Education, Science, Technology, Innovation and Life
Open Access
Sign In
  1. Home
  2. Journals
  3. Automation and Machine Learning
  4. Vol 7, Issue 1, 2026

A Decomposition–Graph Attention–Patch Transformer Framework with Conformal Calibration for Short-Term Wind Power Forecasting

Download as PDF

DOI: 10.23977/autml.2026.070109 | Downloads: 1 | Views: 32

Author(s)

Yirui Wang 1, Yijia Liu 2, Zhaomeng Zhang 3

Affiliation(s)

1 School of Intelligence Science and Technology, Beijing University of Civil Engineering and Architecture, Beijing, China
2 School of Geomatics and Urban Spatial Information‌, Beijing University of Civil Engineering and Architecture, Beijing, China
3 College of Computer Science, Beijing University of Technology, Beijing, China

Corresponding Author

Yirui Wang

ABSTRACT

Accurate short-term wind power forecasting is essential for power system dispatching because wind generation exhibits strong intermittency, non-stationarity, and nonlinear dependence on meteorological conditions. This study proposes a hybrid machine learning framework that integrates multi-resolution signal decomposition, spatio-temporal representation learning, and probabilistic uncertainty calibration for wind power forecasting. First, historical wind power and meteorological time series are decomposed into multiple components to alleviate non-stationary fluctuations and to enhance predictability across frequency bands. Second, spatial dependencies among wind turbines (or measurement sites) are modeled through a graph-based encoder using Graph Attention Network (GAT) to capture dynamic inter-node correlations. Third, temporal patterns are learned by a Patch-based Transformer that processes segmented sequence patches to efficiently represent long-range dependencies for multi-step forecasting horizons. Finally, probabilistic prediction intervals are produced via quantile regression and further calibrated using Conformal Prediction (CP) to improve coverage reliability under distribution shifts and missing observations. Experimental results on wind farm datasets demonstrate that the proposed framework achieves improved accuracy and robustness compared with classical statistical methods and representative deep learning baselines, while providing reliable uncertainty estimates for operational decision-making.

KEYWORDS

Wind power forecasting; Machine learning; Signal decomposition; Graph Attention Network; Patch-based Transformer; Quantile regression; Conformal prediction; Probabilistic forecasting

CITE THIS PAPER

Yirui Wang, Yijia Liu, Zhaomeng Zhang. A Decomposition–Graph Attention–Patch Transformer Framework with Conformal Calibration for Short-Term Wind Power Forecasting. Automation and Machine Learning (2026). Vol. 7, No. 1, 71-81. DOI: http://dx.doi.org/10.23977/autml.2026.070109.

REFERENCES

[1] E. G. S. Nascimento, T. A. C. de Melo, and D. M. Moreira, "A transformer-based deep neural network with wavelet transform for forecasting wind speed and wind energy," Energy, vol. 278, Art. no. 127678, 2023, doi: 10.1016/j.energy.2023.127678.
[2] D. A. Bispo Junior, T. I. Ren, and J. A. Silva, "Short-Term Wind Power Forecasting with Transformer-Based Models Enhanced by Time2Vec and Efficient Attention," Energies, vol. 18, no. 23, Art. no. 6162, 2025, doi: 10.3390/en18236162.
[3] M. Belletreche, A. Baïlek, and E. M. Berkouk, "Hybrid attention-based deep neural networks for short-term wind power forecasting using meteorological data in desert regions," Scientific Reports, vol. 14, Art. no. 21842, 2024, doi: 10.1038/s41598-024-73076-6.
[4] S. Daenens, T. Verstraeten, P.-J. Daems, A. Nowé, and J. Helsen, "Spatio-temporal graph neural networks for power prediction in offshore wind farms using SCADA data," Wind Energy Science, vol. 10, pp. 1137–1152, 2025, doi: 10.5194/wes-10-1137-2025.
[5] C. V. Zuege, J. P. Kuehn, and A. M. Foley, "Wind speed forecasting approach using conformal prediction and feature importance selection," International Journal of Electrical Power & Energy Systems, vol. 168, Art. no. 110700, 2025, doi: 10.1016/j.ijepes.2025.110700.
[6] A. Eikeland, F. M. Bianchi, and G. Reikard, "Probabilistic wind power forecasting with deep learning," Energy and AI, vol. 8, Art. no. 100062, 2022, doi: 10.1016/j.egyai.2022.100062.
[7] E. A. Tuncar, H. T. Tuncar, D. Colak, and H. Bulut, "Wind power forecasting: A comprehensive review of deep learning and hybrid models," Energy Reports, vol. 10, pp. 1106–1135, 2024, doi: 10.1016/j.egyr.2024.07.048.
[8] J. Keisler and E. Le Naour, "WindDragon: automated deep learning for regional wind power forecasting," Environmental Data Science, vol. 4, e19, 2025, doi: 10.1017/eds.2025.10.
[9] H. Alkabbani, F. Hourfar, A. Ahmadian, Q. Zhu, A. Almansoori, and A. Elkamel, "Machine learning-based time series modelling for large-scale regional wind power forecasting: A case study in Ontario, Canada,” Cleaner Energy Systems, vol. 5, Art. no. 100068, 2023, doi: 10.1016/j.cles.2023.100068.
[10] J. Jonkers, D. Nieves Avendano, G. Van Wallendael, and S. Van Hoecke, "A novel day-ahead regional and probabilistic wind power forecasting framework using deep CNNs and conformalized regression forests," Applied Energy, vol. 361, Art. no. 122900, 2024, doi: 10.1016/j.apenergy.2024.122900.
[11] S. Camal, R. Girard, M. Fortin, A. Touron, and L. Dubus, "A conditional and regularized approach for large-scale spatiotemporal wind power forecasting," Sustainable Energy Technologies and Assessments, vol. 65, Art. no. 103743, 2024, doi: 10.1016/j.seta.2024.103743.
[12] S. B. M. Bosma and N. Nazari, "Estimating solar and wind power production using computer vision deep learning techniques on weather maps," Energy Technology, vol. 10, no. 8, Art. no. 2200289, 2022, doi: 10.1002/ente.202200289.
[13] E. S. Miele, N. Ludwig, and A. Corsini, "Multi-horizon wind power forecasting using multi-modal spatio-temporal neural networks," Energies, vol. 16, no. 8, Art. no. 3522, 2023, doi: 10.3390/en16083522.
[14] S. M. J. Jalali, S. Ahmadian, M. Khodayar, A. Khosravi, M. Shafie-khah, S. Nahavandi, and J. P. S. Catalão, "An advanced short-term wind power forecasting framework based on the optimized deep neural network models," International Journal of Electrical Power & Energy Systems, vol. 141, Art. no. 108143, 2022, doi: 10.1016/j.ijepes.2022.108143.
[15] F. Lima, T. I. Ren, and A. Costa, "Wind power forecast based on transformers and clustering of wind farms with temporal and spatial interdependence," in Proc. International Joint Conference on Neural Networks (IJCNN), 2022, pp. 1–6, doi: 10.1109/IJCNN55064.2022.9892309.
[16] Y. Nie, N. H. Nguyen, P. Sinthong, and J. Kalagnanam, "A Time Series is Worth 64 Words: Long-term Forecasting with Transformers," in Proc. International Conference on Learning Representations (ICLR), 2023, doi: 10.48550/arXiv.2211.14730.

Subscription
E-Mail Alert
Most popular papers
Call for Special Issue
Publication Ethics & OA Statement
Plagiarism Policy
Contact Us
Downloads: 4905
Visits: 242217

Sponsors, Associates, and Links

All published work is licensed under a Creative Commons Attribution 4.0 International License.

Copyright © 2016 - 2031 Clausius Scientific Press Inc. All Rights Reserved.