基于STL-Bayesian时空模型的分布式光伏系统异常检测

doi:10.11930/j.issn.1004-9649.202305120

中国电力 ›› 2024, Vol. 57 ›› Issue (5): 222-231.DOI: 10.11930/j.issn.1004-9649.202305120

基于STL-Bayesian时空模型的分布式光伏系统异常检测

刘韵艺¹(), 汤渊¹(), 苏盛²(), 吴裕宙¹(), 王晓倩²()

1. 广东电网有限责任公司东莞供电局，广东东莞　523008
2. 长沙理工大学电气与信息工程学院，湖南长沙　410004

收稿日期:2023-05-27 出版日期:2024-05-28 发布日期:2024-05-16
作者简介:刘韵艺（1990—），女，硕士，工程师，从事电力营销管理研究，E-mail： 516101194@qq.com
汤渊（1990—），男，硕士，高级工程师，从事电力营销管理研究，E-mail：454237103@qq.com
苏盛（1975—），男，通信作者，博士生导师，教授，从事配用电大数据分析与电力气象灾害分析研究，E-mail：eessheng@163.com
吴裕宙（1983—），男，硕士，高级工程师，从事电力营销数据分析与高级应用研究，E-mail：aeou11983@163.com
王晓倩（1998—），女，硕士研究生，从事分布式光伏系统异常状态检测技术研究，E-mail：wxqian21@126.com
基金资助:
国家自然科学基金资助项目（电力调度系统定向攻击的虚构陷阱抗毁性主动安全防护方法研究，51777015）；中国南方电网有限公司科技项目（031900KK52220039）。

Anomaly Detection for Distributed Photovoltaic Systems Based on STL-Bayesian Spatio-Temporal Model

Yunyi LIU¹(), Yuan TANG¹(), Sheng SU²(), Yuzhou WU¹(), Xiaoqian WANG²()

1. Guangdong Power Grid Company Limited Dongguan Power Supply Bureau, Dongguan 523008, China
2. School of Electrical & Information Engineering, Changsha University of Science and Technology, Changsha 410004, China

Received:2023-05-27 Online:2024-05-28 Published:2024-05-16
Supported by:
This work is supported by National Natural Science Foundation of China (Research on Destructive Active Security Protection Method of Fictitious Traps Against Targeted Attacks in Power Dispatch System, No.51777015), Science and Technology Project of CSG (No.031900KK52220039).

摘要/Abstract

摘要：

分布式光伏发电系统一般不配备多种类的传感器和监测设备，反映设备运行状态且可用于异常检测的数据有限。提出了基于STL-Bayesian时空模型的光伏异常状态检测方法，利用气象在时空上的传递性，挖掘光伏发电出力的关联性进而完成异常检测。首先，用季节性分解（seasonal and trend decomposition using loess，STL）将光伏发电有功功率时序数据分解为3个分量；然后，研究不同长度数据输入对分解结果的影响和区域内分量的时空分布特性；接着，通过构建贝叶斯模型分别对趋势分量和剩余分量做短期和超短期空间插值，得到区域内光伏出力；最后，计算真实值与回归值的推土机距离（earth move's distance，EMD）用于检测异常状态。算例分析表明，所提模型在分布式光伏场景检测可逆异常和不可逆异常状态均有较高准确率。

关键词: 分布式光伏, 时序分解, 空间插值, 异常状态检测, 时空分布特性

Abstract:

Distributed photovoltaic (PV) power generation systems generally do not come equipped with a variety of sensors and monitoring devices, limiting the data available for reflecting equipment operation and conducting anomaly detection. This article proposes a PV anomaly detection method based on the STL-Bayesian spatio-temporal model, which utilizes the spatio-temporal transferability of meteorological data to uncover the correlation of PV power output and perform anomaly detection. Firstly, the seasonal and trend decomposition using Loess (STL) is employed to decompose the PV active power time series data into three components. Then, the influence of different lengths of input data on the decomposition results and the spatio-temporal distribution characteristics of the components within the region are investigated. Subsequently, Bayesian models are constructed to perform short-term and ultra-short-term spatial interpolation on the trend component and the residual component, respectively, so as to obtain the PV output within the region. Finally, the earth move's distance (EMD) between the actual values and regression values is calculated to detect abnormal states. The analysis of the algorithm shows that the model has a high accuracy in the detection of both reversible and irreversible anomalies under distributed PV scenarios.

Key words: distributed photovoltaics, time series data decomposition, spatial interpolation, abnormal states detection, spatio-temporal distribution characteristics

刘韵艺, 汤渊, 苏盛, 吴裕宙, 王晓倩. 基于STL-Bayesian时空模型的分布式光伏系统异常检测[J]. 中国电力, 2024, 57(5): 222-231.

Yunyi LIU, Yuan TANG, Sheng SU, Yuzhou WU, Xiaoqian WANG. Anomaly Detection for Distributed Photovoltaic Systems Based on STL-Bayesian Spatio-Temporal Model[J]. Electric Power, 2024, 57(5): 222-231.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.electricpower.com.cn/CN/10.11930/j.issn.1004-9649.202305120

https://www.electricpower.com.cn/CN/Y2024/V57/I5/222

图/表 12

表 1 光伏系统异常状态类型

Table 1 Types of PV abnormal states

异常状态类型	来源	具体异常情况
可逆	安装错误	接线错误；支架不稳定
可逆	环境及气候	阴影或杂物遮挡；冰雹、高温等天气
不可逆	组件永久性故障	面板开裂、线路老化、逆变器过压过温等故障；布线短路

图 1 不同状态下光伏发电功率曲线

Fig.1 PV power curve under different states

图 2 STL分解光伏出力结果

Fig.2 STL decomposition results of PV output

表 2 不同周期STL分解的季节和趋势强度

Table 2 Seasonal and trend intensity of STL decomposition across cycles

输入数据的长度/天	季节强度	趋势强度
7	0.999	0.243
30	0.952	0.565
90	0.927	0.719
365	0.849	0.857

图 3 贝叶斯时空模型流程

Fig.3 Bayesian spatio-temporal model flow chart

图 4 区域内贝叶斯插值分布

Fig.4 Bayesian interpolation distribution in the region

图 5 异常检测流程

Fig.5 Anomaly detection flow chart

图 6 区域内时空插值曲线

Fig.6 Regional spatio-temporal interpolation curves

表 3 不同算法回归的误差

Table 3 Error for different algorithm regressions

算法	E_RMS	E_MA
克里金插值	0.19956	0.12382
STL-Bayesian	0.09293	0.05761

表 4 异常数据与回归数据的距离

Table 4 Distance between abnormal data and regression data

异常数据	EMD距离	欧氏距离
全天下降（k=0.1）	3.87	1.01
全天下降（k=0.2）	4.96	1.12
全天下降（k=0.3）	6.32	1.28
可逆（k=0.1）	3.02	0.95
可逆（k=0.2）	3.19	0.99
可逆（k=0.3）	3.43	1.02

图 7 异常检测结果混淆矩阵

Fig.7 Confusion matrix for anomaly detection results

表 5 异常检测结果

Table 5 Anomaly detection results

算法	异常类型	精确率/%	召回率/%
STL-Bayesian	不可逆故障	96.6	93.5
STL-Bayesian	可逆异常状态	90.0	87.1
LOF	不可逆故障	73.3	78.6
LOF	可逆异常状态	53.3	47.1

参考文献 25

1	曹仟妮, 贾孟硕, 沈沉. 针对复杂安装条件的光伏电站光伏板异常状态检测技术[J]. 中国电机工程学报, 2022, 42 (5): 1917- 1925.
	CAO Qianni, JIA Mengshuo, SHEN Chen. A fault detection scheme for PV modules in large scale PV stations with complex installation conditions[J]. Proceedings of the CSEE, 2022, 42 (5): 1917- 1925.
2	李光辉, 段晨东, 武珊. 基于半监督机器学习法的光伏阵列故障诊断[J]. 电网技术, 2020, 44 (5): 1908- 1913.
	LI Guanghui, DUAN Chendong, WU Shan. Fault diagnosis of PV array based on semi-supervised machine learning[J]. Power System Technology, 2020, 44 (5): 1908- 1913.
3	焦田利, 章坚民, 李熊, 等. 基于空间相关性的大规模分布式用户光伏空间分群方法[J]. 电力系统自动化, 2019, 43 (21): 97- 102, 162.
	JIAO Tianli, ZHANG Jianmin, LI Xiong, et al. Spatial clustering method for large-scale distributed user photovoltaics based on spatial correlation[J]. Automation of Electric Power Systems, 2019, 43 (21): 97- 102, 162.
4	张青山, 王丽婕, 郝颖, 等. 基于卫星云图和晴空模型的分布式光伏电站太阳辐照度超短期预测[J]. 高电压技术, 2022, 48 (8): 3271- 3281.
	ZHANG Qingshan, WANG Lijie, HAO Ying, et al. Ultra-short-term solar irradiance prediction of distributed photovoltaic power stations based on satellite cloud images and clear sky model[J]. High Voltage Engineering, 2022, 48 (8): 3271- 3281.
5	SHI Y C, HE W G, ZHAO J, et al. Expected output calculation based on inverse distance weighting and its application in anomaly detection of distributed photovoltaic power stations[J]. Journal of Cleaner Production, 2020, 253, 119965. DOI
6	张童彦, 廖清芬, 唐飞, 等. 基于气象资源插值与迁移学习的广域分布式光伏功率预测方法[J]. 中国电机工程学报, 2023, 43 (20): 7929- 7940.
	ZHANG Tongyan, LIAO Qingfen, TANG Fei, et al. Wide-area distributed photovoltaic power forecast method based on meteorological resource interpolation and transfer learning[J]. Proceedings of the CSEE, 2023, 43 (20): 7929- 7940.
7	王本涛, 白杨, 邢红涛, 等. 基于STL与MMoE多任务学习的区域多光伏电站超短期功率联合预测方法[J]. 电力系统及其自动化学报, 2022, 34 (9): 17- 23, 31.
	WANG Bentao, BAI Yang, XING Hongtao, et al. Combined ultra-short-term power prediction method for regional multi-photovoltaic power stations based on STL and MMoE multi-task learning[J]. Proceedings of the CSU-EPSA, 2022, 34 (9): 17- 23, 31.
8	孙一浩, 肖先勇, 张文海, 等. 基于伪异常点辨识的关口电能表计量数据异常研究[J]. 电网技术, 2021, 45 (11): 4568- 4576.
	SUN Yihao, XIAO Xianyong, ZHANG Wenhai, et al. Abnormal measurement data of gateway energy meter based on pseudo outlier identification[J]. Power System Technology, 2021, 45 (11): 4568- 4576.
9	熊寿遥, 邓仁毅, 黄创霞, 等. 气象因素对分布式光伏功率的影响分析[J]. 系统科学与数学, 2022, 42 (8): 2092- 2106.
	XIONG Shouyao, DENG Renyi, HUANG Chuangxia, et al. Analysis on the impact of meteorological factors on distributed photovoltaic power[J]. Journal of Systems Science and Mathematical Sciences, 2022, 42 (8): 2092- 2106.
10	杨文海, 程华新, 高亚静, 等. 计及雾霾影响的短期光伏出力预测和负荷预测[J]. 中国电力, 2016, 49 (S): 148- 153.
	YANG Wenhai, CHENG Xinhua, GAO Yajing, et al. Short-term photovoltaic output forecast considering the influence of fog and haze weather[J]. Electric Power, 2016, 49 (S): 148- 153.
11	陆双, 彭曙蓉, 杨云皓, 等. 基于平均影响值-启发式前向搜索的异常光伏用户识别方法[J]. 电力自动化设备, 2022, 42 (2): 106- 111.
	LU Shuang, PENG Shurong, YANG Yunhao, et al. Identification method of abnormal photovoltaic users based on mean impact value and heuristic forward searching[J]. Electric Power Automation Equipment, 2022, 42 (2): 106- 111.
12	徐一伦, 张彬桥, 黄婧, 等. 考虑天气类型和相似日的IWPA-LSSVM光伏发电功率预测[J]. 中国电力, 2023, 56 (2): 143- 149.
	XU Yilun, ZHANG Binqiao, HUANG Jing, et al. Forecast of photovoltaic power based on IWPA-LSSVM considering weather types and similar days[J]. Electric Power, 2023, 56 (2): 143- 149.
13	邵尹池, 袁绍军, 孙荣富, 等. 基于空间相关性的分布式光伏实用化功率预测及误差分析[J]. 中国电力, 2021, 54 (7): 185- 191, 207.
	SHAO Yinchi, YUAN Shaojun, SUN Rongfu, et al. Practical method and error analysis for distributed photovoltaic power prediction based on spatial correlation[J]. Electric Power, 2021, 54 (7): 185- 191, 207.
14	裴哲义, 梁志峰, 陆为华, 等. 广域分布式光伏发电监测与出力估计研究[J]. 中国电力, 2020, 53 (6): 87- 96.
	PEI Zheyi, LIANG Zhifeng, LU Weihua, et al. Wide-area distributed photovoltaic power generation monitoring and output estimation[J]. Electric Power, 2020, 53 (6): 87- 96.
15	GE L J, LIU H X, YAN J, et al. A virtual data collection model of distributed PVs considering spatio-temporal coupling and affine optimization reference[J]. IEEE Transactions on Power Systems, 2023, 38 (4): 3939- 3951.
16	赵振兵, 强一凡, 李信, 等. 基于改进循环神经网络的配电网超短期功率预测方法[J]. 电力科学与技术学报, 2022, 37 (5): 144- 154.
	ZHAO Zhenbing, QIANG Yifan, LI Xin, et al. Ultra-short-term power prediction method of distribution network based on improved recurrent neural network[J]. Journal of Electric Power Science and Technology, 2022, 37 (5): 144- 154.
17	刘明群, 何鑫, 覃日升, 等. 基于改进K-means聚类k值选择算法的配网电压数据异常检测[J]. 电力科学与技术学报, 2022, 37 (6): 91- 99.
	LIU Mingqun, HE Xin, QIN Risheng, et al. Anomaly detection of distribution network voltage data based on improved K-means clustering k-value selection algorithm[J]. Journal of Electric Power Science and Technology, 2022, 37 (6): 91- 99.
18	刘昳娟, 陈云龙, 刘继彦, 等. 基于集成学习的分布式光伏发电功率日前预测[J]. 中国电力, 2022, 55 (9): 38- 45.
	LIU (Die\| Yi)(Juan), CHEN Yunlong, LIU Jiyan, et al. Ensemble learning-based day-ahead power forecasting of distributed photovoltaic generation[J]. Electric Power, 2022, 55 (9): 38- 45.
19	OVASKA S J, VANLANDINGHAM H F, KAMIYA A. Fusion of soft computing and hard computing in industrial applications: an overview[J]. IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews), 2002, 32 (2): 72- 79. DOI
20	AGOUA X G, GIRARD R, KARINIOTAKIS G. Short-term spatio-temporal forecasting of photovoltaic power production[J]. IEEE Transactions on Sustainable Energy, 2018, 9 (2): 538- 546. DOI
21	杨龙飞, 高山, 蔡新雷, 等. 基于拉丁超立方采样的小范围分布式光伏出力时空概率分布生成方法[J]. 南方电网技术, 2023, 17 (9): 38- 48.
	YANG Longfei, GAO Shan, CAI Xinlei, et al. Generating method for spatiotemporal probability distribution of distributed PV output in small range based on Latin hypercube sampling[J]. Southern Power System Technology, 2023, 17 (9): 38- 48.
22	高博, 茆超, 张冲标, 等. 基于时空图网络的分布式光伏发电出力预测[J]. 电力系统及其自动化学报, 2023, 35 (3): 125- 133.
	GAO Bo, MAO Chao, ZHANG Chongbiao, et al. Output prediction of distributed photovoltaic power generation based on spatial-temporal graph neural network[J]. Proceedings of the CSU-EPSA, 2023, 35 (3): 125- 133.
23	邓序之, 刘淇, 叶傲霜, 等. 基于气象耦合特征分析及改进XGBoost算法的用户分布式光伏短期出力预测模型[J]. 南方电网技术, 2023, 17 (12): 80- 89.
	DENG Xuzhi, LIU Qi, YE Aoshuang, et al. Short-term output forecasting model of user side distributed photovoltaic based on meteorological coupling characteristics analysis and improved XGBoost algorithm[J]. Southern Power System Technology, 2023, 17 (12): 80- 89.
24	王守相, 郭陆阳, 赵倩宇, 等. 基于参考序列与全卷积网络的风速数据缺失与异常修复方法[J]. 电力系统自动化, 2023, 47 (9): 129- 136.
	WANG Shouxiang, GUO Luyang, ZHAO Qianyu, et al. Repair method for missed and abnormal wind speed data based on reference sequence and full convolution network[J]. Automation of Electric Power Systems, 2023, 47 (9): 129- 136.
25	赵一鉴, 林利, 王茜蒨, 等. 基于大地距离计算相似度的海上目标轨迹预测[J]. 计算机应用, 2023, 43 (11): 3594- 3598.
	ZHAO Yijian, LIN Li, WANG Qianqian, et al. Trajectory prediction of sea targets based on geodetic distance similarity calculation[J]. Journal of Computer Applications, 2023, 43 (11): 3594- 3598.

[1]	孙通, 张沈习, 曹毅, 曹佳晨, 程浩忠. 计及5G基站可调特性的配电网分布式光伏准入容量鲁棒优化[J]. 中国电力, 2025, 58(2): 140-146.
[2]	周才期, 刘静利, 孙鹏凯, 张玉敏. 分布式光伏波动事件多级区间滚动预警方法[J]. 中国电力, 2024, 57(8): 96-107.
[3]	王海军, 居蓉蓉, 董颖华. 基于时空关联特征与B-LSTM模型的分布式光伏功率区间预测[J]. 中国电力, 2024, 57(7): 74-80.
[4]	张宇, 魏新迟, 冯凯辉, 时珊珊, 孟子涵, 梁媛. 计及分布式光伏承载力的配电网侧独立储能充放电策略[J]. 中国电力, 2024, 57(4): 111-117.
[5]	陈璨, 苏紫诺, 马原, 刘佳林, 王玉庆, 王飞. 基于分层关联性建模的分布式光伏功率超短期概率预测[J]. 中国电力, 2024, 57(12): 50-59.
[6]	杨鹏伟, 赵丽萍, 陈军法, 甄钊, 王飞, 李利明. 基于相似时段匹配与Transformer网络建模的分布式光伏超短期功率预测方法[J]. 中国电力, 2024, 57(12): 60-70.
[7]	陈奇芳, 李若凡, 夏明超, 吴迪, 邵尹池. 计及多维性能评估的新型配电网光伏选址定容方法[J]. 中国电力, 2024, 57(10): 172-178, 207.
[8]	白晶, 金广厚, 孙鹤林, 胡楚樊, 赵晋泉, 陈天华. 高渗透光伏配电网第三方主体调压辅助服务补偿与获取[J]. 中国电力, 2023, 56(4): 95-103,111.
[9]	李娟, 汪家铭, 许苏迪, 姜云龙, 杨雄. 基于量子萤火虫算法的配电网故障恢复策略[J]. 中国电力, 2023, 56(12): 191-198.
[10]	段瑶, 高崇, 程苒, 吴亚雄, 黄烨, 刘志文. 考虑5G基站可调度潜力的配电网分布式光伏最大准入容量评估[J]. 中国电力, 2023, 56(12): 80-85, 99.
[11]	王晶, 袁懿, 邵尹池, 张晋奇, 丁然, 巩彦江. 考虑资源储备度的主动配电网多目标集群划分与电压控制方法[J]. 中国电力, 2023, 56(12): 69-79.
[12]	李世辉, 王琪, 贾晓卜, 吴文传, 梁华洋, 朱琳. 考虑热泵负荷和分布式光伏的配微网协调调度[J]. 中国电力, 2022, 55(9): 29-37.
[13]	陈璨, 白明辉, 张婉明, 席海阔, 尹兆磊, 邵尹池, 王枭枭, 马原. 分布式光伏边界渗透率快速定位及消纳方案择优[J]. 中国电力, 2022, 55(8): 40-50.
[14]	李丰君, 王磊, 赵健, 张建宾, 张世尧, 田杨阳. 基于天气融合和LSTM网络的分布式光伏短期功率预测方法[J]. 中国电力, 2022, 55(11): 149-154.
[15]	左高, 张超, 陈毓春, 于海波, 卢毓东. 基于区间直觉模糊群决策的分布式光伏运维综合评估[J]. 中国电力, 2021, 54(8): 154-163.

基于STL-Bayesian时空模型的分布式光伏系统异常检测

Anomaly Detection for Distributed Photovoltaic Systems Based on STL-Bayesian Spatio-Temporal Model

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 12

参考文献 25

相关文章 15

编辑推荐

Metrics

模态框（Modal）标题

基于STL-Bayesian时空模型的分布式光伏系统异常检测

Anomaly Detection for Distributed Photovoltaic Systems Based on STL-Bayesian Spatio-Temporal Model

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 12

参考文献 25

相关文章 15

编辑推荐

Metrics