基于小波变换和改进快速密度峰值聚类算法的负荷曲线聚类研究
|
摘要
负荷曲线聚类分析是智能电网大数据研究的重要组成部分,是负荷预测、需求侧响应、电网规划、经济运行、费率制定、能效提升等研究与工作的基础。文章利用离散小波变换提取用户负荷数据多时间尺度特征,进而对负荷数据进行聚类分析。针对快速密度峰值聚类算法中局部密度依赖于截断距离和需要人为识别决策图中聚类中心的不足,利用K近邻(K-Nearest Neighbors,KNN)算法的思想重新定义局部密度和距离,并根据向外统计检验的方法实现聚类中心的自动选取。基于某省某行业用户实际负荷数据进行实验,结果将该行业负荷曲线分为正常生产型、双峰型、夜晚生产型、白天生产型、晚高峰型、早高峰型6类,表明了该算法的有效性。
Load curve clustering is an important part of smart grid big data research, which is the basis of load forecasting, demand response, grid planning and operation, rate setting, energy efficiency and so on. In this paper, discrete wavelet transform(DWT) is used to extract the multi-time scale characteristics of load profile, and then clustering analysis is conducted. In order to overcome the shortcomings in density peaks clustering algorithm, i) the local density is sensitive to the cut-off distance and ii) clustering centers identification needs users to preassign two minimum thresholds, the idea of KNN algorithm is used to redefine the local density and distance, and the outward statistical test method is adopted to identify clustering centers automatically. The test results using the given industry real load data in a province show that the industry curves are divided into six types: normal production type, double peaks type, night production type, daytime production type, evening peak type and morning peak type, which shows that the algorithm is effective.
Load curve clustering is an important part of smart grid big data research, which is the basis of load forecasting, demand response, grid planning and operation, rate setting, energy efficiency and so on. In this paper, discrete wavelet transform(DWT) is used to extract the multi-time scale characteristics of load profile, and then clustering analysis is conducted. In order to overcome the shortcomings in density peaks clustering algorithm, i) the local density is sensitive to the cut-off distance and ii) clustering centers identification needs users to preassign two minimum thresholds, the idea of KNN algorithm is used to redefine the local density and distance, and the outward statistical test method is adopted to identify clustering centers automatically. The test results using the given industry real load data in a province show that the industry curves are divided into six types: normal production type, double peaks type, night production type, daytime production type, evening peak type and morning peak type, which shows that the algorithm is effective.
引文
[1]陈毅波,陈乾,眭建新.基于大数据技术的电网运营分析决策系统研究[J].电力信息与通信技术,2015,13(8):128-131.CHEN Yi-bo,CHEN Qian,MU Jian-xin.Researched on power grid operation analysis and decision-making system based on big data technology[J].Electric Power Information and Communication Technology,2015,13(8):128-131.
[2]王继业.大数据在电网企业的应用探索[J].中国电力企业管理2015(17):18-21.
[3]王继业,季知祥,史梦洁,等.智能配用电大数据需求分析与应用研究[J].中国电机工程学报,2015,35(8):1829-1836.WANG Ji-ye,JI Zhi-xiang,SHI Meng-jie,et al.Scenario analysis and application research on big data in smart power distribution and consumption systems[J].Proceedings of the CSEE,2015,35(8):1829-1836.
[4]CHANG R F,LU C N.Load profiling and its applications in power market[C]//Proceedings of the Power Engineering Society General Meeting,2003,IEEE,2003:974-978.
[5]全思平,吴浩.基于数据挖掘技术的短期电力负荷预测[J].电力信息与通信技术,2014,12(3):95-98.QUAN Si-ping,WU Hao.A short-term power load forecasting system using the data mining technology[J].Electric Power Information and Communication Technology,2014,12(3):95-98.
[6]江樱,王志强,戴波.基于大数据的居民用电消费习惯研究与分析[J].电力信息与通信技术,2015,13(11):7-11.JIANG Ying,WANG Zhi-qiang,DAI Bo.Research and analysis of residential electricity consumption behavior based on big data[J].Electric Power Information and Communication Technology,2015,13(11):7-11.
[7]WANG Y,CHEN Q,KANG C,et al.Load profiling and its application to demand response:A review[J].Tsinghua Science and Technology,2015,20(2):117-129.
[8]METS K,DEPUYDT F,DEVELDER C.Two-stage load pattern clustering using fast wavelet transformation[J].IEEETransactions on Smart Grid,2016,7(5):2250-2259.
[9]LI R,LI F,SMITH N D.Multi-resolution load profile clustering for smart metering data[J].IEEE Transactions on Power Systems,2016,31(6):4473-4482.
[10]ZHOU K L,YANG S L,SHEN C.A review of electric load classification in smart grid environment[J].Renewable and Sustainable Energy Reviews,2013(24):103-110.
[11]RODRIGUEZ A,LAIO A.Clustering by fast search and find of density peaks[J].Science,2014,344(6191):1492-1496.
[12]XIE J,GAO H,XIE W,et al.Robust clustering by detecting density peaks and assigning points based on fuzzy weighted K-nearest neighbors[J].Information Sciences,2016(354):19-40.
[13]DU M,DING S,JIA H.Study on density peaks clustering based on k-nearest neighbors and principal component analysis[J].Knowledge-Based Systems,2016(99):135-145.
[14]WANG G,SONG Q.Automatic clustering via outward statistical testing on density metrics[J].IEEE Transactions on Knowledge and Data Engineering,2016,28(8):1971-1985.
[2]王继业.大数据在电网企业的应用探索[J].中国电力企业管理2015(17):18-21.
[3]王继业,季知祥,史梦洁,等.智能配用电大数据需求分析与应用研究[J].中国电机工程学报,2015,35(8):1829-1836.WANG Ji-ye,JI Zhi-xiang,SHI Meng-jie,et al.Scenario analysis and application research on big data in smart power distribution and consumption systems[J].Proceedings of the CSEE,2015,35(8):1829-1836.
[4]CHANG R F,LU C N.Load profiling and its applications in power market[C]//Proceedings of the Power Engineering Society General Meeting,2003,IEEE,2003:974-978.
[5]全思平,吴浩.基于数据挖掘技术的短期电力负荷预测[J].电力信息与通信技术,2014,12(3):95-98.QUAN Si-ping,WU Hao.A short-term power load forecasting system using the data mining technology[J].Electric Power Information and Communication Technology,2014,12(3):95-98.
[6]江樱,王志强,戴波.基于大数据的居民用电消费习惯研究与分析[J].电力信息与通信技术,2015,13(11):7-11.JIANG Ying,WANG Zhi-qiang,DAI Bo.Research and analysis of residential electricity consumption behavior based on big data[J].Electric Power Information and Communication Technology,2015,13(11):7-11.
[7]WANG Y,CHEN Q,KANG C,et al.Load profiling and its application to demand response:A review[J].Tsinghua Science and Technology,2015,20(2):117-129.
[8]METS K,DEPUYDT F,DEVELDER C.Two-stage load pattern clustering using fast wavelet transformation[J].IEEETransactions on Smart Grid,2016,7(5):2250-2259.
[9]LI R,LI F,SMITH N D.Multi-resolution load profile clustering for smart metering data[J].IEEE Transactions on Power Systems,2016,31(6):4473-4482.
[10]ZHOU K L,YANG S L,SHEN C.A review of electric load classification in smart grid environment[J].Renewable and Sustainable Energy Reviews,2013(24):103-110.
[11]RODRIGUEZ A,LAIO A.Clustering by fast search and find of density peaks[J].Science,2014,344(6191):1492-1496.
[12]XIE J,GAO H,XIE W,et al.Robust clustering by detecting density peaks and assigning points based on fuzzy weighted K-nearest neighbors[J].Information Sciences,2016(354):19-40.
[13]DU M,DING S,JIA H.Study on density peaks clustering based on k-nearest neighbors and principal component analysis[J].Knowledge-Based Systems,2016(99):135-145.
[14]WANG G,SONG Q.Automatic clustering via outward statistical testing on density metrics[J].IEEE Transactions on Knowledge and Data Engineering,2016,28(8):1971-1985.