WindFusion-EW: An Ensemble-Weighted Multi-Model Framework for Ultra-Short-Term Wind Power Forecasting
Received: 4 January 2026 | Revised: 13 February 2026 and 19 February 2026 | Accepted: 20 February 2026 | Online: 26 February 2026
Corresponding author: Suong Tieu
Abstract
Ultra-short-term wind power forecasting is an important technical task for supporting stable grid operation and efficient energy dispatch. However, the strong volatility and nonlinear characteristics of the wind limit the performance of conventional single-model forecasting approaches. This paper presents WindFusion-EW, a technical hybrid forecasting framework that combines two-stage signal decomposition with a weighted multi-model ensemble strategy. In this framework, CEEMDAN and Empirical Wavelet Transform (EWT) are sequentially applied to decompose raw wind power time series and extract informative multi-frequency components. Several transformer-based forecasting models are then trained in parallel, and an EnsembleWeighted mechanism dynamically adjusts model contributions based on month-wise validation performance. Experimental results on real-world wind power datasets from France and Turkey show that WindFusion-EW achieves average MAPE values of 1.02% and 1.13%, respectively, outperforming baseline and standalone deep learning models. The results demonstrate that integrating signal processing techniques with adaptive ensemble learning provides an effective and practical solution for ultra-short-term wind power forecasting.
Keywords:
CEEMDAN, EWT, Deep Learning, Wind Energy, WindFusion-EWDownloads
References
Y. Kassem, H. Camur, and A. A. S. Mosbah, "Feasibility Analysis of the Wind Energy Potential in Libya using the RETScreen Expert," Engineering, Technology & Applied Science Research, vol. 13, no. 4, pp. 11277–11289, Aug. 2023. DOI: https://doi.org/10.48084/etasr.6007
M. E. Torres, M. A. Colominas, G. Schlotthauer, and P. Flandrin, "A complete ensemble empirical mode decomposition with adaptive noise," in 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), May 2011, pp. 4144–4147. DOI: https://doi.org/10.1109/ICASSP.2011.5947265
I. Karijadi, S. Y. Chou, and A. Dewabharata, "Wind power forecasting based on hybrid CEEMDAN-EWT deep learning method," Renewable Energy, vol. 218, Dec. 2023, Art. no. 119357. DOI: https://doi.org/10.1016/j.renene.2023.119357
N. Fang, Z. Liu, and S. Fan, "Short-Term Wind Power Prediction Method Based on CEEMDAN-VMD-GRU Hybrid Model," Energies, vol. 18, no. 6, Mar. 2025. DOI: https://doi.org/10.3390/en18061465
Y. Su et al., "Frequency-aware ultra-short-term wind power forecasting using CEEMDAN–VMD–SE and Transformer–GRU networks," Energy, vol. 338, Nov. 2025, Art. no. 138715. DOI: https://doi.org/10.1016/j.energy.2025.138715
B. Lim, S. Ö. Arık, N. Loeff, and T. Pfister, "Temporal Fusion Transformers for interpretable multi-horizon time series forecasting," International Journal of Forecasting, vol. 37, no. 4, pp. 1748–1764, Oct. 2021. DOI: https://doi.org/10.1016/j.ijforecast.2021.03.012
H. Wu, J. Xu, J. Wang, and M. Long, "Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting," in Advances in Neural Information Processing Systems, 2021, vol. 34, pp. 22419–22430.
H. Wu, T. Hu, Y. Liu, H. Zhou, J. Wang, and M. Long, "TimesNet: Temporal 2D-Variation Modeling for General Time Series Analysis." arXiv, Apr. 12, 2023.
J. Tang, J. Lv, G. Yue, and X. Chen, "A Dual Decomposition and Dynamic Weighting-Based Integrated Forecasting Method for Ultra-Short-Term Wind Power," International Journal of High Speed Electronics and Systems, June 2025, Art. no. 2540639. DOI: https://doi.org/10.1142/S0129156425406394
Y. Wang, H. Xu, R. Zou, F. Zhang, and Q. Hu, "Dynamic non-constraint ensemble model for probabilistic wind power and wind speed forecasting," Renewable and Sustainable Energy Reviews, vol. 204, Oct. 2024, Art. no. 114781. DOI: https://doi.org/10.1016/j.rser.2024.114781
H. L. T. Khoa, "coderkhongodo/data_wind." Jan. 29, 2026, [Online]. Available: https://github.com/coderkhongodo/data_wind.
"Wind Turbine Scada Dataset." Kaggle, [Online]. Available: https://www.kaggle.com/datasets/berkerisen/wind-turbine-scada-dataset.
Downloads
How to Cite
License
Copyright (c) 2026 Thanh-Long Nguyen, Nguyen Khanh Vy, Minh Y. Nguyen, Huynh Ly Tan Khoa, Suong Tieu
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain the copyright and grant the journal the right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) after its publication in ETASR with an acknowledgement of its initial publication in this journal.
