Comparative evaluation of recurrent and attention-based deep learning models for wind speed forecasting
FME Transactions · 2025 · 53 (4) · 693-704
DOI: 10.5937/fme2504693a ↗Kurz gefasst
Fachartikel in der Zeitschrift FME Transactions, 2025. Verfasst von Qahtani Al und 3 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 36 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Wind speed forecasting is critical for enhancing wind energy output and improving efficiency. Although traditional statistical and physical models are commonly used but fail to capture the complex, non-linear, and time-varying nature of meteorological data. Therefore, this study explores four deep learning architectures: Recurrent Neural Network (RNN), Gated Recurrent Unit (GRU), Long Short-Term Memory (LSTM), and transformer. These models are analyzed using multivariate and hourly meteorological data from the Al-Ahsa region over 24 years (2001-2024). The dataset includes five features: wind speed at 10 meters above ground (used as a reference), air temperature at 2 meters, specific humidity at 2 meters, wind direction at 10 meters, and surface pressure. The models were trained on 80% of the dataset, validated on 10%, and tested on the remaining 10%. The GRU model achieved the best performance with RMSE = 0.3126 m/s and R² = 0.9759.
Zitieren
APA
Al, Q., Abolouz, M., Mohandes, M., & Rehman, S. (2025). Comparative evaluation of recurrent and attention-based deep learning models for wind speed forecasting. FME Transactions, 53(4), 693-704. https://doi.org/10.5937/fme2504693a
Vancouver
Al Q, Abolouz M, Mohandes M, Rehman S. Comparative evaluation of recurrent and attention-based deep learning models for wind speed forecasting. FME Transactions. 2025;53(4):693-704. doi:10.5937/fme2504693a
BibTeX
@article{Al2025,
author = {Al, Qahtani and Abolouz, Mohamed and Mohandes, Mohamed and Rehman, Shafiqur},
title = {Comparative evaluation of recurrent and attention-based deep learning models for wind speed forecasting},
journal = {FME Transactions},
year = {2025},
volume = {53},
number = {4},
pages = {693--704},
doi = {10.5937/fme2504693a}
}
EUVIMED