Short-Term Probabilistic Power Forecasting of Photovoltaic Clusters Based on Spatial Encoded Image Generation

doi:10.16183/j.cnki.jsjtu.2024.427

Abstract

Abstract:

A short-term probabilistic forecasting method for distributed photovoltaic (PV) clusters based on spatial encoding maps is proposed in this paper. First, the distributed PV clusters are encoded into a two-dimensional matrix according to their spatial layout, thereby forming PV output feature maps and weather feature maps. Then, using a conditional generative adversarial network as the forecasting model, the short-term power forecasting problem of the cluster is transformed into a problem of generating future multi-period PV output feature maps under given conditions. Multiple forecast scenarios can be generated by adjusting the input random vectors, thereby constructing probabilistic forecasting results. Finally, the effectiveness of the forecast results is comprehensively evaluated by three metrics: point-to-point error, interval coverage rate, and data distribution similarity. Case study results based on data from the Alice Springs distributed PV project at the Desert Knowledge Australia Solar Centre indicate that the proposed method can fully exploit the spatiotemporal correlations within the cluster, simultaneously providing forecasting results for all PV units in the cluster. Compared with traditional point-wise forecasting methods, the proposed approach achieves higher prediction accuracy and computational efficiency.

Key words: photovoltaic (PV) cluster, short-term probabilistic forecasting, spatial encoding, conditional generative adversarial network (cGAN)

CLC Number:

TM769

CHEN Junjie, LI Yiyan, ZHOU Zhenghao, YAN Zheng. Short-Term Probabilistic Power Forecasting of Photovoltaic Clusters Based on Spatial Encoded Image Generation[J]. Journal of Shanghai Jiao Tong University, 2026, 60(7): 1160-1172.

Figures/Tables 16

Fig.1

Fig.2

Fig.3

Tab.1

Tab.2

Fig.4

Tab.3

Fig.5

Fig.6

Fig.7

Tab.4

Fig.8

Fig.9

Tab.5

Fig.10

Tab.6

References 36

[11]	ZHAO Long, YANG Bo, LU Zhipeng, et al. Ultra-short-term cluster distributed photovoltaic power prediction method based on spectral clustering and AM-LSTM[J]. Distribution & Utilization, 2023, 40(7): 10-17.
[12]	王于波, 郝玲, 徐飞, 等. 分布式光伏集群发电功率波动模式识别与超短期概率预测[J]. 上海交通大学学报, 2024, 58(9): 1334-1343. doi: 10.16183/j.cnki.jsjtu.2023.048
	WANG Yubo, HAO Ling, XU Fei, et al. Pattern recognition and ultra-short-term probabilistic forecasting of power fluctuating in aggregated distributed photovoltaics clusters[J]. Journal of Shanghai Jiao Tong University, 2024, 58(9): 1334-1343.
[13]	杨茂, 贾梦琦, 张薇, 等. 基于功率重构和时序特性约束的长预见期光伏集群功率预测[J]. 电力系统自动化, 2024, 48(15): 102-111.
	YANG Mao, JIA Mengqi, ZHANG Wei, et al. Power forecasting for photovoltaic cluster in long forecasting period based on power reconstruction and temporal characteristic constraints[J]. Automation of Electric Power Systems, 2024, 48(15): 102-111.
[14]	LOBO M G, SANCHEZ I. Regional wind power forecasting based on smoothing techniques, with application to the Spanish peninsular system[J]. IEEE Transactions on Power Systems, 2012, 27(4): 1990-1997. doi: 10.1109/TPWRS.2012.2189418 URL
[15]	白永祥, 房大中, 侯佑华, 等. 内蒙古电网区域风电功率预测系统[J]. 电网技术, 2010, 34(10): 157-162.
	BAI Yongxiang, FANG Dazhong, HOU Youhua, et al. Regional wind power forecasting system for Inner Mongolia power grid[J]. Power System Technology, 2010, 34(10): 157-162.
[16]	刘运超, 杨宁, 崔承刚, 等. 基于时空信息的区域内光伏电站功率预测[J]. 科学技术与工程, 2023, 23(34): 14596-14602.
	LIU Yunchao, YANG Ning, CUI Chenggang, et al. Power prediction of regional photovoltaic plants based on spatio-temporal information[J]. Science Technology & Engineering, 2023, 23(34): 14596-14602.
[17]	欧阳永健, 缪希仁, 林蔚青, 等. 基于分层有向图与动态时空相关性的小区域光伏超短期预测方法[J]. 电网技术, 2024, 48(6): 2458-2468.
	OUYANG Yongjian, MIAO Xiren, LIN Weiqing, et al. Small area PV ultra-short-term prediction method using stratified digraph and dynamic spatial-temporal correlation[J]. Power System Technology, 2024, 48(6): 2458-2468.
[1]	IEA PVPS. Snapshot of global PV markets 2024[R/OL]. (2024-04-01)[2024-12-15]. https://iea-pvps.org/snapshot-reports/snapshot-2024/.
[2]	WANG F, XUAN Z M, ZHEN Z, et al. A day-ahead PV power forecasting method based on LSTM-RNN model and time correlation modification under partial daily pattern prediction framework[J]. Energy Conversion & Management, 2020, 212: 112766.
[3]	李安寿, 陈琦, 王子才, 等. 光伏发电系统功率预测方法综述[J]. 电气传动, 2016, 46(6): 93-96.
	LI Anshou, CHEN Qi, WANG Zicai, et al. Review of power forecast methods for photovoltaic generating system[J]. Electric Drive, 2016, 46(6): 93-96.
[4]	王洪坤, 葛磊蛟, 李宏伟, 等. 分布式光伏发电的特性分析与预测方法综述[J]. 电力建设, 2017, 38(7): 1-9. doi: 10.3969/j.issn.1000-7229.2017.07.001
	WANG Hongkun, GE Leijiao, LI Hongwei, et al. A review on characteristic analysis and prediction method of distributed PV[J]. Electric Power Construction, 2017, 38(7): 1-9. doi: 10.3969/j.issn.1000-7229.2017.07.001
[5]	栗峰, 丁杰, 周才期, 等. 新型电力系统下分布式光伏规模化并网运行关键技术探讨[J]. 电网技术, 2024, 48(1): 184-196.
	LI Feng, DING Jie, ZHOU Caiqi, et al. Key technologies of large-scale grid-connected operation of distributed photovoltaic under new-type power system[J]. Power System Technology, 2024, 48(1): 184-196.
[6]	阮呈隆, 李康平, 李正辉, 等. 分布式光伏集群功率短期预测的空间互补特性初探[J]. 电力系统自动化, 2024, 48(3): 42-50.
	RUAN Chenglong, LI Kangping, LI Zhenghui, et al. Preliminary study on spatial complementarity characteristics of short-term power prediction for distributed photovoltaic clusters[J]. Automation of Electric Power Systems, 2024, 48(3): 42-50.
[7]	LIU J C, LI K J, XUE W P. Photovoltaic solar power prediction using iPSO-based data clustering and AdaLSTM network[J]. Energies, 2024, 17(7): 1624. doi: 10.3390/en17071624 URL
[18]	赖昌伟, 黎静华, 陈博, 等. 光伏发电出力预测技术研究综述[J]. 电工技术学报, 2019, 34(6): 1201-1217.
	LAI Changwei, LI Jinghua, CHEN Bo, et al. Review of photovoltaic power output prediction technology[J]. Transactions of China Electrotechnical Society, 2019, 34(6): 1201-1217.
[19]	黎静华, 骆怡辰, 杨舒惠, 等. 可再生能源电力不确定性预测方法综述[J]. 高电压技术, 2021, 47(4): 1144-1155.
	LI Jinghua, LUO Yichen, YANG Shuhui, et al. Review of uncertainty forecasting methods for renewable energy power[J]. High Voltage Engineering, 2021, 47(4): 1144-1155.
[20]	王思懿, 盛万兴, 刘科研, 等. 基于天气分类的改进CNN-LSTM区域光伏功率单值和概率预测方法[J]. 新型电力系统, 2023, 1(3): 283-292.
	WANG Siyi, SHENG Wanxing, LIU Keyan, et al. Improved CNN-LSTM method for single value and interval prediction of regional photovaltaic power based on weather classification[J]. New Type Power Systems, 2023, 1(3): 283-292.
[21]	康重庆, 夏清, 刘梅. 电力系统负荷预测[M]. 2版. 北京: 中国电力出版社, 2017.
	KANG Chongqing, XIA Qing, LIU Mei. Power system load forecasting[M]. 2nd ed. Beijing: China Electric Power Press, 2017.
[22]	赵耀, 高少炜, 李东东, 等. 基于天气相似聚类与QRNN的短期光伏功率区间概率预测[J]. 电力系统自动化, 2023, 47(23): 152-161.
	ZHAO Yao, GAO Shaowei, LI Dongdong, et al. Short-term interval probability prediction of photovoltaic power based on weather similarity clustering and quantile regression neural network[J]. Automation of Electric Power Systems, 2023, 47(23): 152-161.
[23]	GOODFELLOW I J, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial nets[C]// Proceedings of the 27th International Conference on Neural Information Processing Systems. Montreal, Canada: NIPS, 2014: 1-9.
[8]	IEA. Trends in PV applications 2022[EB/OL]. (2022-02-01)[2024-12-15]. https://iea-pvps.org/trends_reports/trends-2022/.
[9]	张小萌, 白恺, 柳玉, 等. 大规模分布式光伏短期集群功率预测综述[J]. 华北电力技术, 2017(6): 1-7.
	ZHANG Xiaomeng, BAI Kai, LIU Yu, et al. Overview for large-scale distributed photovoltaic power prediction with clustered method[J]. North China Electric Power, 2017(6): 1-7.
[10]	麻吕斌, 潘国兵, 蒋群, 等. 基于EOF-DBSCAN-GRU的分布式光伏集群出力预测方法研究[J]. 太阳能学报, 2024, 45(1): 39-46. doi: 10.19912/j.0254-0096.tynxb.2022-1461
	MA Lübin, PAN Guobing, JIANG Qun, et al. Research on distributed pv cluster power output forecasting method based on EOF-DBSCAN-GRU[J]. Acta Energiae Solaris Sinica, 2024, 45(1): 39-46. doi: 10.19912/j.0254-0096.tynxb.2022-1461
[11]	赵龙, 杨波, 卢志鹏, 等. 基于谱聚类和AM-LSTM的分布式光伏集群超短期预测方法[J]. 供用电, 2023, 40(7): 10-17.
[24]	MIRZA M, OSINDERO S. Conditional generative adversarial nets[PP/OL]. V1. arXiv (2014-11-06)[2024-12-15]. https://doi.org/10.48550/arXiv.1411.1784.
[25]	林珊, 王红, 齐林海, 等. 基于条件生成对抗网络的短期负荷预测[J]. 电力系统自动化, 2021, 45(11): 52-60.
	LIN Shan, WANG Hong, QI Linhai, et al. Short-term load forecasting based on conditional generative adversarial network[J]. Automation of Electric Power Systems, 2021, 45(11): 52-60.
[26]	孙浩, 万灿, 曹照静, 等. 基于条件生成对抗网络曲线生成的短期负荷概率预测[J]. 电力系统自动化, 2023, 47(23): 189-199.
	SUN Hao, WAN Can, CAO Zhaojing, et al. Short-term load probabilistic forecasting based on conditional generative adversarial network curve generation[J]. Automation of Electric Power Systems, 2023, 47(23): 189-199.
[27]	Yulara Solar System. DKA solar centre[EB/OL]. (2016-02-24) [2024-12-15]. http://dkasolarcentre.com.au/download/notes-on-the-data/.
[28]	董骁翀, 孙英云, 蒲天骄. 基于条件生成对抗网络的可再生能源日前场景生成方法[J]. 中国电机工程学报, 2020, 40(17): 5527-5535.
	DONG Xiaochong, SUN Yingyun, PU Tianjiao. Day-ahead scenario generation of renewable energy based on conditional GAN[J]. Proceedings of the CSEE, 2020, 40(17): 5527-5535.
[29]	HENDRYCKS D, GIMPEL K. Gaussian error linear units (GELUs)[PP/OL]. V5. arXiv (2023-06-06)[2024-12-15]. https://doi.org/10.48550/arXiv.1606.08415.
[30]	VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]// Proceedings of the 31st International Conference on Neural Information Processing Systems. Long Beach, USA: ACM, 2017: 6000-6010.
[31]	李芬, 杨勇, 胡超, 等. 基于天气类型识别的跟踪式光伏发电系统能量评估及优化[J]. 太阳能学报, 2020, 41(11): 112-119.
	LI Fen, YANG Yong, HU Chao, et al. Energy evaluation analysis and optimization of tracking photovoltaic generation system based on weather type recognition[J]. Acta Energiae Solaris Sinica, 2020, 41(11): 112-119.
[32]	ALRASHIDI M, RAHMAN S. Short-term photovoltaic power production forecasting based on novel hybrid data-driven models[J]. Journal of Big Data, 2023, 10: 26. doi: 10.1186/s40537-023-00706-7
[33]	IHEANETU K J. Solar photovoltaic power forecasting: A review[J]. Sustainability, 2022, 14(24): 17005. doi: 10.3390/su142417005 URL
[34]	韩叶林, 张展耀, 俞伊丽, 等. 基于LSTM及分位数回归理论的配电变压器重过载概率预测[J]. 电力大数据, 2023, 26(6): 1-8.
	HAN Yelin, ZHANG Zhanyao, YU Yili, et al. Probabilistic prediction of heavy or overload distribution transformer based on LSTM and quantile regression theory[J]. Power Systems & Big Data, 2023, 26(6): 1-8.
[35]	易淑智, 刘云凯, 杨帆, 等. 基于改进门控循环单元分位数回归的短期负荷概率预测[J]. 智慧电力, 2021, 49(8): 84-90.
	YI Shuzhi, LIU Yunkai, YANG Fan, et al. Short-term load probability forecasting based on improved quantile regression of gated recurrent unit[J]. Smart Power, 2021, 49(8): 84-90.
[36]	王晓东, 鞠邦国, 刘颖明, 等. 基于QR-NFGLSTM与核密度估计的风电功率概率预测[J]. 太阳能学报, 2022, 43(2): 479-485. doi: 10.19912/j.0254-0096.tynxb.2020-0478
	WANG Xiaodong, JU Bangguo, LIU Yingming, et al. Probability prediction of wind power based on QR-NFGLSTM and kernel density estimation[J]. Acta Energiae Solaris Sinica, 2022, 43(2): 479-485. doi: 10.19912/j.0254-0096.tynxb.2020-0478

序号	变量	$\stackrel{-}{\mathit{r}}$
1	风速/(m·s^-1)	0.2429
2	温度/℃	0.4437
3	相对湿度/%	-0.4317
4	全球水平辐射/(W·m^-2)	0.9338
5	漫射水平辐射/(W·m^-2)	0.5386
6	风向/(°)	-0.0318
7	每日降雨量/mm	-0.0331
8	倾斜面全球辐射/(W·m^-2)	0.9366
9	倾斜面漫射辐射/(W·m^-2)	0.2542

层数	名称	参数	取值
Layer 1	3D卷积层	卷积核大小	3×3×3
	3D卷积层	输出通道数	2
	3D卷积层	卷积步长	3×1×1
	3D卷积层	边缘补零数	1×1×1
	激活函数	激活函数种类	GELU
	批标准化	数量	2
Layer 2	3D卷积层	卷积核大小	3×3×3
	3D卷积层	输出通道数	4
	3D卷积层	卷积步长	3×1×1
	3D卷积层	边缘补零数	1×1×1
	激活函数	激活函数种类	GELU
	批标准化	数量	4
Layer 3	3D卷积层	卷积核大小	3×3×3
	3D卷积层	输出通道数	1
	3D卷积层	卷积步长	3×1×1
	3D卷积层	边缘补零数	1×1×1
	激活函数	激活函数种类	GELU
	批标准化	数量	1
Layer 4	展平	输出层输入	1100
	全连接层	输出特征矩阵大小	2400
	激活函数	激活函数种类	tanh
	重塑	生成器输出	(1,(24,10,10))

层数	名称	参数	取值
Layer 1	展平	判别器输入	21600
Layer 2	全连接层	层输出	5400
	激活函数	激活函数种类	GELU
Layer 3	全连接层	层输出	1350
	激活函数	激活函数种类	GELU
Layer 4	全连接层	层输出	1
	激活函数	激活函数种类	Sigmoid

天气类型	100%置信区间覆盖率/%	90%置信区间覆盖率/%	100%置信区间宽度/ kW	90%置信区间宽度/ kW
晴天	98.3	96.8	0.3064	0.2520
阴天	91.7	81.4	0.6170	0.4368
雨天	93.7	80.2	0.9288	0.5956

方法	L_RMSE/kW	L_MAPE/%
集群预测	0.3105	7.33
单点预测	0.3932	11.60