Comparative analysis of optimized LSTM architectures for long-term solar power forecasting

Year 2026, Volume: 10 Issue: 1 , 28 - 43 , 20.04.2026

Melisa Türker , Celal Yelgel , Övgü Ceyda Yelgel

https://doi.org/10.35860/iarej.1794285

https://izlik.org/JA22WB23BY

Abstract

Accurate long-term forecasting of photovoltaic (PV) power output is vital for ensuring the reliability and efficiency of renewable energy systems integrated into smart grids. This study proposes a systematic deep learning framework using Long Short-Term Memory (LSTM) networks, exploring the combined effects of network depth, activation functions (ReLU vs. Leaky ReLU), and optimization algorithms (Adam, Nadam, RMSprop, SGD) on predictive accuracy. A real-world dataset, recorded over one year in Van, Turkey, was used to evaluate the performance of 24 distinct model configurations. The results reveal that deeper LSTM architectures significantly improve generalization, particularly when paired with Leaky ReLU activation. Among all configurations, the four-layer LSTM model optimized with Nadam and activated with Leaky ReLU achieved the best forecasting performance, reaching a minimum SMAPE of 15.4%. This result represents a substantial improvement over conventional setups and demonstrates the effectiveness of adaptive optimization and nonlinear activation in capturing complex temporal patterns. The findings offer practical implications for enhancing grid stability, optimizing solar energy dispatch, and informing energy policy planning. The proposed model serves as a robust and scalable tool for energy stakeholders aiming to improve long-term solar forecasting in dynamically changing environments.

Keywords

Long-Term Solar PV Power Forecasting , Deep Learning (LSTM) Model , Sustainable Energy Systems

References

• 1. Foster, R., Ghassemi, M., and Cota, A., Solar energy: Renewable energy and the environment. 2009.
• 2. Diagne, M., David, M., Lauret, P., Boland, J., and Schmutz, N., Review of solar irradiance forecasting methods and a proposition for small-scale insular grids. 2013. 27: p. 65-76.
• 3. Calvet, W., et al., Locally resolved investigation of wedged Cu(In,Ga)Se2 films prepared by physical vapor deposition using hard X-ray photoelectron and X-ray fluorescence spectroscopy. Thin Solid Films, 2015. 582: p. 361–365.
• 4. Cheng, J., Zhang, M., Wu, G., Wang, X., Zhou, J., and Cen, K., Optimizing CO₂ reduction conditions to increase carbon atom conversion using a Pt-RGO||Pt-TNT photoelectrochemical cell. Solar Energy Materials and Solar Cells, 2015. 132: p. 606–614.
• 5. Lorenz, E., Hurka, J., Heinemann, D., and Beyer, H.G., Irradiance Forecasting for the Power Prediction of Grid-Connected Photovoltaic Systems. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2009. 2(1): p. 2–10.
• 6. Khaligh, A., and Onar, O.C., Energy harvesting: Solar, wind, and ocean energy conversion systems. 2017.
• 7. Jebli, I., Belouadha, F.Z., Kabbaj, M.I., and Tilioua, A., Prediction of solar energy guided by Pearson correlation using machine learning. Energy, 2021. 224: p. 120109.
• 8. Al Badwawi, R., Abusara, M., and Mallick, T., A Review of Hybrid Solar PV and Wind Energy System. Smart Science, 2015. 3(3): p. 127–138.
• 9. Mirzapour, F., Lakzaei, M., Varamini, G., Teimourian, M., and Ghadimi, N., A new prediction model of battery and wind-solar output in hybrid power system. Journal of Ambient Intelligence and Humanized Computing, 2019. 10(1): p. 77–87.
• 10. Kumar, K.P., and Saravanan, B., Recent techniques to model uncertainties in power generation from renewable energy sources and loads in microgrids – A review. Renewable and Sustainable Energy Reviews, 2017. 71: p. 348–358.
• 11. Abdel-Nasser, M., and Mahmoud, K., Accurate photovoltaic power forecasting models using deep LSTM-RNN. Neural Computing and Applications, 2019. 31(7): p. 2727–2740.
• 12. Mellit, A., Massi Pavan, A., and Lughi, V., Short-term forecasting of power production in a large-scale photovoltaic plant. Solar Energy, 2014. 105: p. 401–413.
• 13. Livera, A., Theristis, M., Makrides, G., and Georghiou, G.E., Advanced diagnostic approach of failures for grid-connected photovoltaic (PV) systems. In Proceedings of the 35th European Photovoltaic Solar Energy Conference, 2018.
• 14. Das, U.K., Tey, K.S., Seyedmahmoudian, M., Mekhilef, S., Idris, M.Y.I., Van Deventer, W., and Horan, B., Forecasting of photovoltaic power generation and model optimization: A review. Renewable and Sustainable Energy Reviews, 2018. 81: p. 912–928.
• 15. Gensler, A., Henze, J., Sick, B., and Raabe, N., Deep Learning for solar power forecasting – An approach using AutoEncoder and LSTM Neural Networks. In: IEEE International Conference on Systems, Man, and Cybernetics, 2016. p. 2858–2865.
• 16. Perera, M., De Hoog, J., Bandara, K., and Halgamuge, S., Multi-resolution, multi-horizon distributed solar PV power forecasting with forecast combinations. Expert Systems with Applications, 2022. 205: p. 117690.
• 17. Zhu, J., Li, M., Luo, L., Zhang, B., Cui, M., and Yu, L., Short-term PV power forecast methodology based on multi-scale fluctuation characteristics extraction. Renewable Energy, 2023. 208: p. 141–151.
• 18. Khan, Z.A., Hussain, T., and Baik, S.W., Dual stream network with attention mechanism for photovoltaic power forecasting. Applied Energy, 2023. 338: p. 120916.
• 19. Li, Y., Huang, W., Lou, K., Zhang, X., and Wan, Q., Short-term PV power prediction based on meteorological similarity days and SSA-BiLSTM. Systems and Soft Computing, 2024. 6: p. 200084.
• 20. Luo, X., Zhang, D., and Zhu, X., Deep learning based forecasting of photovoltaic power generation by incorporating domain knowledge. Energy, 2021. 225: p. 120240.
• 21. Gensler, A., Henze, J., Sick, B., and Raabe, N., Deep Learning for solar power forecasting - An approach using AutoEncoder and LSTM Neural Networks. IEEE International Conference on Systems, Man, and Cybernetics, 2016: p. 2858–2865.
• 22. Elsaraiti, M., and Merabet, A., Solar power forecasting using deep learning techniques. IEEE Access, 2022. 10: p. 31692–31698.
• 23. Qing, X., and Niu, Y., Hourly day-ahead solar irradiance prediction using weather forecasts by LSTM. Energy, 2018. 148: p. 461-468.
• 24. Gao, M., Li, J., Hong, F., and Long, D., Day-ahead power forecasting in a large-scale photovoltaic plant based on weather classification using LSTM. Energy, 2019. 187: p. 115838.
• 25. Han, S., Qiao, Y., Yan, J., Liu, Y., Li, L., and Wang, Z., Mid-to-long term wind and photovoltaic power generation prediction based on copula function and long short term memory network. Applied Energy, 2019. 239: p. 181-191.
• 26. Wang, K., Qi, X., and Liu, H., A comparison of day-ahead photovoltaic power forecasting models based on deep learning neural network. Applied Energy, 2019. 251: p. 113315.
• 27. Gao, M., Li, J., Hong, F., and Long, D., Day-ahead power forecasting in a large-scale photovoltaic plant based on weather classification using LSTM. Energy, 2019. 187: p. 115838.
• 28. Srivastava, S., and Lessmann, S., A comparative study of LSTM neural networks in forecasting day-ahead global horizontal irradiance with satellite data. Solar Energy, 2018. 162: p. 232–247.
• 29. Zhou, H., Zhang, Y., Yang, L., Liu, Q., Yan, K., and Du, Y., Short-Term photovoltaic power forecasting based on long short term memory neural network and attention mechanism. IEEE Access, 2019. 7: p. 78063–78074.
• 30. Yu, Y., Cao, J., and Zhu, J., An LSTM Short-Term Solar Irradiance Forecasting under Complicated Weather Conditions. IEEE Access, 2019. 7: p. 145651–145666.
• 31. Ghimire, S., Deo, R.C., Raj, N., and Mi, J., Deep solar radiation forecasting with convolutional neural network and long short-term memory network algorithms. Applied Energy, 2019. 253: p. 113541.
• 32. Wang, K., Qi, X., and Liu, H., Photovoltaic power forecasting based on LSTM-Convolutional Network. Energy, 2019. 189: p. 116225.
• 33. Qing, X., and Niu, Y., Hourly day-ahead solar irradiance prediction using weather forecasts by LSTM. Energy, 2018. 148: p. 461–468.
• 34. Shin, D., and Kim, C.B., Short term forecast model for solar power generation using RNN-LSTM. Journal of Advanced Navigation Technology, 2018. 22: p. 233-239.
• 35. Correa-Jullian, C., Cardemil, J.M., López Droguett, E., and Behzad, M., Assessment of Deep Learning techniques for Prognosis of solar thermal systems. Renewable Energy, 2020. 145: p. 2178–2191.
• 36. Wen, L., Zhou, K., Yang, S., and Lu, X., Optimal load dispatch of community microgrid with deep learning based solar power and load forecasting. Energy, 2019. 171: p. 1053–1065.
• 37. Hochreiter, S., and Schmidhuber, J., Long short-term memory. Neural Computation, 1997. 9(8): p. 1735–1780.
• 38. Rao, G., Huang, W., Feng, Z., and Cong, Q., LSTM with sentence representations for document-level sentiment classification. Neurocomputing, 2018. 308: p. 49–57.
• 39. Chai, T., and Draxler, R.R., Root mean square error (RMSE) or mean absolute error (MAE)? Arguments against avoiding RMSE in the literature. Geoscientific Model Development, 2014. 7(3): p. 1247–1250.
• 40. Ayyed, N.A., and Al-Sinjary, A.M., Comparison of Some Optimizers in Long Short-Term Memory Networks with an Application to Tigris River Water Imports. Asian Journal of Probability and Statistics, 2025. 27(1): p. 56–68.
• 41. Yelgel Ö.C., Yelgel, C. The Role of Machine Learning Methods fro Renewable Energy Technologies in Advances in Energy Recovery and Efficiency Technologies, 2024.
• 42. Lewinson, E. Choosing the correct error metric: MAPE vs. sMAPE. Towards Data Science. [Online] Available: https://towardsdatascience.com/choosing-the-correct-error-metric-mape-vs-smape-5328dec53fac
• 43. Konečný, J., and Richtárik, P., Semi-Stochastic Gradient Descent Methods. Frontiers in Applied Mathematics and Statistics, 2017. 3(9): p. 238564.
• 44. Türker, M., Yelgel, C., and Yelgel, Ö.C., Long-Term Prediction of Solar Panel Power Output with Artificial Intelligence Techniques. Ejons International Journal on Mathematic, Engineering and Natural Sciences, 2025. 9: p. 130.