基于优化SAX和带权负荷特性指标的AP聚类用户用电行为分析
|
摘要
智能电表的推广和安装,使用户侧累积了海量用电数据。特征提取和聚类分析作为有效的数据处理手段,有助于挖掘用电数据中隐藏的宝贵信息,提取用户的用电行为特性。为提取有效直观的负荷特性,本文提出利用优化SAX和带权负荷指标的AP聚类算法,对负荷曲线进行聚类。针对AP聚类复杂度较高的问题,首先利用SAX算法对负荷曲线进行降维并提取特征,利用基于模拟退火粒子群算法,优化确定合理的字符数和状态数;然后结合负荷特性指标,运用改进AP聚类算法,对负荷曲线进行聚类,聚类过程中采用熵权法对负荷特性指标进行客观赋权,避免指标设置的主观性。基于聚类结果,对各类用户的用电行为以及需求响应潜力进行分析。案例分析验证了该算法的高效性和有效性,并可应用于电网公司决策,如负荷预测、异常检测和提供差异化服务等。
The installation of smart meters has resulted in the accumulation of massive electricity data at the demand side. Feature extraction and clustering analysis, as effective data processing means,can help utilities to mine the valuable information hidden in the data and extract customer behavior characteristics. In order to extract effective and intuitive load characteristics, this paper proposes a clustering algorithm based on affinity propagation(AP) clustering algorithm with optimized symbolic aggregate approximation(SAX) and weighted load characteristic indices. First, to solve the problem of high complexity, SAX algorithm is applied to reduce the dimension of load curves, and the appropriate symbolic representation scheme is obtained by simulated annealing particle swarm optimization. Then,combined with the load characteristic indices, the improved AP clustering algorithm is utilized to cluster load curves. In addition, to avoid the subjectivity of the index setting, entropy weighting method is used to objectively weight the load characteristic indices. Based on the clustering results, the consumption behavior and demand response potential of different customers are analyzed. Case studies indicate that the proposed methods and approaches is efficient and effective, which can be applied in utilities for decision making, such as load forecasting, anomaly detection, differential services, etc.
The installation of smart meters has resulted in the accumulation of massive electricity data at the demand side. Feature extraction and clustering analysis, as effective data processing means,can help utilities to mine the valuable information hidden in the data and extract customer behavior characteristics. In order to extract effective and intuitive load characteristics, this paper proposes a clustering algorithm based on affinity propagation(AP) clustering algorithm with optimized symbolic aggregate approximation(SAX) and weighted load characteristic indices. First, to solve the problem of high complexity, SAX algorithm is applied to reduce the dimension of load curves, and the appropriate symbolic representation scheme is obtained by simulated annealing particle swarm optimization. Then,combined with the load characteristic indices, the improved AP clustering algorithm is utilized to cluster load curves. In addition, to avoid the subjectivity of the index setting, entropy weighting method is used to objectively weight the load characteristic indices. Based on the clustering results, the consumption behavior and demand response potential of different customers are analyzed. Case studies indicate that the proposed methods and approaches is efficient and effective, which can be applied in utilities for decision making, such as load forecasting, anomaly detection, differential services, etc.
引文
[1]朱永利,李莉,宋亚奇,等.ODPS平台下的电力设备监测大数据存储与并行处理方法[J].电工技术学报,2017,32(9):199-210.Zhu Yongli,Li Li,Song Yaqi,et al.Storage and parallel processing of big data of power equipment condition monitoring on ODPS platform[J].Transactions of China Electrotechnical Society,2017,32(9):199-210.
[2]Wang Yi,Chen Qixin,Hong Tao,et al.Review of smart meter data analytics:applications,methodologies,and challenges[J].IEEE Transactions on Smart Grid,2018,DOI:10.1109/TSG.2018.2818167.
[3]吴金浩,杨秀媛,孙骏.基于主成分分析法的风电功率短期组合预测[J].电气技术,2016,17(7):41-47.Wu Jinhao,Yang Xiuyuan,Sun Jun.The combination forecasting model for wind farm power based on PCA[J].Electrical Engineering,2016,17(7):41-47.
[4]邹经鑫,陈伟根,万福,等.油纸绝缘老化拉曼光谱特征量提取及诊断方法[J].电工技术学报,2018,33(5):1133-1142.Zou Jingxin,Chen Weigen,Wan Fu,et al.The raman spectral feature extraction and diagnosis of oil-paper insulation ageing[J].Transactions of China Electrotechnical Society,2018,33(5):1133-1142.
[5]Frey B J,Dueck D.Clustering by passing messages between data points[J].Science,2007,315(5814):972-976.
[6]Lin J,Keogh E,Lonardi S,et al.A symbolic representation of time series,with implications for streaming algorithms[C]//ACM SIGMOD Workshop on Research Issues in Data Mining and Knowledge Discovery,2003:2-11.
[7]Wang Yi,Chen Qixin,Kang Chongqing,et al.Clustering of electricity consumption behavior dynamics toward big data applications[J].IEEETransactions on Smart Grid,2016,7(5):2437-2447.
[8]Pham N D,Le Q L,Dang T K.Two novel adaptive symbolic representations for similarity search in time series databases[C]//Web Conference,IEEE,Busan,South Korea,2010:181-187.
[9]Ahmed A M,Bakar A A,Hamdan A R.Harmony Search algorithm for optimal word size in symbolic time series representation[C]//Data Mining and Optimization,IEEE,Putrajaya,Malaysia,2011:57-62.
[10]Bondu A,BoulléM,Cornuéjols A.Symbolic representation of time series:a hierarchical coclustering formalization[C]//International Workshop on Advanced Analytics and Learning on Temporal Data,Springer International Publishing,Porto,Portugal,2015:3-16.
[11]Rechy-Ramírez F,Mesa H G A,Mezura-Montes E,et al.Times series discretization using evolutionary programming[C]//International Conference on Artificial Intelligence:Advances in Soft Computing,SpringerVerlag,Guadalajara,Mexico,2011:225-234.
[12]Marquez-Grajales A,Acosta-Mesa H G,MezuraMontes E.An adaptive symbolic discretization scheme for the classification of tempo ral datasets using NSGA-II[C]//IEEE International Autumn Meeting on Power,Electronics and Computing,IEEEIxtapa,Mexico,2017:1-8.
[13]Fuad M M M.Variable-chromosome-length genetic algorithm for time series discretization[M].Database and Expert Systems Applications.Springer International Publishing,Regensburg,Germany,2016:418-425.
[14]Acosta-Mesa H G,Cruz-Ramírez N,García-Lopez DA.Entropy based linear approximation algorithm for time series discretization.advances in artificial intelligence and applications[J].Research in Computers,2008,32:214-224.
[15]Aghabozorgi S,Shirkhorshidi A S,Wah T Y.Timeseries clustering-a decade review[J].Information Systems,2015,53(C):16-38.
[16]Al-Otaibi R,Jin N,Wilcox T,et al.Feature construction and calibration for clustering daily load curves from smart-meter data[J].IEEE Transactions on Industrial Informatics,2017,12(2):645-654.
[17]李欣然,徐振华,宋军英,等.基于功率空间的分时段负荷模型参数在线修正[J].电工技术学报,2012,27(8):147-156.Li Xinran,Xu Zhenhua,Song Junying,et al.On-line revising algorithm for load model parameters of substation in different daily periods based on the measured active power[J].Transactions of China Electrotechnical Society,2012,27(8):147-156.
[18]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.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.
[19]龚钢军,陈志敏,陆俊,等.智能用电用户行为分析的聚类优选策略[J].电力系统自动化,2018,42(2):58-63.Gong Gangjun,Chen Zhimin,Lu Jun,et al.Clustering optimization strategy for electricity consumption behavior analysis in smart grid[J].Automation of Electric Power Systems,2018,42(2):58-63.
[20]沈小军,周冲成,吕洪.基于运行数据的风电机组间风速相关性统计分析[J].电工技术学报,2017,32(16):265-274.Shen Xiaojun,Zhou Chongcheng,LüHong.Statistical analysis of wind speed correlation between wind turbines based on operational data[J].Transactions of China Electrotechnical Society,201732(16):265-274.
[21]孙建军,张世泽,曾梦迪,等.考虑分时电价的主动配电网柔性负荷多目标优化控制[J].电工技术学报,2018,33(2):401-412.Sun Jianjun,Zhang Shize,Zeng Mengdi,et al.Multi-objective optimal control for flexible load in active distribution network considering time-of-use tariff[J].Transactions of China Electrotechnical Society,2018,33(2):401-412.
[22]黄俊辉,汪惟源,王海潜,等.基于模拟退火遗传算法的交直流系统无功优化与电压控制研究[J].电力系统保护与控制,2016,44(10):37-43.Huang Junhui,Wang Weiyuan,Wang Haiqian,et al.Study of hybrid genetic algorithm and annealing algorithm on reactive power optimization and voltage control in AC/DC transmission system[J].Power System Protection and Control,2016,44(10):37-43.
[23]Commission for Energy Regulation.CER Smart Metering Project[EB/OL].2012,http://www.ucd.ie/issda/data/commissionforenergyregulationcer/
[2]Wang Yi,Chen Qixin,Hong Tao,et al.Review of smart meter data analytics:applications,methodologies,and challenges[J].IEEE Transactions on Smart Grid,2018,DOI:10.1109/TSG.2018.2818167.
[3]吴金浩,杨秀媛,孙骏.基于主成分分析法的风电功率短期组合预测[J].电气技术,2016,17(7):41-47.Wu Jinhao,Yang Xiuyuan,Sun Jun.The combination forecasting model for wind farm power based on PCA[J].Electrical Engineering,2016,17(7):41-47.
[4]邹经鑫,陈伟根,万福,等.油纸绝缘老化拉曼光谱特征量提取及诊断方法[J].电工技术学报,2018,33(5):1133-1142.Zou Jingxin,Chen Weigen,Wan Fu,et al.The raman spectral feature extraction and diagnosis of oil-paper insulation ageing[J].Transactions of China Electrotechnical Society,2018,33(5):1133-1142.
[5]Frey B J,Dueck D.Clustering by passing messages between data points[J].Science,2007,315(5814):972-976.
[6]Lin J,Keogh E,Lonardi S,et al.A symbolic representation of time series,with implications for streaming algorithms[C]//ACM SIGMOD Workshop on Research Issues in Data Mining and Knowledge Discovery,2003:2-11.
[7]Wang Yi,Chen Qixin,Kang Chongqing,et al.Clustering of electricity consumption behavior dynamics toward big data applications[J].IEEETransactions on Smart Grid,2016,7(5):2437-2447.
[8]Pham N D,Le Q L,Dang T K.Two novel adaptive symbolic representations for similarity search in time series databases[C]//Web Conference,IEEE,Busan,South Korea,2010:181-187.
[9]Ahmed A M,Bakar A A,Hamdan A R.Harmony Search algorithm for optimal word size in symbolic time series representation[C]//Data Mining and Optimization,IEEE,Putrajaya,Malaysia,2011:57-62.
[10]Bondu A,BoulléM,Cornuéjols A.Symbolic representation of time series:a hierarchical coclustering formalization[C]//International Workshop on Advanced Analytics and Learning on Temporal Data,Springer International Publishing,Porto,Portugal,2015:3-16.
[11]Rechy-Ramírez F,Mesa H G A,Mezura-Montes E,et al.Times series discretization using evolutionary programming[C]//International Conference on Artificial Intelligence:Advances in Soft Computing,SpringerVerlag,Guadalajara,Mexico,2011:225-234.
[12]Marquez-Grajales A,Acosta-Mesa H G,MezuraMontes E.An adaptive symbolic discretization scheme for the classification of tempo ral datasets using NSGA-II[C]//IEEE International Autumn Meeting on Power,Electronics and Computing,IEEEIxtapa,Mexico,2017:1-8.
[13]Fuad M M M.Variable-chromosome-length genetic algorithm for time series discretization[M].Database and Expert Systems Applications.Springer International Publishing,Regensburg,Germany,2016:418-425.
[14]Acosta-Mesa H G,Cruz-Ramírez N,García-Lopez DA.Entropy based linear approximation algorithm for time series discretization.advances in artificial intelligence and applications[J].Research in Computers,2008,32:214-224.
[15]Aghabozorgi S,Shirkhorshidi A S,Wah T Y.Timeseries clustering-a decade review[J].Information Systems,2015,53(C):16-38.
[16]Al-Otaibi R,Jin N,Wilcox T,et al.Feature construction and calibration for clustering daily load curves from smart-meter data[J].IEEE Transactions on Industrial Informatics,2017,12(2):645-654.
[17]李欣然,徐振华,宋军英,等.基于功率空间的分时段负荷模型参数在线修正[J].电工技术学报,2012,27(8):147-156.Li Xinran,Xu Zhenhua,Song Junying,et al.On-line revising algorithm for load model parameters of substation in different daily periods based on the measured active power[J].Transactions of China Electrotechnical Society,2012,27(8):147-156.
[18]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.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.
[19]龚钢军,陈志敏,陆俊,等.智能用电用户行为分析的聚类优选策略[J].电力系统自动化,2018,42(2):58-63.Gong Gangjun,Chen Zhimin,Lu Jun,et al.Clustering optimization strategy for electricity consumption behavior analysis in smart grid[J].Automation of Electric Power Systems,2018,42(2):58-63.
[20]沈小军,周冲成,吕洪.基于运行数据的风电机组间风速相关性统计分析[J].电工技术学报,2017,32(16):265-274.Shen Xiaojun,Zhou Chongcheng,LüHong.Statistical analysis of wind speed correlation between wind turbines based on operational data[J].Transactions of China Electrotechnical Society,201732(16):265-274.
[21]孙建军,张世泽,曾梦迪,等.考虑分时电价的主动配电网柔性负荷多目标优化控制[J].电工技术学报,2018,33(2):401-412.Sun Jianjun,Zhang Shize,Zeng Mengdi,et al.Multi-objective optimal control for flexible load in active distribution network considering time-of-use tariff[J].Transactions of China Electrotechnical Society,2018,33(2):401-412.
[22]黄俊辉,汪惟源,王海潜,等.基于模拟退火遗传算法的交直流系统无功优化与电压控制研究[J].电力系统保护与控制,2016,44(10):37-43.Huang Junhui,Wang Weiyuan,Wang Haiqian,et al.Study of hybrid genetic algorithm and annealing algorithm on reactive power optimization and voltage control in AC/DC transmission system[J].Power System Protection and Control,2016,44(10):37-43.
[23]Commission for Energy Regulation.CER Smart Metering Project[EB/OL].2012,http://www.ucd.ie/issda/data/commissionforenergyregulationcer/