基于改进CNN-Autoformer网络的光伏功率短期概率预测方法

朱文志; 郭力; 刘一欣; 李彦榕; 李西良; 吴翠姑

doi:10.19912/j.0254-0096.tynxb.2024-1968

PDF(1596 KB)

太阳能学报 ›› 2026, Vol. 47 ›› Issue (3) : 678-689. DOI: 10.19912/j.0254-0096.tynxb.2024-1968

基于改进CNN-Autoformer网络的光伏功率短期概率预测方法

朱文志¹, 郭力¹, 刘一欣¹, 李彦榕¹, 李西良^2,3, 吴翠姑^2,3

作者信息 +

SHORT-TERM PROBABILISTIC PREDICTION OF PV POWER BASED ON IMPROVED CNN-AUTOFORMER NETWORK

Zhu Wenzhi¹, Guo Li¹, Liu Yixin¹, Li Yanrong¹, Li Xiliang^2,3, Wu Cuigu^2,3

Author information +

文章历史 +

摘要

针对光伏功率短期预测面临的不确定性难以准确刻画问题,提出一种基于改进CNN-Autoformer网络的光伏功率短期概率预测方法。首先,采用卷积神经网络提取并建立数值天气预报高维气象特征与光伏功率之间的映射关系;其次,利用自组织映射神经网络对天气类型进行削减归类,并作为光伏功率日序列的离散特征。在此基础上,构建时序Autoformer网络深度分解光伏功率时间序列,引入自相关机制捕捉光伏功率时间序列的周期性和趋势性特征;最后,结合极大似然估计与梯度优化,通过概率密度估计层输出得到光伏功率的概率分布参数。算例结果表明,相较于传统预测方法,所提方法在提高光伏功率概率预测性能方面具有明显优势。

Abstract

Addressing the challenge of accurately characterizing uncertainties in short-term photovoltaic （PV） power forecasting, this paper proposes a short-term PV probabilistic prediction method based on an improved CNN-Autoformer network. Firstly, convolutional neural network is used to extract and establish a mapping relationship between high-dimensional meteorological features and PV output based on numerical weather prediction. Secondly, a self-organizing map （SOM） neural network is employed to reduce and categorize weather types as discrete features of the daily PV sequence. Based on this, a temporal Autoformer network is constructed to deeply decompose the PV sequence, incorporating an autocorrelation mechanism to capture the periodicity and trend features. Finally, combining maximum likelihood estimation with gradient optimization, the parameters of the PV output probabilistic distribution are derived through a probability density estimation layer. Simulation results demonstrate that the proposed method can effectively improve the performance of PV probabilistic prediction compared to the comparative methods.

导出引用

朱文志, 郭力, 刘一欣, 李彦榕, 李西良, 吴翠姑. 基于改进CNN-Autoformer网络的光伏功率短期概率预测方法[J]. 太阳能学报. 2026, 47(3): 678-689 https://doi.org/10.19912/j.0254-0096.tynxb.2024-1968

Zhu Wenzhi, Guo Li, Liu Yixin, Li Yanrong, Li Xiliang, Wu Cuigu. SHORT-TERM PROBABILISTIC PREDICTION OF PV POWER BASED ON IMPROVED CNN-AUTOFORMER NETWORK[J]. Acta Energiae Solaris Sinica. 2026, 47(3): 678-689 https://doi.org/10.19912/j.0254-0096.tynxb.2024-1968

中图分类号： TM615

参考文献

[1] TOUBEAU J F, BOTTIEAU J, DE GRÈVE Z, et al. Data-driven scheduling of energy storage in day-ahead energy and reserve markets with probabilistic guarantees on real-time delivery[J]. IEEE transactions on power systems, 2021, 36(4): 2815-2828.
[2] 赵洪山, 孙承妍, 温开云, 等. 基于有向图卷积循环网络的分布式光伏出力超短期预测方法[J]. 太阳能学报, 2024, 45(8): 281-288.
ZHAO H S, SUN C Y, WEN K Y, et al.Ultra-short-term prediction method of distributed photovoltaic power output based on directed graph convolution recurrent network[J]. Acta energiae solaris sinica, 2024, 45(8): 281-288.
[3] 孙玉玺, 刘寅韬, 耿光超, 等. 基于多模式增量更新的短期光伏功率预测方法[J]. 太阳能学报, 2024, 45(9): 386-393.
SUN Y X, LIU Y T, GENG G C, et al.Short-term photovoltaic power forecasting method based on multi-mode incremental update[J]. Acta energiae solaris sinica, 2024, 45(9): 386-393.
[4] WU T, HU R F, ZHU H Y, et al.Combined IXGBoost-KELM short-term photovoltaic power prediction model based on multidimensional similar day clustering and dual decomposition[J]. Energy, 2024, 288: 129770.
[5] XU S W, WU W C.Tractable reformulation of two-side chance-constrained economic dispatch[J]. IEEE transactions on power systems, 2022, 37(1): 796-799.
[6] 张宇华, 时鑫洋, 颜楠楠, 等. 逆向云灰色关联相似日的EEMD-RL-GWO-LSTM区域风光功率短期预测[J]. 太阳能学报, 2024, 45(10): 144-152.
ZHANG Y H, SHI X Y, YAN N N, et al.Short-term prediction of regional wind-solar power of EEMD-RL-GWO-LSTM on reverse cloud grey correlation similar days[J]. Acta energiae solaris sinica, 2024, 45(10): 144-152.
[7] 时培明, 郭轩宇, 杜清灿, 等. 基于TCN-BiLSTM-Attention-ESN的光伏功率预测[J]. 太阳能学报, 2024, 45(9): 304-316.
SHI P M, GUO X Y, DU Q C, et al.Photovoltaic power prediction based on TCN-BiLSTM-Attention-ESN[J]. Acta energiae solaris sinica, 2024, 45(9): 304-316.
[8] JUNG Y, JUNG J, KIM B, et al.Long short-term memory recurrent neural network for modeling temporal patterns in long-term power forecasting for solar PV facilities: case study of South Korea[J]. Journal of cleaner production, 2020, 250: 119476.
[9] KORKMAZ D.SolarNet: a hybrid reliable model based on convolutional neural network and variational mode decomposition for hourly photovoltaic power forecasting[J]. Applied energy, 2021, 300: 117410.
[10] ABOU HOURAN M, SALMAN BUKHARI S M, ZAFAR M H, et al. COA-CNN-LSTM: coati optimization algorithm-based hybrid deep learning model for PV/wind power forecasting in smart grid applications[J]. Applied energy, 2023, 349: 121638.
[11] LIN T Y, WANG Y X, LIU X Y, et al.A survey of transformers[J]. AI open, 2022, 3: 111-132.
[12] KHAN Z A, HUSSAIN T, BAIK S W.Dual stream network with attention mechanism for photovoltaic power forecasting[J]. Applied energy, 2023, 338: 120916.
[13] ZHOU H Y, ZHANG S H, PENG J Q, et al.Informer: beyond efficient transformer for long sequence time-series forecasting[J]. Proceedings of the AAAI conference on artificial intelligence, 2021, 35(12): 11106-11115.
[14] WU H X, XU J H, WANG J M, et al.Autoformer: decomposition transformers with auto-correlation for long-term series forecasting[C]//Neural Information Processing Systems, 2021.
[15] KITAEV N, KAISER Ł, LEVSKAYA A. Reformer: the efficient transformer[EB/OL].2020: arXiv: 2001.04451. https://arxiv.org/abs/2001.04451
[16] BAN G H, CHEN Y, XIONG Z H, et al.The univariate model for long-term wind speed forecasting based on wavelet soft threshold denoising and improved Autoformer[J]. Energy, 2024, 290: 130225.
[17] CHEN J, PENG T, QIAN S J, et al.An error-corrected deep Autoformer model via Bayesian optimization algorithm and secondary decomposition for photovoltaic power prediction[J]. Applied energy, 2025, 377: 124738.
[18] SAEED A, LI C S, GAN Z H, et al.A simple approach for short-term wind speed interval prediction based on independently recurrent neural networks and error probability distribution[J]. Energy, 2022, 238: 122012.
[19] 王东风, 刘婧, 黄宇, 等. 结合太阳辐射量计算与CNN-LSTM组合的光伏功率预测方法研究[J]. 太阳能学报, 2024, 45(2): 443-450.
WANG D F, LIU J, HUANG Y, et al.Photovoltaic power prediction method combinating solar radiation calculation and CNN-LSTM[J]. Acta energiae solaris sinica, 2024, 45(2): 443-450.
[20] 师浩琪, 郭力, 刘一欣, 等. 基于多源气象预报总辐照度修正的光伏功率短期预测[J]. 电力自动化设备, 2022, 42(3): 104-112.
SHI H Q, GUO L, LIU Y X, et al.Short-term forecasting of photovoltaic power based on total irradiance correction of multi-source meteorological forecast[J]. Electric power automation equipment, 2022, 42(3): 104-112.
[21] 马乐乐, 孔小兵, 郭磊, 等. 基于最大重叠离散小波变换和深度学习的光伏功率预测[J]. 太阳能学报, 2024, 45(5): 576-583.
MA L L, KONG X B, GUO L, et al.Photovoltaic power forecasting based on maximum overlap discrete wavelet transform and deep learning[J]. Acta energiae solaris sinica, 2024, 45(5): 576-583.
[22] 孟亦康, 许野, 王鑫鹏, 等. 基于相似日选取和PCA-LSTM的光伏出力组合预测模型研究[J]. 太阳能学报, 2024, 45(7): 453-461.
MENG Y K, XU Y, WANG X P, et al.Research on photovoltaic output combination prediction model based on similar day selection and PCA-LSTM[J]. Acta energiae solaris sinica, 2024, 45(7): 453-461.
[23] JIANG Y Q, GAO T L, DAI Y X, et al.Very short-term residential load forecasting based on deep-autoformer[J]. Applied energy, 2022, 328: 120120.
[24] WANG Z Q, CHEN Z H, YANG Y, et al.A hybrid Autoformer framework for electricity demand forecasting[J]. Energy reports, 2023, 9: 3800-3812.
[25] 韩宇超, 同向前, 邓亚平. 基于概率密度估计与时序Transformer网络的风功率日前区间预测[J]. 中国电机工程学报, 2024, 44(23): 9285-9296.
HAN Y C, TONG X Q, DENG Y P.Probabilistic distribution estimation and temporal Transformer-based interval prediction in day-ahead wind power prediction[J]. Proceedings of the CSEE, 2024, 44(23): 9285-9296.