arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2610.38976cs.LG
收藏

物理信息机器学习在太阳能发电预测中的效用

The Utility of Physics-Informed Machine Learning in Solar Power Production Forecasting

Kempelen Institute of Intelligent Technologies(肯佩伦智能技术学院);Faculty of Information Technology, Brno University of Technology(布尔诺理工大学信息技术学院);Know Center Research GmbH(诺中心研究有限公司);Signal Processing and Speech Communication Laboratory, Graz University of Technology(格拉茨理工大学信号处理与语音通信实验室);Research Center ENERGETIC, Graz University of Technology(格拉茨理工大学ENERGETIC研究中心)

Peter Pavlík, Franz M. Rohrhofer, Bernhard C. Geiger

AI总结:

本研究探究物理信息机器学习在大规模光伏发电预测中的效用,开发集成ADR物理模型与TSMixer架构的混合模型,发现其在测试精度等方面未较纯数据驱动基线有一致提升,对日前光伏预测益处有限。

摘要:

物理信息机器学习(PIML)与混合建模作为结合物理知识和深度学习灵活性的方法已受到广泛关注。本研究探究这些方法是否能在大规模光伏(PV)发电预测中实现其常被宣称的优势。我们开发两种混合预测架构,通过特征增强和残差学习将ADR物理光伏效率模型与TSMixer神经预测架构集成。所提模型在包含33个欧洲国家3713个光伏设施超300万个样本的数据集上评估,我们评估收敛速度、预测精度、跨区域迁移性和数据效率。尽管物理信息模型训练时收敛速度略快,但与纯数据驱动基线相比,它们在测试集精度、泛化性或有限训练数据下的性能上未实现一致提升。这些发现表明,对于在大数据集上训练的现代深度预测架构,显式集成ADR物理模型对日前光伏预测的实际益处有限。

英文摘要:

Physics-informed machine learning (PIML) and hybrid modeling have attracted significant attention as approaches for combining physical knowledge with the flexibility of deep learning. In this work, we investigate whether these approaches deliver their commonly claimed benefits for large-scale photovoltaic (PV) power forecasting. We develop two hybrid forecasting architectures that integrate the ADR physical PV efficiency model with the TSMixer neural forecasting architecture through feature augmentation and residual learning. The proposed models are evaluated on a dataset comprising over 3 million samples from 3,713 PV installations across 33 European countries. We assess convergence speed, forecasting accuracy, transferability across geographical regions, and data efficiency. Although the physics-informed models seem to exhibit slightly faster convergence during training, they do not achieve consistent improvements in test-set accuracy, generalization, or performance under limited training data compared to the purely data-driven baseline. These findings suggest that, for modern deep forecasting architectures trained on large datasets, explicitly incorporating the ADR physical model provides limited practical benefit for day-ahead PV forecasting.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑