[1]陈苏豫,顾亦然,张腾飞.基于 DLT-Kmedoids 算法的用电负荷聚类分析[J].计算机技术与发展,2024,34(04):205-211.[doi:10. 3969 / j. issn. 1673-629X. 2024. 04. 031]
 CHEN Su-yu,GU Yi-ran,ZHANG Teng-fei.Power Load Clustering Analysis Based on DLT-Kmedoids[J].,2024,34(04):205-211.[doi:10. 3969 / j. issn. 1673-629X. 2024. 04. 031]
点击复制

基于 DLT-Kmedoids 算法的用电负荷聚类分析()
分享到:

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
34
期数:
2024年04期
页码:
205-211
栏目:
新型计算应用系统
出版日期:
2024-04-10

文章信息/Info

Title:
Power Load Clustering Analysis Based on DLT-Kmedoids
文章编号:
1673-629X(2024)04-0205-07
作者:
陈苏豫1 顾亦然12 张腾飞1
1. 南京邮电大学 自动化学院、人工智能学院,江苏 南京 210023;2. 南京邮电大学 智慧校园研究中心,江苏 南京 210023
Author(s):
CHEN Su-yu1 GU Yi-ran12 ZHANG Teng-fei1
1. School of Automation and School of Artificial Intelligence,Nanjing University of Posts and Telecommunications,Nanjing 210023,China;
2. Smart Campus Research Center,Nanjing University of Posts and Telecommunications,Nanjing 210023,China
关键词:
用电负荷数据动态时间弯曲LB_Keogh聚类用电模式
Keywords:
power load datadynamic time warpingLB_Keoghclusteringpower consumption mode
分类号:
TP181;TM743
DOI:
10. 3969 / j. issn. 1673-629X. 2024. 04. 031
摘要:
针对高校用电负荷中传统聚类算法直接应用于时间序列聚类效果准确性较低的问题,提出一种融合 DTW 距离、LB_Keogh 距离以及时间窗口的 DLT - Kmedoids 算法,以提高聚类算法应用于时间序列的准确性以及算法效率。 DLT -Kmedoids 算法使用 DTW 计算时序数据之间的距离取代传统的欧氏距离度量方式,提高了相似性度量算法精度,同时也提高了聚类算法的准确性和复杂度,引入 LB_Keogh 距离在计算 DTW 距离之前过滤掉大部分不可能是最优匹配序列的序列,对于剩下的序列再使用 DTW 逐个比较,进一步降低算法的复杂度。 最后结合高校建筑用电负荷时间序列数据进行分析,通过与主流聚类算法进行比较,表明该算法对于高校用电负荷数据的聚类任务,能够更准确地识别相似的负荷模式,并以更高的效率进行聚类分析。
Abstract:
Aiming at the problem of low accuracy in directly applying traditional clustering algorithms to time series clustering in theelectricity load of universities,a DLT-Kmedoids?
algorithm combining DTW distance,LB_Keogh distance and time window is proposedto improve the accuracy and efficiency of clustering algorithm applied to time series.?
The DLT - Kmedoids algorithm uses DTW tocalculate the distance between time series data instead of traditional Euclidean distance measurement,improving the accuracy of similaritymeasurement algorithms and also improving the accuracy and complexity of clustering algorithms. LB_Keogh distance is introduced tofilter out most sequences that are unlikely to be optimal matching sequences before calculating DTW distance, and DTW is used tocompare the remaining sequences one by one to further reduce the complexity?
of the algorithm. Finally,we analyze the time series data of electricity consumption in university buildings, and compare it with mainstream clustering algorithms. It is showed that the proposedalgorithm can more accurately identify similar load patterns and perform clustering analysis with higher efficiency for the clustering task ofelectricity consumption data in universities.
更新日期/Last Update: 2024-04-10