刘洋,于海东,刘文彬,黄敏,李立生,张世栋

• 1:国网山东省电力公司电力科学研究院
• 2:山东省智能电网技术创新中心

摘要(Abstract)：

设备故障、天气环境等因素导致分布式光伏发电系统产生大量异常数据，对其安全稳定运行造成严重影响。为了准确识别和剔除存在的异常数据，提出一种基于动态时间弯曲（DTW）-两阶四分位的分布式光伏发电异常数据辨识方法。首先，通过对比相似辐照度下光伏功率均值实现连续型异常数据识别与剔除，采用基于同时段光伏功率均值剔除异常数据，并考虑光伏发电曲线的波动性，采用基于DTW与欧氏距离的综合曲线相似度判定方法剔除连续型异常数据，更全面地考虑了数据的波动特性，提高了连续型异常数据辨识和剔除效果；其次，提出DTW-两阶四分位异常数据辨识算法，采用一阶变化率和二阶变化率对融合后的数据进行离散型异常数据剔除，有效识别和剔除离散型异常数据；最后，根据异常数据识别和剔除结果判断是否出现故障。实验结果表明：所提算法剔除异常数据后能更好地拟合正常光伏功率数据分布情况，相较于四分位法和3-Sigma算法，所提算法剔除异常数据前后线性相关程度变化分别提高了58.15%和68.41%，辨识效果更佳。

关键词(KeyWords)： 分布式光伏;异常数据辨识;动态时间弯曲;两阶四分位

Abstract:

Keywords:

基金项目(Foundation): 国家电网有限公司科技项目（520626210014）~~

作者(Author): 刘洋,于海东,刘文彬,黄敏,李立生,张世栋

DOI: 10.19666/j.rlfd.202402037

参考文献(References)：

• [1]陈骏嚎,张娜,刘广忱,等.基于EOSSA-ELM的光伏短期输出功率预测[J].可再生能源, 2022, 40(7):890-898.CHEN Junhao, ZHANG Na, LIU Guangchen, et al.Photovoltaic short-term output power forecasting based on EOSSA-ELM[J]. Renewable Energy Resources, 2022,40(7):890-898.
• [2]石珵,刘朋矩,杜治钢,等.数字孪生辅助的智能楼宇多模态通信资源管理方法[J].电信科学, 2023, 39(1):60-71.SHI Cheng, LIU Pengju, DU Zhigang, et al. Digital twinassisted multi-mode communication resource management methods for smart buildings[J].Telecommunications Science, 2023, 39(1):60-71.
• [3]王超,胡浩,郑炼,等.基于智慧蓄电池的光伏储能系统及其控制策略[J].可再生能源, 2022, 40(4):506-512.WANG Chao, HU Hao, ZHENG Lian, et al. Photovoltaic energy storage system based on smart battery and its control strategy[J]. Renewable Energy Resources, 2022,40(4):506-512.
• [4]张娜,任强,刘广忱,等.基于VMD-GWO-ELMAN的光伏功率短期预测方法[J].中国电力, 2022, 55(5):57-65.ZHANG Na, REN Qiang, LIU Guangchen, et al. PV power short-term forecasting method based on VMDGWO-ELMAN[J]. Electric Power, 2022, 55(5):57-65.
• [5]张翔,王瞾,周振宇,等.考虑光伏集成5G基站用能模式的多主体共享储能优化配置[J].电测与仪表,2023, 60(3):97-106.ZHANG Xiang, WANG Zhao, ZHOU Zhenyu, et al.Optimal configuration of shared energy storage system for multi-entities considering PV integrated 5G base station energy consumption mode[J]. Electrical Measurement&Instrumentation, 2023, 60(3):97-106.
• [6]时珉,尹瑞,胡傲宇,等.基于滑动标准差计算的光伏阵列异常数据清洗办法[J].电力系统保护与控制,2020, 48(6):108-114.SHI Min, YIN Rui, HU Aoyu, et al. A novel photovoltaic array outlier cleaning algorithm based on moving standard deviation[J]. Power System Protection and Control, 2020,48(6):108-114.
• [7]张翔,廖海君,周振宇,等.含规模化5G基站租赁共享储能的配电网混合博弈优化调度[J].电测与仪表,2023, 60(5):23-32.ZHANG Xiang, LIAO Haijun, ZHOU Zhenyu, et al.Hybrid game optimal dispatching for distribution network with large-scale 5G base station leasing shared energy storage[J]. Electrical Measurement&Instrumentation,2023, 60(5):23-32.
• [8]吴家葆,曾国辉,张振华,等.基于K-means分层聚类的TCN-GRU和LSTM动态组合光伏短期功率预测[J].可再生能源, 2023, 41(8):1015-1022.WU Jiabao, ZENG Guohui, ZHANG Zhenhua, et al.Dynamic combination of TCN-GRU and LSTM photovoltaic short-term power prediction based on Kmeans hierarchical clustering[J]. Renewable Energy Resources, 2023, 41(8):1015-1022.
• [9]刘岩,李文文,周丽霞,等.基于高斯混合模型的光伏发电出力中高比例异常数据检测方法研究[J].电测与仪表, 2021, 58(9):14-21.LIU Yan, LI Wenwen, ZHOU Lixia, et al. Research of high proportion of outliers detection method in photovoltaic power output data based on Gaussian mixture model[J]. Electrical Measurement&Instrumentation, 2021, 58(9):14-21.
• [10]郭泽阳,程拥军,温家铭,等.基于支持向量机的分布式光伏异常数据序列提取[J].中国测试, 2023, 49(7):156-161.GUO Zeyang, CHENG Yongjun, WEN Jiaming, et al.Extraction of abnormal data sequence of photovoltaic power station based on support vector machine[J]. China Measurement&Test, 2023, 49(7):156-161.
• [11]赵晋斌,张建平,毛玲,等.基于PSO-Soft attention双向LSTM算法的光伏发电量预测研究[J].智慧电力,2022, 50(3):1-7.ZHAO Jinbin, ZHANG Jianping, MAO Ling, et al.Photovoltaic power generation forecasting based on PSOsoft attention bidirectional LSTM algorithm[J]. Smart Power, 2022, 50(3):1-7.
• [12]叶林,崔宝丹,李卓,等.分布式光伏高比例异常运行数据组合识别方法[J].电力系统自动化, 2022, 46(20):74-82.YE Lin, CUI Baodan, LI Zhuo, et al. Combined identification method for high proportion of abnormal operation data in photovoltaic power station[J].Automation of Electric Power Systems, 2022, 46(20):74-82.
• [13]梅玉杰,李勇,周王峰,等.基于机器学习的配电网异常缺失数据动态清洗方法[J].电力系统保护与控制,2023, 51(7):158-169.MEI Yujie, LI Yong, ZHOU Wangfeng, et al. Dynamic data cleaning method of abnormal and missing data in a distribution network based on machine learning[J]. Power System Protection and Control, 2023, 51(7):158-169.
• [14]黄颖祺.基于机器学习的用户侧异常数据诊断研究[J].制造业自动化, 2023, 45(8):49-55.HUANG Yingqi. Research on user-side data abnormal diagnosis based on machine learning[J]. Manufacturing Automation, 2023, 45(8):49-55.
• [15]刘丰硕,刘然,董子慧.基于k-means聚类的智能电表异常动态数据检测方法[J].电子设计工程, 2023,31(5):84-88.LIU Fengshuo, LIU Ran, DONG Zihui. Abnormal dynamic data detection method for smart meters based on k-means clustering[J]. Electronic Design Engineering,2023, 31(5):84-88.
• [16]赵天辉,张耀,王建学.基于空间密度聚类和异常数据域的负荷异常值识别方法[J].电力系统自动化,2021, 45(10):97-105.ZHAO Tianhui, ZHANG Yao, WANG Jianxue.Identification method of load outlier based on densitybased spatial clustering and outlier boundaries[J].Automation of Electric Power Systems, 2021, 45(10):97-105.
• [17]王锋,高欣,贾欣,等.一种基于对数区间隔离森林的电力调度数据异常检测集成算法[J].电网技术, 2021,45(12):4818-4827.WANG Feng, GAO Xin, JIA Xin, et al. An anomaly detection ensemble algorithm based on loginterval isolation[J]. Power System Technology, 2021, 45(12):4818-4827.
• [18]王璨,马金辉,王松,等.基于大数据的配电网运行线损异常诊断模型构建[J].自动化与仪表, 2023, 38(3):96-99.WANG Can, MA Jinhui, WANG Song, et al. Construction of abnormal diagnosis model of distribution network operation line loss based on big data[J]. Automation&Instrumentation, 2023, 38(3):96-99.
• [19]刘震宇.基于支持向量机的电力系统调度数据异常检测方法[J].自动化技术与应用, 2023, 42(6):24-27.LIU Zhenyu. Anomaly detection method of power system dispatching data based on support vector machine[J].Techniques of Automation and Applications, 2023, 42(6):24-27.
• [20]杨茂,黄鑫.基于光照过程特征分析的光伏功率异常数据识别算法[J].电力系统自动化, 2019, 43(6):64-69.YANG Mao, HUANG Xin. Abnormal data identification algorithm for photovoltaic power based on characteristics analysis of illumination process[J]. Automation of Electric Power Systems, 2019, 43(6):64-69.
• [21]吴悦,张忠会,游志刚,等.基于可视化和数据融合的分布式光伏电气设备异常运行状态检测方法[J].制造业自动化, 2021, 43(9):16-19.WU Yue, ZHANG Zhonghui, YOU Zhigang. Abnormal operation state detection method of electrical equipment in photovoltaic power station based on visualization and data fusion[J]. Manufacturing Automation, 2021, 43(9):16-19.
• [22]马速良,武亦文,李建林,等.聚类分析架构下基于遗传算法的电池异常数据检测方法[J].电网技术, 2023,47(2):859-867.MA Suliang, WU Yiwen, LI Jianlin, et al. Anomaly detection for battery data based on genetic algorithm under cluster analysis framework[J]. Power System Technology, 2023, 47(2):859-867.
• [23]周嘉琪,毕利.基于GAN的光伏逆变器数据异常检测技术[J].电力系统保护与控制, 2022, 50(1):133-140.ZHOU Jiaqi, BI Li. Abnormal detection technology of photovoltaic inverter data based on GAN[J]. Power System Protection and Control, 2022, 50(1):133-140.
• [24]曹俊,王珂,曹瑞峰,等.基于区域光伏发电曲线相似度的异常运行分析[J].电力大数据, 2022, 25(7):43-52.CAO Jun, WANG Ke, CAO Ruifeng, et al. Operational anomaly analysis based on regional photovoltaic power generation curve similarity[J]. Power Systems and Big Data, 2022, 25(7):43-52.
• [25]郭辉,杨国清,姚李孝,等.基于综合相似日和功率相关性的分布式光伏预测功率修正[J].电网与清洁能源,2018, 34(9):52-58.GUO Hui, YANG Guoqing, YAO Lixiao, et al. Correction of predictive power of PV plants based on integrated similar days and power correlations[J]. Power System and Clean Energy, 2018, 34(9):52-58.
• [26]罗智骁.基于聚类和相似日理论的光伏预测数据预处理模型研究[D].北京:华北电力大学, 2021:1.LUO Zhixiao. Research on photovoltaic prediction data preprocessing model based on clustering and similar day theory[D]. Beijing:North China Electric Power University, 2021:1.
• [27]邢晨,温蜜,张照贝.基于SA的光伏出力概率密度预测方法[J].计算机仿真, 2023, 40(5):154-161.XING Chen, WEN Mi, ZHANG Zhaobei. Photovoltaic output probability density forecasting method based on Seq2seq-attention(SA)[J]. Computer Simulation, 2023,40(5):154-161.
• [28]薛文琦.基于深度学习的光伏数据异常修复与出力预测方法研究[D].长沙:湖南大学, 2022:1.XUE Wenqi. Research on photovoltaic data anomaly repair and output prediction method based on deep learning[D]. Changsha:Hunan University, 2022:1.
• [29]吴永斌,张建忠,邓富金,等.基于方差变化率判据-四分位的风电场功率异常数据识别[J].电力工程技术,2023, 42(4):141-148.WU Yongbin, ZHANG Jianzhong, DENG Fujin, et al.Identification of wind farm power anomaly data based on variance change rate criterion-quartile[J]. Electric Power Engineering Technology, 2023, 42(4):141-148.
• [30]马良玉,耿妍竹,袁乃正,等.基于四分位与CFSFDP的风电机组异常数据清洗方法[J].电力科学与工程,2023, 39(6):9-16.MA Liangyu, GENG Yanzhu, YUAN Naizheng, et al.Abnormal data cleaning method of wind turbine based on quartile and CFSFDP[J]. Electric Power Science and Engineering, 2023, 39(6):9-16.
• [31]杨茂,孟玲建,李大勇,等.基于类3σ准则的光伏功率异常数据识别[J].可再生能源, 2018, 36(10):1443-1448.YANG Mao, MENG Lingjian, LI Dayong, et al.Identification of abnormal data of photovoltaic power based on class 3σ[J]. Renewable Energy Resources, 2018,36(10):1443-1448.
• [32]胡傲宇.光伏发电系统异常数据清洗及其应用研究[D].北京:华北电力大学, 2020:1.HU Aoyu. Research oil abnormal data cleaning of photovoltaic power generation and its application[D].Beijing:North China Electric Power University, 2020:1.
• [33]李智勇,周剑,俞蕙,等.基于特征聚类分析的大规模发电数据异常辨识方法[J].南方能源建设, 2018, 5(1):35-39.LI Zhiyong, ZHOU Jian, YU Hui, et al. An abnormal data identification method of large-scale generation data based on cluster analysis[J]. Southern Energy Construction,2018, 5(1):35-39.