基于多特征和LOF的用户负荷突变检测Abrupt user load change detection based on multiple features and LOF algorithm
曾静,娄冰,吕娜,邓隽,王冠明
ZENG Jing,LOU Bing,LYU Na,DENG Jun,WANG Guanming
摘要(Abstract)：
负荷突变对电网冲击力大，会造成电网频率和功率振荡。为了对复杂且大体量的用户负荷异常数据进行排查，提出了多特征与LOF（局部离群因子）算法相结合的检测方法。提取负荷数据统计特征平均值、标准差以及波形特征离散系数、峭度、波形因子和脉冲因子，经过PCA（主成分分析）降维后获得有效特征，并采用LOF算法对每天的用户负荷异常数据进行检测。此检测算法在以阿里云为基础的浙电数据中台中运行，结果表明所提方法能够每天定时在海量量测数据中将负荷突变的用户查找出来，实现了在线检测并具有较高的准确率。
The sudden load changes impact power grids by frequency and power oscillations. In order to distinguish the complex and massive abnormal user load data, this paper proposes a method combining multiple features and LOF(local outlier factor) algorithm. Firstly, the statistical characteristic mean value, standard deviation, waveform characteristic dispersion coefficient, kurtosis, waveform factor and pulse factor of load data are extracted, and the effective features are obtained through dimensionality reduction of PCA(principal component analysis). Furthermore, the LOF algorithm is used to detect abnormal user load data every day. This detection algorithm is used in the Zhejiang power data center based on Alibaba cloud. The results show that it can detect users with abrupt load changes in massive measured data at fixed times of every day and realizes online detection with high accuracy.
关键词(KeyWords)：
机器学习;LOF算法;负荷突变;大数据
machine learning;LOF algorithm;abrupt load change;big data
基金项目(Foundation):
作者(Author):
曾静,娄冰,吕娜,邓隽,王冠明
ZENG Jing,LOU Bing,LYU Na,DENG Jun,WANG Guanming
DOI: 10.19585/j.zjdl.202302012
参考文献(References)：
- [1]方凯杰，黄奇峰，杨世海，等.一种非侵入式居民负荷突变事件三阈值检测方法和系统：CN112396007B[P].2021-08-24.
- [2]黄建元.电力系统中配电线路运行故障检测技术[J].电子乐园，2019(16):278.
- [3]何尧，梁宏池，连鸿松，等.基于滑动窗口和多元高斯分布的变压器油色谱异常值检测[J].高压电器，2020,56(1):203-209.HE Yao,LIANG Hongchi,LIAN Hongsong,et al.Outlier detection of power transformer oil chromatographic data based on algorithm sliding windows and multivariate Gaussian distribution[J].High Voltage Apparatus,2020,56(1):203-209.
- [4]严英杰，盛戈皞，刘亚东，等.基于滑动窗口和聚类算法的变压器状态异常检测[J].高电压技术，2016,42(12):4020-4025.YAN Yingjie,SHENG Gehao,LIU Yadong,et al.Anomalous state detection of power transformer based on algorithm sliding windows and clustering[J].High Voltage Engineering,2016,42(12):4020-4025.
- [5]钱宇骋，甄超，季坤，等.变压器在线监测数据异常值检测与清洗[J].哈尔滨理工大学学报，2020,25(5):15-22.QIAN Yucheng,ZHEN Chao,JI Kun,et al.Transformer online monitoring data abnormal value detection and cleaning[J].Journal of Harbin University of Science and Technology,2020,25(5):15-22.
- [6]文旭，王浩，黄刚，等.基于因子分析的母线负荷异常数据辨识方法[J].重庆大学学报，2021,44(8):91-102.WEN Xu,WANG Hao,HUANG Gang,et al.Identification method of abnormal data in bus load based on factor analysis[J]. Journal of Chongqing University,2021,44(8):91-102.
- [7]熊燕，李艳红.基于数据挖掘的电力系统异常数据辨识与调整[J].安徽电气工程职业技术学院学报，2010,15(3):11-15.XIONG Yan,LI Yanhong. Power system identification and adjustment of bad data based on data mining[J].Journal of Anhui Electrical Engineering Professional Technique College,2010,15(3):11-15.
- [8]李航.基于LOF的快速密度峰值聚类的电力数据异常值检测方法研究[D].兰州：兰州理工大学，2019.LI Hang.Research on detection method of outlier value of power data based on fast density peak clustering and LOF[D].Lanzhou:Lanzhou University of Technology,2019.
- [9]刘冬兰，马雷，刘新，等.基于深度学习的电力大数据融合与异常检测方法[J].计算机应用与软件，2018,35(4):61-64.LIU Donglan,MA Lei,LIU Xin,et al. Deep learning based anomaly detection approach for power big data[J].Computer Applications and Software,2018,35(4):61-64.
- [10]孙胜博，张凯，冯剑，等.基于负荷特性分析的电力用户用电行为特征研究[J].内蒙古电力技术，2020,38(1):40-44.SUN Shengbo,ZHANG Kai,FENG Jian,et al.Research on characteristics of customer electricity behavior based on load characteristics[J]. Inner Mongolia Electric Power,2020,38(1):40-44.
- [11]肖白，徐潇，宋坤，等.空间电力负荷预测中异常数据的辨识与处理[J].东北电力大学学报，2013,33（增刊1）:45-50.XIAO Bai,XU Xiao,SONG Kun,et al. Abnormal data identification and treatment in spatial electric load forecasting[J]. Journal of Northeast Dianli University,2013,33(S1):45-50.
- [12]赵振翔.模块化的多机牵引提速道岔转辙机智能测试仪的研究与设计[D].兰州：兰州交通大学，2018.ZHAO Zhenxiang. Research and design of a smart high speed railway point system based on S700K[D].Lanzhou:Lanzhou Jiatong University,2018.
- [13]孙滢涛，张锋明，陈水标，等.基于多域特征提取的电力数据异常检测方法[J].电力系统及其自动化学报，2022,34(6):105-113.SUN Yingtao,ZHANG Fengming,CHEN Shuibiao,et al.Power data anomaly detection algorithm based on multidomain feature extraction[J]. Proceedings of the CSUEPSA,2022,34(6):105-113.
- [14]张美林，李俊萩，张晴晖，等.基于熵和波形特征的木材损伤断裂过程声发射信号处理[J].林业工程学报，2022,7(2):159-166.ZHANG Meilin,LI Junqiu,ZHANG Qinghui,et al.Acoustic emission signal processing and analysis of wood damage and fracture process based on entropy and waveform characteristics[J]. Journal of Forestry Engineering,2022,7(2):159-166.
- [15]KIRBY M,SIROVICH L. Application of the KarhunenLoeve procedure for the characterization of human faces[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1990,12(1):103-108.
- [16]张戈，盖赟.局部离群因子算法（LOF）在异常检测中的应用研究[J].网络安全技术与应用，2020(11):49-50.ZHANG Ge,GAI Yun. Research on the application of LOF in anomaly detection[J].Network Security Technology&Application,2020(11):49-50.
- [17]BREUNIG M M,KRIEGEL H P,NG R T,et al.LOF:identifying density-based local outliers[C]//Proceedings of the 2000 ACM SIGMOD international conference on Management of data,16 May 2000,Dallas,Texas,USA.ACM,2000:93-104.
- [18]周莹.基于MaxCompute的大数据云服务平台的设计与实现[D].武汉：华中科技大学，2017.ZHOU Ying.Design and implementation of big data cloud service platform based on MaxCompute[D].Wuhan:Huazhong University of Science and Technology,2017.