针对光伏输出功率波动显著且预测难度较大的问题,提出一种基于相似日聚类的WOA-VMD-Transformer的组合光伏功率预测模型。首先,利用K-means++算法进行相似日聚类;然后,采用鲸鱼优化算法（WOA）对变分模态分解（VMD）的参数进行寻优,将光伏功率序列分解为多个本征模态函数（IMF）;将IMF分量和气象因子加权合并成新的特征向量输入后续模型;并基于TCN-Transformer模型,分别预测不同天气类型下的IMF,叠加后得到预测值。最后,以澳大利亚中部爱丽丝泉沙漠太阳能研究中心的Hanwha Solar光伏场站一年的光伏发电和气象数据作为实例,对模型的有效性加以验证。消融实验和综合评估表明,所提模型在各类天气下均可取得较高的预测精度。
Abstract
To solve the problem that PV power fluctuates significantly and is difficult to predict, this paper proposes a combined PV power prediction model based on similar day clustering and an WOA-WMD-TCN-Transformer model. Firstly, K-means ++ is used to cluster similar days. Then WOA was used to optimize VMD parameters, and the PV power sequence was decomposed into multiple Intrinsic Mode functions（IMFs）. The IMF components and meteorological factors were weighted and combined into a new feature vector and fed into the subsequent model. Based on TCN-Transformer, IMF under different weather conditions can be predicted separately and the predicted value can be obtained after superposition. Finally, the photovoltaic power generation and meteorological data of Hanwha Solar Photovoltaic Station, a desert solar Research Center in Alice Springs, Central Australia, were used as an example to verify the validity of the model. Ablation experiments and comprehensive evaluation show that the proposed model can achieve high prediction accuracy under various weather conditions.
关键词
预测 /
深度学习 /
变分模态分解 /
相似日聚类 /
TCN-Transformer /
光伏
Key words
forecasting /
deep learning /
variational mode decomposition /
similar day clustering /
TCN-Transformer /
photovoltaic
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 谢小瑜, 周俊煌, 张勇军. 深度学习在泛在电力物联网中的应用与挑战[J]. 电力自动化设备, 2020, 40(4): 77-87.
XIE X Y, ZHOU J H, ZHANG Y J.Application and challenge of deep learning in ubiquitous power Internet of Things[J]. Electric power automation equipment, 2020, 40(4): 77-87.
[2] KHALID H M, FLITTI F, MAHMOUD M S, et al.Wide area monitoring system operations in modern power grids: a Median regression function-based state estimation approach towards cyber attacks[J]. Sustainable energy, grids and networks, 2023, 34: 101009.
[3] YANG D Z, JIRUTITIJAROEN P, WALSH W M.Hourly solar irradiance time series forecasting using cloud cover index[J]. Solar energy, 2012, 86(12): 3531-3543.
[4] SHI J, LEE W J, LIU Y Q, et al.Forecasting power output of photovoltaic systems based on weather classification and support vector machines[J]. IEEE transactions on industry applications, 2012, 48(3): 1064-1069.
[5] 马伟, 乔颖, 谢丽蓉, 等. 考虑气象特征波动划分阈值的双目标短期风功率预测[J]. 高电压技术, 2022, 48(10): 4154-4162.
MA W, QIAO Y, XIE L R, et al.Short-term wind power prediction with dual targets considering the threshold of meteorological characteristic fluctuation partition[J]. High voltage engineering, 2022, 48(10): 4154-4162.
[6] 孟安波, 许炫淙, 陈嘉铭, 等. 基于强化学习和组合式深度学习模型的超短期光伏功率预测[J]. 电网技术, 2021, 45(12): 4721-4728.
MENG A B, XU X C, CHEN J M, et al.Ultra short term photovoltaic power prediction based on reinforcement learning and combined deep learning model[J]. Power system technology, 2021, 45(12): 4721-4728.
[7] 王粟, 隗磊锋, 曾亮. 基于GWO-SVM与随机森林的组合光伏功率预测模型[J]. 昆明理工大学学报(自然科学版), 2021, 46(5): 82-88.
WANG S, KUI L F, ZENG L.A combined model for photovoltaic power forecasting based on GWO-SVM and random forest[J]. Journal of Kunming University of Science and Technology (natural sciences), 2021, 46(5): 82-88.
[8] 叶畅, 柳丹, 曹侃. 基于云图特征的超短期光伏发电功率预测模型[J]. 电网与清洁能源, 2023, 39(10): 70-79.
YE C, LIU D, CAO K.An ultra-short-term photovoltaic power forecasting model based on cloud features[J]. Power system and clean energy, 2023, 39(10): 70-79.
[9] LI M, WANG W, HE Y, et al.Deep learning model for short-term photovoltaic power forecasting based on variational mode decomposition and similar day clustering[J]. Computers and electrical engineering, 2024, 115: 109116.
[10] 殷豪, 欧祖宏, 陈德, 等. 基于二次模式分解和级联式深度学习的超短期风电功率预测[J]. 电网技术, 2020, 44(2): 445-453.
YIN H, OU Z H, CHEN D, et al.Ultra-short-term wind power prediction based on two-layer mode decomposition and cascaded deep learning[J]. Power system technology, 2020, 44(2): 445-453.
[11] 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.
MA L B, PAN G B, JIANG Q, 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.
[12] MACQUEEN J.Some methods for classification and analysis of multivariate observations[R]. US government technology report, 1966.
[13] DRAGOMIRETSKIY K, ZOSSO D.Variational mode decomposition[J]. IEEE transactions on signal processing, 2014, 62(3): 531-544.
[14] XU Y.Development of commercial vehicle E-axle system based on NVH performance optimization[R]. SAE report, 2020.
[15] 李争, 张杰, 徐若思, 等. 基于相似日聚类和PCC-VMD-SSA-KELM模型的短期光伏功率预测[J]. 太阳能学报, 2024, 45(2): 460-468.
LI Z, ZHANG J, XU R S, et al.Short term photovoltaic power prediction based on similar day clustering and PCC-VMD-SSA-KELM model[J]. Acta energiae solaris sinica, 2024, 45(2): 460-468.
[16] CHEN C, LI X M, YANG L J, et al.Application of variational mode decomposition in power system harmonic detection[J]. Power system protection and control, 2018, 46(14): 63-70.
[17] LEA C, FLYNN M D, VIDAL R, et al.Temporal convolutional networks for action segmentation and detection[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition(CVPR), Honolulu, HI, USA, 2017: 1003-1012.
[18] SUBAKAN C, RAVANELLI M, CORNELL S, et al.Attention is all you need in speech separation[C]//ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Toronto, ON, Canada, 2021: 21-25.
[19] 张程, 林谷青, 匡宇. 基于MEEMD-QUATRE-BILSTM的短期光伏出力区间预测[J]. 太阳能学报, 2023, 44(11): 40-54.
ZHANG C, LIN G Q, KUANG Y.Short-term PV output interval prediction based on MEEMD-QUATRE-BILSTM[J]. Acta energiae solaris sinica, 2023, 44(11): 40-54.
基金
广州市科技局重点研发计划：基于漂浮式海上风电平台风光融合的绿氢制取关键技术研究（202402gx0012）
PDF(3964 KB)
