基于无监督极限学习机的用电负荷模式提取
|
摘要
能源互联网的建设及发展使得用能端数据不断积累,充分利用用电信息,挖掘典型的用电负荷模式,是协调规划能源互联网并为其用户提供个性化服务的基础。用电负荷模式的提取通常以对负荷曲线进行聚类分析为基础,作为聚类算法输入的负荷曲线特征对聚类效果影响较大,高维输入容易使传统聚类算法表现不佳,对负荷曲线特征进行优化选取成为了一个重要问题。提出一种基于降维的负荷曲线聚类方法,首先用无监督极限学习机对原始负荷序列数据集进行低维嵌入,然后采用k-means算法对提取的低维特征进行聚类。给出了采用无监督极限学习机结合k-means方法进行负荷模式提取的具体流程。通过算例对比了传统k-means、其它降维算法加k-means与所提方法,实验结果证明所提方法聚类效果更好且效率较高,区分出的负荷曲线类别更能反映实际规律,得到的不同典型负荷曲线之间差异性更大。基于无监督极限学习机的聚类方法因其有效性可应用到用电负荷模式提取过程中。
Construction and development of Energy Internet keeps data of electricity consumption side constantly accumulating. Making full use of electricity information and mining typical load pattern of electricity consumption is the basis for coordinating planning of Energy Internet and providing personalized services to its users. Extracting load pattern of electricity consumption is usually based on cluster analysis about load curves. The load curve characteristics, as input of clustering algorithm, has a great influence on clustering effect. High-dimensional input is prone to make traditional clustering algorithm perform poorly, so it is important to optimize selection of load curve characteristics. A novel load curve clustering method based on dimension reduction is proposed. Firstly, original load sequence data set is embedded in low-dimensional space with unsupervised extreme learning machine, then the low-dimensional features are clustered with k-means algorithm. The specific process of load pattern extraction with unsupervised extreme learning machine combined with k-means method is given. The proposed method is compared to traditional k-means algorithm and other dimension reduction algorithms with k-means for clustering. Experimental results show that the proposed method has better clustering effect and higher efficiency. Curve types distinguished with the proposed method can better reflect actual law, and there is bigger difference among different typical load curves. The proposed method can be applied to extraction process of electricity consumption load pattern for its effectiveness.
Construction and development of Energy Internet keeps data of electricity consumption side constantly accumulating. Making full use of electricity information and mining typical load pattern of electricity consumption is the basis for coordinating planning of Energy Internet and providing personalized services to its users. Extracting load pattern of electricity consumption is usually based on cluster analysis about load curves. The load curve characteristics, as input of clustering algorithm, has a great influence on clustering effect. High-dimensional input is prone to make traditional clustering algorithm perform poorly, so it is important to optimize selection of load curve characteristics. A novel load curve clustering method based on dimension reduction is proposed. Firstly, original load sequence data set is embedded in low-dimensional space with unsupervised extreme learning machine, then the low-dimensional features are clustered with k-means algorithm. The specific process of load pattern extraction with unsupervised extreme learning machine combined with k-means method is given. The proposed method is compared to traditional k-means algorithm and other dimension reduction algorithms with k-means for clustering. Experimental results show that the proposed method has better clustering effect and higher efficiency. Curve types distinguished with the proposed method can better reflect actual law, and there is bigger difference among different typical load curves. The proposed method can be applied to extraction process of electricity consumption load pattern for its effectiveness.
引文
[1]曾鸣,杨雍琦,刘敦楠,等.能源互联网“源–网–荷–储”协调优化运营模式及关键技术[J].电网技术,2016,40(1):114-124.Zeng Ming,Yang Yongqi,Liu Dunnan,et al.“Generationgrid-load-storage”coordinative optimal operation mode of energy internet and key technologies[J].Power System Technology,2016,40(1):114-124(in Chinese).
[2]蒋菱,王旭东,于建成,等.基于分布式计算的海量用电数据分析技术研究[J].计算机技术与发展,2016,26(12):176-181.Jiang Ling,Wang Xudong,Yu Jiancheng,et al.Research on power usage behavior analysis based on distributed computing[J].Computer Technology and Development,2016,26(12):176-181(in Chinese).
[3]Picard M L,潘旭阳.从法国公共电力企业的视角看大数据带来的挑战和机遇[J].电网技术,2015,39(11):3109-3113.Picard M L,Pan Xuyang.Big data challenges and opportunities for a utility as EDF[J].Power System Technology,2015,39(11):3109-3113(in Chinese).
[4]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(in Chinese).
[5]彭显刚,赖家文,陈奕.基于聚类分析的客户用电模式智能识别方法[J].电力系统保护与控制,2014,42(19):68-73.Peng Xiangang,Lai Jiawen,Chen Yi.Application of clustering analysis in typical power consumption profile analysis[J].Power System Protection and Control,2014,42(19):68-73(in Chinese).
[6]颜庆国,薛溟枫,范洁,等.有序用电用户负荷特性分析方法研究[J].江苏电机工程,2014,33(6):48-50,54.Yan Qingguo,Xue Mingfeng,Fan Jie,et al.Load property analysis method for demanders participating orderly power utilization[J].Jiangsu Electrical Engineering,2014,33(6):48-50,54(in Chinese).
[7]王星华,陈卓优,彭显刚.一种基于双层聚类分析的负荷形态组合识别方法[J].电网技术,2016,40(5):1495-1501.Wang Xinghua,Chen Zhuoyou,Peng Xiangang.A new combinational electrical load analysis method based on bilayer clustering analysis[J].Power System Technology,2016,40(5):1495-1501(in Chinese).
[8]罗滇生,杜乾,别少勇,等.基于负荷分解的居民差异化用电行为特性分析[J].电力系统保护与控制,2016,44(21):29-33.Luo Diansheng,Du Qian,Bie Shaoyong,et al.Analysis of differentiation residential electricity consumption characteristic based on power load decomposition[J].Power System Protection and Control,2016,44(21):29-33(in Chinese).
[9]杨永标,颜庆国,王冬,等.居民用户智能用电建模及优化仿真分析[J].电力系统自动化,2016,40(3):46-51.Yang Yongbiao,Yan Qingguo,Wang Dong,et al.Intelligent electricity consumption modeling and optimal simulations for residential users[J].Automation of Electric Power Systems,2016,40(3):46-51(in Chinese).
[10]王春雷,梁小放,章坚民,等.基于用电采集系统的负荷特性曲线聚类分析[J].浙江电力,2014(7):6-10.Wang Chunlei,Liang Xiaofang,Zhang Jianmin,et al.Cluster analysis of load characteristic curve based on electricity consumption information acquisition system[J].Zhejiang Electric Power,2014(7):6-10(in Chinese).
[11]庄池杰,张斌,胡军,等.基于无监督学习的电力用户异常用电模式检测[J].中国电机工程学报,2016,36(2):379-387.Zhuang Chijie,Zhang Bin,Hu Jun,et al.Anomaly detection for power consumption patterns based on unsupervised learning[J].Proceedings of the CSEE,2016,36(2):379-387(in Chinese).
[12]赵莉,候兴哲,胡君,等.基于改进K-means算法的海量智能用电数据分析[J].电网技术,2014,38(10):2715-2720.Zhao Li,Hou Xingzhe,Hu Jun,et al.Improved K-means algorithm based analysis on massive data of intelligent power utilization[J].Power System Technology,2014,38(10):2715-2720(in Chinese).
[13]和敬涵,卢育梓,陆金耀,等.基于“进化”主成分分析法的用户分类及其应用[J].电力建设,2017,38(3):101-107.He Jinghan,Lu Yuzi,Lu Jinyao,et al.User classification method based on’evolution’PCA and its application[J].Electric Power Construction,2017,38(3):101-107(in Chinese).
[14]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.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(in Chinese).
[15]王德文,孙志伟.一种基于内存计算的电力用户聚类分析方法[J].智能系统学报,2015,10(4):569-576.Wang Dewen,Sun Zhiwei.A method for cluster analysis of electric power consumers based on in-memory computing[J].CAAI Transactions on Intelligent Systems,2015,10(4):569-576(in Chinese).
[16]朱文俊,王毅,罗敏,等.面向海量用户用电特性感知的分布式聚类算法[J].电力系统自动化,2016,40(12):21-27.Zhu Wenjun,Wang Yi,Luo Min,et al.Distributed clustering algorithm for awareness of electricity consumption characteristics of massive consumers[J].Automation of Electric Power Systems,2016,40(12):21-27(in Chinese).
[17]张进,徐苒茨,徐健,等.居民用户用电的聚类分析[J].电力与能源,2016,37(2):180-184.Zhang Jin,Xu Ranci,Xu Jian,et al.Clustering analysis of residential electricity utilization[J].Power&Energy,2016,37(2):180-184(in Chinese).
[18]Huang G,Song S J,Gupta J N D,et al.Semi-supervised and unsupervised extreme learning machines[J].IEEE Transactions on Cybernetics,2014,44(12):2405-2417.
[19]Huang G B,Zhu Q Y,Siew C K.Extreme learning machine:Theory and applications[J].Neurocomputing,2006,70(1-3):489-501.
[2]蒋菱,王旭东,于建成,等.基于分布式计算的海量用电数据分析技术研究[J].计算机技术与发展,2016,26(12):176-181.Jiang Ling,Wang Xudong,Yu Jiancheng,et al.Research on power usage behavior analysis based on distributed computing[J].Computer Technology and Development,2016,26(12):176-181(in Chinese).
[3]Picard M L,潘旭阳.从法国公共电力企业的视角看大数据带来的挑战和机遇[J].电网技术,2015,39(11):3109-3113.Picard M L,Pan Xuyang.Big data challenges and opportunities for a utility as EDF[J].Power System Technology,2015,39(11):3109-3113(in Chinese).
[4]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(in Chinese).
[5]彭显刚,赖家文,陈奕.基于聚类分析的客户用电模式智能识别方法[J].电力系统保护与控制,2014,42(19):68-73.Peng Xiangang,Lai Jiawen,Chen Yi.Application of clustering analysis in typical power consumption profile analysis[J].Power System Protection and Control,2014,42(19):68-73(in Chinese).
[6]颜庆国,薛溟枫,范洁,等.有序用电用户负荷特性分析方法研究[J].江苏电机工程,2014,33(6):48-50,54.Yan Qingguo,Xue Mingfeng,Fan Jie,et al.Load property analysis method for demanders participating orderly power utilization[J].Jiangsu Electrical Engineering,2014,33(6):48-50,54(in Chinese).
[7]王星华,陈卓优,彭显刚.一种基于双层聚类分析的负荷形态组合识别方法[J].电网技术,2016,40(5):1495-1501.Wang Xinghua,Chen Zhuoyou,Peng Xiangang.A new combinational electrical load analysis method based on bilayer clustering analysis[J].Power System Technology,2016,40(5):1495-1501(in Chinese).
[8]罗滇生,杜乾,别少勇,等.基于负荷分解的居民差异化用电行为特性分析[J].电力系统保护与控制,2016,44(21):29-33.Luo Diansheng,Du Qian,Bie Shaoyong,et al.Analysis of differentiation residential electricity consumption characteristic based on power load decomposition[J].Power System Protection and Control,2016,44(21):29-33(in Chinese).
[9]杨永标,颜庆国,王冬,等.居民用户智能用电建模及优化仿真分析[J].电力系统自动化,2016,40(3):46-51.Yang Yongbiao,Yan Qingguo,Wang Dong,et al.Intelligent electricity consumption modeling and optimal simulations for residential users[J].Automation of Electric Power Systems,2016,40(3):46-51(in Chinese).
[10]王春雷,梁小放,章坚民,等.基于用电采集系统的负荷特性曲线聚类分析[J].浙江电力,2014(7):6-10.Wang Chunlei,Liang Xiaofang,Zhang Jianmin,et al.Cluster analysis of load characteristic curve based on electricity consumption information acquisition system[J].Zhejiang Electric Power,2014(7):6-10(in Chinese).
[11]庄池杰,张斌,胡军,等.基于无监督学习的电力用户异常用电模式检测[J].中国电机工程学报,2016,36(2):379-387.Zhuang Chijie,Zhang Bin,Hu Jun,et al.Anomaly detection for power consumption patterns based on unsupervised learning[J].Proceedings of the CSEE,2016,36(2):379-387(in Chinese).
[12]赵莉,候兴哲,胡君,等.基于改进K-means算法的海量智能用电数据分析[J].电网技术,2014,38(10):2715-2720.Zhao Li,Hou Xingzhe,Hu Jun,et al.Improved K-means algorithm based analysis on massive data of intelligent power utilization[J].Power System Technology,2014,38(10):2715-2720(in Chinese).
[13]和敬涵,卢育梓,陆金耀,等.基于“进化”主成分分析法的用户分类及其应用[J].电力建设,2017,38(3):101-107.He Jinghan,Lu Yuzi,Lu Jinyao,et al.User classification method based on’evolution’PCA and its application[J].Electric Power Construction,2017,38(3):101-107(in Chinese).
[14]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.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(in Chinese).
[15]王德文,孙志伟.一种基于内存计算的电力用户聚类分析方法[J].智能系统学报,2015,10(4):569-576.Wang Dewen,Sun Zhiwei.A method for cluster analysis of electric power consumers based on in-memory computing[J].CAAI Transactions on Intelligent Systems,2015,10(4):569-576(in Chinese).
[16]朱文俊,王毅,罗敏,等.面向海量用户用电特性感知的分布式聚类算法[J].电力系统自动化,2016,40(12):21-27.Zhu Wenjun,Wang Yi,Luo Min,et al.Distributed clustering algorithm for awareness of electricity consumption characteristics of massive consumers[J].Automation of Electric Power Systems,2016,40(12):21-27(in Chinese).
[17]张进,徐苒茨,徐健,等.居民用户用电的聚类分析[J].电力与能源,2016,37(2):180-184.Zhang Jin,Xu Ranci,Xu Jian,et al.Clustering analysis of residential electricity utilization[J].Power&Energy,2016,37(2):180-184(in Chinese).
[18]Huang G,Song S J,Gupta J N D,et al.Semi-supervised and unsupervised extreme learning machines[J].IEEE Transactions on Cybernetics,2014,44(12):2405-2417.
[19]Huang G B,Zhu Q Y,Siew C K.Extreme learning machine:Theory and applications[J].Neurocomputing,2006,70(1-3):489-501.