Deep Embedding Clustering Method for Daily Load Based onConvolutional Auto-Encoder

doi:10.12204/j.issn.1000-7229.2021.01.015

Abstract

Abstract:

Clustering of load data is an important foundation for analyzing electrical big data. Aiming at the difficulty of extracting sequential features of high-dimensional daily load data, and the reduction of accuracy of load clustering due to the separation of feature extraction and clustering processing, a deep embedding clustering method based on one dimensional convolutional auto-encoder (DEC-1D-CAE) is proposed for daily load data in this paper. Firstly, a one-dimensional convolutional auto-encoder is used to extract sequential features contained in the load curve. Then, a user-defined clustering layer is used for soft division of the extracted load feature vector. Finally, the Kullback-Leibler divergence (KLD) is used as loss function to jointly optimize convolutional auto-encoder and the clustering layer to obtain the clustering result. A numerical experiment were carried out and the results of the proposed method are better than K-means, 1D-CAE+K-means and DEC-1D-CAE on both Davies-Bouldin index (DBI) and Calinski-Harabasz index (CHI), which indicate that the proposed method can effectively improve the accuracy of daily load clustering.

Key words: load clustering, convolutional auto-encoder(CAE), deep embedding clustering method (DEC), sequential feature extraction

CLC Number:

TM73

HUANG Dongmei, LIN Xiaoxiang, HU Anduo, SUN Jinzhong. Deep Embedding Clustering Method for Daily Load Based onConvolutional Auto-Encoder[J]. ELECTRIC POWER CONSTRUCTION, 2021, 42(1): 132-138.

Figures/Tables 7

Fig.1 Structure of 1D-CAE model

Fig.2 Flowchart of deep embedding clustering method of daily load applying one-dimensional convolutional auto-encoder

Fig.3 Comparison of load sequence reconstruction

Fig.4 Relative error of load reconstruction

Fig.5 Clustering results of the proposed method

Fig.6 Clustering results of K-means algorithm

Table 1 Comparison of four clustering methods

References 22

[1]	陆俊, 陈志敏, 龚钢军 , 等. 基于极限学习机的居民用电行为分类分析方法[J]. 电力系统自动化, 2019,43(2):97-104.
	LU Jun, CHEN Zhimin, GONG Gangjun , et al. Classification analysis method for electricity consumption behavior based on extreme learning machine algorithm[J]. Automation of Electric Power Systems, 2019,43(2):97-104.
[2]	孙毅, 顾玮, 李彬 , 等. 面向售电侧改革的用户分层聚类与套餐推荐方法[J]. 电网技术, 2018,42(2):447-454.
	SUN Yi, GU Wei, LI Bin , et al. Electricity-retail-market-reform oriented recommendation method of user’s hierarchical clustering and pricing packages[J]. Power System Technology, 2018,42(2):447-454.
[3]	黄俊铭, 朱建全, 庄远灿 . 基于动态RBF神经网络的广义电力负荷建模[J]. 电网技术, 2018,42(2):591-597.
	HUANG Junming, ZHU Jianquan, ZHUANG Yuancan . Generalized power load modeling based on dynamic RBF neural network[J]. Power System Technology, 2018,42(2):591-597.
[4]	TEERARATKUL T, O’NEILL D, LALL S . Shape-based approach to household electric load curve clustering and prediction[J]. IEEE Transactions on Smart Grid, 2018,9(5):5196-5206.
[5]	张铁峰, 顾明迪 . 电力用户负荷模式提取技术及应用综述[J]. 电网技术, 2016,40(3):804-811.
	ZHANG Tiefeng, GU Mingdi . Overview of electricity customer load pattern extraction technology and its application[J]. Power System Technology, 2016,40(3):804-811.
[6]	LIN S F, LI F X, TIAN E W , et al. Clustering load profiles for demand response applications[J]. IEEE Transactions on Smart Grid, 2019,10(2):1599-1607.
[7]	黄毅成, 杨洪耕 . 改进遗传K均值算法在负荷特性分类的应用[J]. 电力系统及其自动化学报, 2014,26(7):70-75.
	HUANG Yicheng, YANG Honggeng . Application of improved genetic and K-means algorithm on load characteristics classification[J]. Proceedings of the CSU-EPSA, 2014,26(7):70-75.
[8]	苏适, 李康平, 严玉廷 , 等. 基于密度空间聚类和引力搜索算法的居民负荷用电模式分类模型[J]. 电力自动化设备, 2018,38(1):129-136.
	SU Shi, LI Kangping, YAN Yuting , et al. Classification model of residential power consumption mode based on DBSCAN and gravitational search algorithm[J]. Electric Power Automation Equipment, 2018,38(1):129-136.
[9]	PHETSANGKAT P, CHALERMYANONT K, DUANGSOITHONG R . Hierarchical clustering electric load: Case study in lower south region of Thailand [C]//2019 16th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON). Pattaya, Chonburi, Thailand: IEEE, 2019: 881-884.
[10]	SUN Y, GU W, JUN L , et al. Fuzzy clustering algorithm-based classification of daily electrical load patterns [C]//2015 12th International Conference on Fuzzy Systems and Knowledge Discovery (FSKD). Zhangjiajie, China: IEEE, 2015: 50-54.
[11]	PARSONS L, HAQUE E, LIU H . Subspace clustering for high dimensional data[J]. ACM SIGKDD Explorations Newsletter, 2004,6(1):90-105.
[12]	宋亚奇, 周国亮, 朱永利 . 智能电网大数据处理技术现状与挑战[J]. 电网技术, 2013,37(4):927-935.
	SONG Yaqi, ZHOU Guoliang, ZHU Yongli . Present status and challenges of big data processing in smart grid[J]. Power System Technology, 2013,37(4):927-935.
[13]	张江林, 张亚超, 洪居华 , 等. 基于离散小波变换和模糊K-modes的负荷聚类算法[J]. 电力自动化设备, 2019,39(2):100-106, 122.
	ZHANG Jianglin, ZHANG Yachao, HONG Juhua , et al. A load clustering algorithm based on discrete wavelet transform and fuzzy K-modes[J]. Electric Power Automation Equipment, 2019,39(2):100-106, 122.
[14]	ZHONG S Y, TAM K S . Hierarchical classification of load profiles based on their characteristic attributes in frequency domain[J]. IEEE Transactions on Power Systems, 2015,30(5):2434-2441. doi: 10.1109/TPWRS.2014.2362492 URL
[15]	刘思, 李林芝, 吴浩 , 等. 基于特性指标降维的日负荷曲线聚类分析[J]. 电网技术, 2016,40(3):797-803. doi: 10.13335/j.1000-3673.pst.2016.03.020 URL
	LIU Si, LI Linzhi, WU Hao , et al. Cluster analysis of daily load curves using load pattern indexes to reduce dimensions[J]. Power System Technology, 2016,40(3):797-803. doi: 10.13335/j.1000-3673.pst.2016.03.020 URL
[16]	宋军英, 何聪, 李欣然 , 等. 基于特征指标降维及熵权法的日负荷曲线聚类方法[J]. 电力系统自动化, 2019,43(20):65-76.
	SONG Junying, HE Cong, LI Xinran , et al. Daily load curve clustering method based on feature index dimension reduction and entropy weight method[J]. Automation of Electric Power Systems, 2019,43(20):65-76.
[17]	李阳, 刘友波, 刘俊勇 , 等. 基于形态距离的日负荷数据自适应稳健聚类算法[J]. 中国电机工程学报, 2019,39(12):3409-3420.
	LI Yang, LIU Youbo, LIU Junyong , et al. Self-adaptive and robust clustering algorithm for daily load profiles based on morphological distance[J]. Proceedings of the CSEE, 2019,39(12):3409-3420.
[18]	赵文清, 龚亚强 . 基于Kernel K-means的负荷曲线聚类[J]. 电力自动化设备, 2016,36(6):203-207.
	ZHAO Wenqing, GONG Yaqiang . Load curve clustering based on Kernel K-means[J]. Electric Power Automation Equipment, 2016,36(6):203-207.
[19]	XIE J Y, GIRSHICK R, FARHADI A . Unsupervised deep embedding for clustering analysis[J/OL]. ( 2016-05-24)[2020-06-01]. https://arxiv.org/abs/1511.06335.
[20]	袁非牛, 章琳, 史劲亭 , 等. 自编码神经网络理论及应用综述[J]. 计算机学报, 2019,42(1):203-230.
	YUAN Feiniu, ZHANG Lin, SHI Jinting , et al. Theories and applications of auto-encoder neural networks: A literature survey[J]. Chinese Journal of Computers, 2019,42(1):203-230.
[21]	VAN DER MAATEN L, HINTON G . Visualizing data using t-SNE[J]. Journal of Machine Learning Research, 2008,9:2579-2605.
[22]	张斌, 庄池杰, 胡军 , 等. 结合降维技术的电力负荷曲线集成聚类算法[J]. 中国电机工程学报, 2015,35(15):3741-3749. doi: 10.13334/j.0258-8013.pcsee.2015.15.001 URL
	ZHANG Bin, ZHUANG Chijie, HU Jun , et al. Ensemble clustering algorithm combined with dimension reduction techniques for power load profiles[J]. Proceedings of the CSEE, 2015,35(15):3741-3749.