Comparing Deep Learning and Ensemble Models for Photovoltaic Power Forecasting

Edris Shamsulhaq; Fikri Arif Wicaksana

doi:10.33019/

Authors

Edris Shamsulhaq Nusa Putra University
Fikri Arif Wicaksana Nusa Putra University

DOI:

https://doi.org/10.33019/

Keywords:

Deep Learning, Ensemble Learning, LSTM, Photovoltaic Faorecasting, SHAP Analysis, XGBoost

Abstract

Precise forecasting of photovoltaic (PV) power is crucial for maintaining grid stability and effectively incorporating solar energy. This study compares four forecasting models, LSTM, GRU, Random Forest, and XGBoost, using three years of real operational data (2023–2025) from the DKASC solar database. Fourteen input features were constructed from irradiance measurements, meteorological variables, rainfall, wind direction, and cyclical time features. A sliding window of 36 time steps captured temporal dependencies, while the dataset was chronologically split into training, validation, and testing subsets. Model performance was assessed using R², RMSE, MAE, and MAPE. Results show LSTM achieved the highest accuracy with R² of 0.9647 and RMSE of 88.40 W, whereas XGBoost obtained the lowest MAPE of 19.18%. SHAP analysis and an ablation study were conducted to identify influential features and evaluate their contributions to forecasting performance. SHAP analysis identified global horizontal radiation, global tilted radiation, and cyclical temporal features, particularly day_cos and hour_cos, as the most influential predictors, while wind speed and daily rainfall showed limited impact. The findings demonstrate that deep learning models can provide reliable PV power forecasts when combined with informative feature engineering and appropriate temporal window selection.

Downloads

Download data is not yet available.

References

[1] Z. A. Aljawary et al., “Sustainable Energy Generation through Photovoltaic Systems: Current Trends and Future Perspectives,” pp. 1–6, Dec. 2025, doi: 10.1109/MECON67253.2025.11276917.

[2] T. Cang, “Comprehensive Exploration of Solar Photovoltaic Technology: Enhancing Efficiency, Integrating Energy Storage, and Addressing Environmental and Economic Challenges,” Applied and Computational Engineering, vol. 123, no. 1, pp. 10–16, Jan. 2025, doi: 10.54254/2755-2721/2025.19565.

[3] S. Kalogirou, Solar Energy Engineering: Processes and Systems, Elsevier, 2024.

[4] B. Sivadharshini et al., “Solar Energy Forecasting Models for Grid Integration and Power Balancing,” E3S Web of Conferences, vol. 540, p. 10022, Jun. 2024, doi: 10.1051/E3SCONF/202454010022.

[5] C. Wan et al., “Photovoltaic and solar power forecasting for smart grid energy management,” CSEE Journal of Power and Energy Systems, vol. 1, no. 4, pp. 38–46, Jan. 2016, doi: 10.17775/CSEEJPES.2015.00046.

[6] W. A. Beckman, N. Blair, and J. A. Duffie, Solar Engineering of Thermal Processes, Photovoltaics and Wind, 5th ed., Wiley, 2021.

[7] N. Y. Hendrikx et al., “All sky imaging-based short-term solar irradiance forecasting with Long Short-Term Memory networks,” Solar Energy, vol. 272, p. 112463, Apr. 2024, doi: 10.1016/J.SOLENER.2024.112463.

[8] M. Jordan, J. Kleinberg, and B. Schölkopf, “Pattern Recognition and Machine Learning,” Springer, 2006.

[9] A. Nadeem et al., “AI-Driven precision in solar forecasting: Breakthroughs in machine learning and deep learning,” AIMS Geosciences, vol. 10, no. 4, pp. 684–734, 2024, doi: 10.3934/GEOSCI.2024035.

[10] I. Goodfellow, Y. Bengio, and A. Courville, Deep Learning, MIT Press, 2016.

[11] G. James, D. Witten, T. Hastie, and R. Tibshirani, An Introduction to Statistical Learning, Springer, 2021, doi: 10.1007/978-1-0716-1418-1.

[12] R. Shwartz-Ziv and A. Armon, “Tabular Data: Deep Learning is Not All You Need,” Information Fusion, vol. 81, pp. 84–90, Jun. 2022, doi: 10.1016/j.inffus.2021.11.011.

[13] C. F. Mbey et al., “A Novel Deep Learning-Based Data Analysis Model for Solar Photovoltaic Power Generation and Electrical Consumption Forecasting in the Smart Power Grid,” Applied Computational Intelligence and Soft Computing, vol. 2024, 2024, doi: 10.1155/2024/9257508.

[14] T. N. Nguyen and F. Müsgens, “A meta-analysis of solar forecasting based on skill score,” Journal of Renewable and Sustainable Energy, vol. 18, no. 2, Mar. 2026, doi: 10.1063/5.0300682.

[15] P. N. L. M. Radzi et al., “Optimizing solar power forecasting with metaheuristic algorithms and deep learning models for photovoltaic grid connected systems,” Scientific Reports, vol. 15, no. 1, Dec. 2025, doi: 10.1038/s41598-025-23822-1.

[16] A. J. Lari et al., “Using Machine Learning Algorithms to Forecast Solar Energy Power Output,” Electronics, vol. 14, no. 5, Mar. 2025, doi: 10.3390/electronics14050866.

[17] H. Alizadegan et al., “Solar Energy Production Forecasting: A Comparative Study of Bi-LSTM, LSTM, XGBoost Models with Activation Function Analysis,” May 2024. doi: 10.20944/preprints202405.0994.v1.

[18] Y. Xu, X. Ji, and Z. Zhu, “A photovoltaic power forecasting method based on the LSTM-XGBoost-EEDA-SO model,” Scientific Reports, vol. 15, no. 1, Dec. 2025, doi: 10.1038/s41598-025-16368-9.

[19] A. D. Huynh and T. K. Nguyen, “The Comparison of GRU and LSTM in Solar Power Generation Forecasting Application,” International Journal of Science and Research Archive, vol. 13, no. 1, pp. 1360–1370, Sep. 2024, doi: 10.30574/ijsra.2024.13.1.1831.

[20] D. K. Avoussou, A. A. Salami, and D. Barchiesi, “Interval forecasting of solar irradiation using LSTM, XGBoost and bandwidth optimization of KDE by differential evolution,” Sep. 2026, Solar Compass, doi: 10.1016/j.solcom.2026.100171.

[21] F. Aksan et al., “A Comparative Study of Machine Learning Models for PV Energy Prediction in an Energy Community,” Energies, vol. 18, no. 22, Nov. 2025, doi: 10.3390/en18225980.

[22] A. Mollasalehi and A. Farhadi, “Solar and Wind Power Forecasting: A Comparative Review of LSTM, Random Forest, and XGBoost Models,” Sep. 2025. [Online]. Available: http://arxiv.org/abs/2509.24059.

[23] Desert Knowledge Australia, “Desert Knowledge Australia Solar Centre (DKASC) Historical Dataset (2023–2025),” Desert Knowledge Australia Solar Centre.

[24] A. Zhang, Z. C. Lipton, M. Li, and A. J. Smola, Dive into Deep Learning, Cambridge University Press, 2023.

[25] S. Hochreiter and J. Schmidhuber, “Long Short-Term Memory,” Neural Computation, vol. 9, pp. 1735–1780, Jun. 1997, doi: 10.1162/neco.1997.9.8.1735.

[26] F. Chollet, Deep Learning with Python, 2nd ed., Manning, 2021.

[27] J. Chung, C. Gulcehre, K. Cho, and Y. Bengio, “Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling,” Dec. 2014. [Online]. Available: https://arxiv.org/pdf/1412.3555.

[28] T. Hastie, R. Tibshirani, and J. Friedman, The Elements of Statistical Learning, Springer, 2009, doi: 10.1007/978-0-387-84858-7.

[29] L. Breiman, “Random forests,” Machine Learning, vol. 45, no. 1, pp. 5–32, Oct. 2001, doi: 10.1023/A:1010933404324.

[30] J. Han, M. Kamber, and J. Pei, Data Mining: Concepts and Techniques, 3rd ed., Morgan Kaufmann, 2011.

[31] T. Chen and C. Guestrin, “XGBoost: A Scalable Tree Boosting System,” in Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 785–794, 2016, doi: 10.1145/2939672.2939785.

[32] S. M. Lundberg and S.-I. Lee, “A Unified Approach to Interpreting Model Predictions,” in Advances in Neural Information Processing Systems, vol. 30, pp. 4766–4775, 2017.

[33] C. Molnar, Interpretable Machine Learning: A Guide for Making Black Box Models Explainable, 2nd ed., 2022. [Online]. Available: http://leanpub.com/interpretable-machine-learning.