Forecasting Wind Speed with Autoregressive and Long-Short Term Memory Neural Network Models

Year 2025, Volume: 9 Issue: 1, 98 - 121, 31.12.2025

Serkan Ansay , Bayram Köse , İbrahim Işıklı , Ceyda Mülayim , Bekir Ertilav

https://doi.org/10.61969/jai.1789834

Abstract

The growing need for sustainable and efficient energy systems has intensified interest in accurate renewable energy forecasting. In particular, wind speed forecasting is vital for the reliable integration of wind energy into power systems. This study investigates and compares the performance of three different approaches for wind speed prediction: Autoregressive (AR), Long Short-Term Memory (LSTM) neural networks, and a hybrid AR–LSTM model. Real wind speed data collected from İzmir, Turkey, were used in the experiments. The AR model, a linear statistical method, was evaluated alongside the LSTM model, a deep learning method capable of capturing long-term temporal dependencies. A hybrid model was also developed to benefit from the strengths of both. Additionally, noise reduction techniques such as Moving Average and Gaussian Filtering were applied to enhance data quality and model accuracy. The results demonstrated that the LSTM model achieved the lowest RMSE value (0.084), outperforming both the AR and hybrid models. This suggests that LSTM-based models are more suitable for capturing complex and nonlinear patterns in wind speed data. The findings contribute to the development of intelligent forecasting systems for efficient renewable energy management.

Keywords

Autoregressive , Artificial neural networks , Long-short term memory , Wind energy , Modeling of energy systems

Ethical Statement

This research did not involve human participants or animals and therefore did not require ethical approval.

Thanks

The authors would like to thank the General Directorate of Meteorology for their valuable contributions.

References

• Akın, E., & Şahin, M. E. (2024). A study on deep learning and artificial neural network models. EMO Scientific Journal, 14(1), 27–38.
• Albayrak, A. S. (2014). Autoregression techniques alternative to the least squares technique in the presence of autocorrelation and an application. Süleyman Demirel University Faculty of Economics and Administrative Sciences Journal, 19(1), 1–20.
• Altan, A., & Karasu, S. (2020). Investigation of the effect of decomposition methods on wind speed forecasting model performance defined by deep learning algorithm. European Journal of Science and Technology, (20), 844–853. https://doi.org/10.31590/ejosat.785699
• Ansay, S., & Köse, B. (2023). Precipitation forecast with artificial neural networks method. Journal of AI, 7(1), 15–31. https://doi.org/10.61969/jai.1310918
• Barış, C., Bilgin, A. C., & Altan, A. (2021). Determining the effect of intrinsic mode functions obtained by the empirical mode decomposition on the performance of deep learning based wind speed prediction model. European Journal of Science and Technology, (32), 654-660. https://doi.org/10.31590/ejosat.1026742
• Bento, P. M. R., Pombo, J. A. N., Calado, M. R. A., & Mariano, S. J. P. S. (2021). Stacking ensemble methodology using deep learning and ARIMA models for short-term load forecasting. Energies, 14(21), 7378. https://doi.org/10.3390/en14217378
• Berman, D. S., Buczak, A. L., Chavis, J. S., & Corbett, C. L. (2019). A survey of deep learning methods for cyber security. Information, 10(4), 122. https://doi.org/10.3390/info10040122
• Biswal, A., Borah, M. D., & Hussain, Z. (2021). Chapter eleven - Music recommender system using restricted Boltzmann machine with implicit feedback. In S. Kim & G. C. Deka (Eds.), Hardware Accelerator Systems for Artificial Intelligence and Machine Learning (Vol. 122, pp. 367–402). Elsevier. https://doi.org/10.1016/bs.adcom.2021.01.001
• Chang, C.-H. (2015). Deep and shallow architecture of multilayer neural networks. IEEE Transactions on Neural Networks and Learning Systems, 26(10), 2477–2486. https://doi.org/10.1109/TNNLS.2014.2387439
• Chen, H. (2023). A novel wind model downscaling with statistical regression and forecast. Journal of Cleaner Production, 426, 140217. https://doi.org/10.1016/j.jclepro.2023.140217
• Chen, H., Zhang, Q., & Birkelund, Y. (2022). Machine learning forecasts of Scandinavian numerical weather prediction wind model residuals with control theory for wind energy. Energy Reports, 8, 661–668. https://doi.org/10.1016/j.egyr.2022.08.105
• Çelebi, S. B., & Fidan, Ş. (2024). RNN-based time series analysis for wind turbine energy forecasting. International Journal of Engineering and Innovative Research, 6(1), 15–28. https://doi.org/10.47933/ijeir.1387314
• İnik, Ö. (2017). Datasets and software libraries used in deep learning. In E. Ülker (Ed.), 1st International Symposium on Multidisciplinary Studies and Innovative Technologies Proceedings Book. https://www.set-science.com/manage/uploads/ISAS-?go=d1001a2417e2b87d5b7c53e16c5e1675&conf_id=1&paper_id=16
• Jailani, N. L. M., et al. (2023). Investigating the power of LSTM-based models in solar energy forecasting. Processes, 11(5), 1382. https://doi.org/10.3390/pr11051382
• Karaman, Ö. A., & Bektaş, Y. (2023). Long-term electricity demand forecasting with machine learning and optimization methods: The case of Turkey. Journal of Engineering Sciences and Research, 5(2), 285–292. https://doi.org/10.46387/bjesr.1306577
• Kiriş, Z. N., Beyca, Ö. F., & Kosanoğlu, F. (2022). Recurrent neural networks based wind speed forecasting models: A case study of Yalova. Journal of Intelligent Systems: Theory and Applications, 5(2), 178–188. https://doi.org/10.38016/jista.1120383
• Köse, B., Aygün, H., & Pak, S. (2023). Parameter estimation of the wind speed distribution model by dragonfly algorithm. Journal of the Faculty of Engineering and Architecture of Gazi University, 38(3), 1747–1756. https://doi.org/10.17341/gazimmfd.935689
• Liu, L., Sun, Q., Wennersten, R., & Chen, Z. (2023). Day-ahead forecast of photovoltaic power based on a novel stacking ensemble method. IEEE Access, 11, 113593–113604. https://doi.org/10.1109/ACCESS.2023.3323526
• Navamani, T. M. (2019). Chapter 7 - Efficient deep learning approaches for health informatics. In A. K. Sangaiah (Ed.), Deep Learning and Parallel Computing Environment for Bioengineering Systems (pp. 123–137). Academic Press. https://doi.org/10.1016/B978-0-12-816718-2.00014-2
• Ömrüuzun, B., & Saldanlı, A. (2019). Cryptocurrency price modeling with artificial neural networks [Master’s thesis, İstanbul University].
• Özdemir, M. (2023). Investigation of Borsa Istanbul’s asymmetric dynamics with quantile autorepression approach. International Review of Economics and Management, 11(1), 57–74. https://doi.org/10.18825/iremjournal.1283918
• Özkişi, H., & Topaloğlu, M. (2017). The estimation of the photovoltaic cell productivity with the use of artificial neural network. Journal of Information Technologies, 10(3), 247–253. https://doi.org/10.17671/gazibtd.331035
• Öztürk, G., & Eldoğan, O. (2024). Prediction of multivariate chaotic time series using GRU, LSTM and RNN. Sakarya University Journal of Computer and Information Sciences, 7(2), 156–172. https://doi.org/10.35377/saucis...1404116
• Sarker, I. H. (2021). Deep learning: A comprehensive overview on techniques, taxonomy, applications and research directions. SN Computer Science, 2, 420. https://doi.org/10.1007/s42979-021-00815-1
• Ser, G., & Bati, C. T. (2019). Determining the best model with deep neural networks: Keras application on mushroom data. Yüzüncüyıl University Journal of Agricultural Sciences, 29(3), 406–417. https://doi.org/10.29133/yyutbd.505086
• Serov, A. N., Ivanenko, K. A., & Chumachenko, D. A. (2021). Applying a moving average filter to reduce the frequency measurement error for the case applying by the zero crossing technique. In 2021 Systems of Signal Synchronization, Generating and Processing in Telecommunications (SYNCHROINFO) (pp. 1–6). IEEE. https://doi.org/10.1109/SYNCHROINFO51390.2021.9488363
• Sucu, İ. (2019). The effect of artifiticial intelligence on society and artificial intelligence the view of artificial intelligence in the context of film (I.A.). International Journal of Textbooks and Educational Materials, 2(2), 203–215. https://dergipark.org.tr/tr/pub/ijotem/issue/50671/643204
• Şeker, A., Diri, B., & Balık, H. H. (2017). A review about deep learning methods and applications. Gazi Journal of Engineering Sciences, 3(3), 47–64. https://dergipark.org.tr/en/pub/gmbd/issue/31064/372661
• Şenkal, S., & Emeksiz, C. (2023). The effect of data decomposition on prediction performance in wind speed prediction with artificial neural network. International Scientific and Vocational Studies Journal, 7(2), 213–223. https://doi.org/10.47897/bilmes.1406384
• Ubal, C., Di-Giorgi, G., Contreras-Reyes, J. E., & Salas, R. (2023). Predicting the long-term dependencies in time series using recurrent artificial neural networks. Machine Learning and Knowledge Extraction, 5(4), 1340–1358. https://doi.org/10.3390/make5040068
• Vásquez-Coronel, J. A., Mora, M., & Vilches, K. (2023). A review of multilayer extreme learning machine neural networks. Artificial Intelligence Review, 56(11), 13691–13742. https://doi.org/10.1007/s10462-023-10478-4
• Vennerød, C. B., Kjærran, A., & Bugge, E. S. (2021). Long short-term memory RNN. arXiv. https://doi.org/10.48550/arXiv.2105.06756
• Zhang, F., & Fu, J. (2016). Noise elimination based on moving average by Gaussian distribution weighting method. In 2016 2nd International Conference on Control, Automation and Robotics (ICCAR) (pp. 169–172). IEEE. https://doi.org/10.1109/ICCAR.2016.7486720