
郝晓弘
李亚岚
顾群
裴婷婷
宋吉祥
周强
HAO Xiaohong;LI Yalan;GU Qun;PEI Tingting;SONG Jixiang;ZHOU Qiang(School of Electrical Engineering and Information Engineering,Lanzhou University of Technology,Lanzhou 730050,China;School of Computer and Communication,Lanzhou University of Technology,Lanzhou 730050,China;State Grid Gansu Electric Power Company Wind Power Technology Company,Lanzhou 730000,China)
出处
《传感器与微系统》
CSCD
2020年第12期140-142,共3页
Transducer and Microsystem Technologies
基金
甘肃省部级资助项目(5227221600KQ)。
关键词
电力负荷
聚类
动态时间归整
直方图法
K均值
power load
clustering
dynamic time warping(DTW)
histogram method
K-means