Comparative Analysis of Deep Recurrent Neural Network Models for Ultra Short-Term Wind Power Forecasting

The rapid growth of global wind power capacity has made accurate forecasting crucial for grid stability, economical allocation and efficient operation of wind farms. In this paper, we investigate the performance of various deep recurrent neural network (DRNN) models, including GRU, LSTM, BiGRU, and BiLSTM, deeper LTSM and LSTM-GRU for ultra short-term wind power forecasting. Furthermore, we incorporate a self-attention mechanism into LSTM (SA-LSTM) and GRU (SA-GRU) models to capture complex temporal dependencies and enhance forecasting accuracy. Extensive experiments on real-world wind power datasets demonstrate that the SA-LSTM and SA-GRU modes slightly outperform conventional architectures with an accuracy of 97%, while all evaluated models exhibit strong capabilities in capturing short-term fluctuations in wind power. These results emphasize the potential of advanced deep learning models to support real-time wind farm control, energy management and grid integration.

[1]   Wind power capacity in the Net Zero Scenario, 2015–2030, Int. Energy Agency (IEA). Available: https://www.iea.org/data-and-statistics/charts/wind-power-capacity-in-the-net-zero-scenario-2015-2030

[2]   F. Rodr?guez, A.M. Florez-Tapia, L. Font?n, A. Galarza, Very short-term wind power density forecasting through artificial neural networks for microgrid control, Renew. Energy 145 (2020) 1517–1527.

[3]   G. Yu, L. Shen, Q. Dong, G. Cui, S. Wang, D. Xin, X. Chen, W. Lu, Ultra-short-term wind power forecasting techniques: Comparative analysis and future trends, Front. Energy Res. 11 (2024) 1345004.

[4]   W.C. Tsai, C.M. Hong, C.S. Tu, W.M. Lin, C.H. Chen, A review of modern wind power generation forecasting technologies, Sustainability 15 (2023) 10757.

[5]   M.H. Lipu, M.S. Miah, M.A. Hannan, A. Hussain, M.R. Sarker, A. Ayob, M.H.M. Saad, M.S. Mahmud, Artificial intelligence based hybrid forecasting approaches for wind power generation: Progress, challenges and prospects, IEEE Access 9 (2021) 102460–102489.

[6]   H.H. Cevik, M. Cunka?, K. Polat, A new multistage short-term wind power forecast model using decomposition and artificial intelligence methods, Physica A 534 (2019) 122177.

[7]   B. Kosovic, S.E. Haupt, D. Adriaansen, S. Alessandrini, G. Wiener, L. Delle Monache, Y. Liu, S. Linden, T. Jensen, W. Cheng, M. Politovich, A comprehensive wind power forecasting system integrating artificial intelligence and numerical weather prediction, Energies 13 (2020) 1372.

[8]   M. Godinho, R. Castro, Comparative performance of AI methods for wind power forecast in Portugal, Wind Energy 24 (2021) 39–53.

[9]   L. Xiang, J. Liu, X. Yang, A. Hu, H. Su, Ultra-short term wind power prediction applying a novel model named SATCN-LSTM, Energy Convers. Manag. 252 (2022) 115036.

[10] S. Qureshi, F. Shaikh, L. Kumar, F. Ali, M. Awais, A.E. Gürel, Short-term forecasting of wind power generation using artificial intelligence, Environ. Chall. 11 (2023) 100722.

[11]  L. Liu, J. Liu, Y. Ye, H. Liu, K. Chen, D. Li, X. Dong, M. Sun, Ultra-short-term wind power forecasting based on deep Bayesian model with uncertainty, Renew. Energy 205 (2023) 598–607.

[12] Q. Li, X. Ren, F. Zhang, L. Gao, B. Hao, A novel ultra-short-term wind power forecasting method based on TCN and Informer models, Comput. Electr. Eng. 120 (2024) 109632.

[13] S. Yang, J. Zhou, X. Gu, Y. Mei, J. Duan, A comprehensive framework of the decomposition-based hybrid method for ultra-short-term wind power forecasting with on-site application, Energy 313 (2024) 133911.

[14] B. Kishore, K. Kabilan, L.S. Rahul, G.P. Priyadarshan, E.P. Vishnutheerth, R. Satheesh, M.L. Kolhe, Advancing short-term wind power forecasting by AI-driven models for improved accuracy, Electr. Eng. 107 (2025) 14795–14810.

[15]  C. Zhang, X. Xiao, Y. Wang, M.Z. Hou, S. Huang, W. Hu, M. Hu, R. Huang, Fine-grained ultra-short-term wind power forecasting based on Temporal Fusion Transformers integrated with turbine power time series clustering, Energy (2025) 137995.

[16]  X. Shu, L. Gao, F. Zhang, X. Ren, L. Qin, Y. Wang, X. Wu, Ultra-short-term wind power forecasting method based on optimized decomposition and deep learning, Energy Convers. Manag. X (2025) 101315.

[17] W. Li, T. Yang, J. Yang, L. Peng, Predicting ultra-short-term wind power combinations under extreme weather conditions, IEEE Access (2025).

[18]  P. Xu, S. Zhang, P. Luo, X. Wang, Y. Yan, D. Qi, Data-feature hybrid optimization for ultra-short-term wind power forecasting, Int. J. Electr. Power Energy Syst. 176 (2026) 111742.

[19] Mellit, A.M. Pavan, V. Lughi, Deep learning neural networks for short-term photovoltaic power forecasting, Renew. Energy 172 (2021) 276–288.

[20] Hochreiter, S. and Schmidhuber, J.,. Long short-term memory. Neural computation, 9(8), (1997) pp.1735-1780.

[21] Han, Z., Zhao, J., Leung, H., Ma, K.F. and Wang, W., 2019. A review of deep learning models for time series prediction. IEEE Sensors Journal, 21(6), pp.7833-7848.

[22] Boubaker, S., Benghanem, M., Mellit, A., Lefza, A., Kahouli, O. and Kolsi, L., Deep neural networks for predicting solar radiation at Hail Region, Saudi Arabia. IEEE Access, 2021, 36719-36729.

[23] Gamboa, J.C.B. Deep learning for time-series analysis. arXiv, 2017, preprint arXiv:1701.01887.

[24] Cho, K., Van Merriënboer, B., Bahdanau, D., & Bengio, Y. On the properties of neural machine translation: Encoder–decoder approaches. 2014, ArXiv:1409.1259.

[25] Xu, L., Xu, W., Cui, Q., Li, M., Luo, B. and Tang, Y., 2023. Deep heuristic evolutionary regression model based on the fusion of BiGRU and BiLSTM. Cognitive Computation, 15(5), pp.1672-1686.

[26] Sagheer, A. and Kotb, M.,. Unsupervised pre-training of a deep LSTM-based stacked autoencoder for multivariate time series forecasting problems. Scientific reports, 9(1), (2019), p.19038.

[27]  W. Li, F. Qi, M. Tang, Z. Yu, Bidirectional LSTM with self-attention mechanism and multi-channel features for sentiment classification, Neurocomputing 387 (2020) 63–77.

[28] Kilinc, H.C., Apak, S., Ozkan, F., Ergin, M.E. and Yurtsever, A., 2024. Multimodal fusion of optimized GRU–LSTM with self-attention layer for hydrological time series forecasting. Water Resources Management, 38(15), pp.6045-6062.

[29]   Z. Fu, H. Yang, A.M.C. So, W. Lam, L. Bing, N. Collier, On the effectiveness of parameter-efficient fine-tuning, Proc. AAAI Conf. Artif. Intell. 37 (2023) 12799–12807.

[30] P. Dhal, C. Azad, A fine-tuning deep learning with multi-objective-based feature selection approach for the classification of text, Neural Comput. Appl. 36 (2024) 3525–3553.