Short-term solar energy forecasting using machine learning models: A comparative study

Alex Ricardo Guamán Andrade, Diego Alejandro García Saraguro, Julio Francisco Guallo Paca, Johanna Gabriela del Pozo Naranjo

Resumen


The increasing deployment of renewable energy systems within decentralized energy frameworks, including microgrids, off-grid installations, and smart residential environments, demands accurate short-term solar power forecasts to ensure both system reliability and operational efficiency. This research presents a comparative analysis of four machine learning algorithms for forecasting: Random Forest, Extreme Gradient Boosting, Support Vector Regression, and Autoregressive Integrated Moving Average (ARIMA). Using solar generation data collected at 15-minute intervals over a three-day period, the effectiveness of the models was evaluated based on the Root Mean Square Error (RMSE), Coefficient of Determination (R²), and visual correspondence with actual generation patterns. The results of this research reveal that ensemble-based machine learning methodologies, specifically RF and XGBoost, consistently outperform both conventional statistical methods such as ARIMA and kernel-based techniques such as SVR. Among them, RF achieved the lowest RMSE and the highest R², indicating exceptional accuracy and generalization capabilities. These results highlight the effectiveness of ensemble approaches in capturing nonlinear dynamics and temporal dependencies inherent in solar power generation, supporting their use in real-time forecasting applications relevant to distributed renewable energy systems.

Palabras clave


Short-range forecasting; Solar power prediction; Random Forest, XGBoost; Support Vector Regression; ARIMA.

Texto completo:

PDF HTML

Referencias


Kocakusak, D., Senick, J., & Andrews, C. J. (2023). Implementing the energy transition: lessons from New Jersey’s residential solar industry.Climate Policy. https://doi.org/10.1080/14693062.2023.2202208

Shi, G., & Eftekharnejad, S. (2016, September 1). Impact of solar forecasting on power system planning.North American Power Symposium. https://doi.org/10.1109/NAPS.2016.7747909

Lin, Y., Duan, D., Hong, X., Cheng, X., Yang, L., & Cui, S. (2020, May 29).Very-Short-Term Solar Forecasting with Long Short-Term Memory (LSTM) Network. https://doi.org/10.1109/AEEES48850.2020.9121512

Guimarães, T., Costa, L. M., Leite, H., & Azevedo, L. F. (2020, September 1).A Hybrid Approach to Load Forecast at a Micro Grid level through Machine Learning algorithms. https://doi.org/10.1109/SEST48500.2020.9203308

Solano, E. S., & Affonso, C. M. (2023). Solar Irradiation Forecasting Using Ensemble Voting Based on Machine Learning Algorithms.Sustainability. https://doi.org/10.3390/su15107943

Liu, F., Qiao, R., Shen, C., & Luo, L. (2017).Designing ensemble learning algorithms using kernel methods. https://doi.org/10.1504/IJMISSP.2017.10009116

Chodakowska, E., Nazarko, J., Nazarko, Ł., Rabayah, H., Abendeh, R., & Alawneh, R. (2023). ARIMA Models in Solar Radiation Forecasting in Different Geographic Locations.Energies. https://doi.org/10.3390/en16135029

Ziyabari, S., Du, L., & Biswas, S. (2020, June 14). A Spatio-temporal Hybrid Deep Learning Architecture for Short-term Solar Irradiance Forecasting.Photovoltaic Specialists Conference. https://doi.org/10.1109/PVSC45281.2020.9300789

Wu, S.-F., Chang, C.-Y., & Lee, S.-J. (2015, March 2).Time series forecasting with missing values. https://doi.org/10.4108/ICST.INISCOM.2015.258269

Bourchtein, A., & Bourchtein, L. (2008).A Splitting Scheme for Large-Scale Atmosphere Dynamics Models. https://doi.org/10.1007/978-3-540-69848-7_51

Basavarajaiah, D. M., & Narasimha Murthy, B. (2020).Random Forest and Concept of Decision Tree Model. https://doi.org/10.1007/978-981-15-8210-3_3

Khan, M., & Srivastava, K. (2020, January 17). Regression Model for Better Generalization and Regression Analysis.International Conference on Machine Learning. https://doi.org/10.1145/3380688.3380691

Bentéjac, C., Csörgo, A., & Martínez-Muñoz, G. (2019). A Comparative Analysis of XGBoost.arXiv: Learning. https://doi.org/10.1007/S10462-020-09896-5

Huang, H., Zheng, Z., Xiao, J., Xu, G., & Wang, J. (2019).XGBoost disease probability predicting method, system and storage medium.

Tiao, G. C. (2001).Time Series: ARIMA Methods. https://doi.org/10.1016/B978-0-08-097086-8.42182-3

Alsharif, M. H., Younes, M. K., & Kim, J. (2019). Time Series ARIMA Model for Prediction of Daily and Monthly Average Global Solar Radiation: The Case Study of Seoul, South Korea.Symmetry. https://doi.org/10.3390/SYM11020240

WU, T. (2006). SVM Based Algorithm for Regressive Modeling with Accurate Data Input-interval Number Output.Control and Decision.

Bisoi, R., Dash, D. R., Dash, P. K., & Tripathy, L. N. (2022). An efficient robust optimized functional link broad learning system for solar irradiance prediction.Applied Energy. https://doi.org/10.1016/j.apenergy.2022.119277

Willmott, C. J., & Matsuura, K. (2005). Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance.Climate Research. https://doi.org/10.3354/CR030079

Narmadha, J., Kaavya, G. N., & Preethii, S. D. (2017, April 1). Analysis on electricity generation forecasting system.International Conference on Electric Information and Control Engineering. https://doi.org/10.1109/ICEICE.2017.8191901

Liebermann, S., Um, J.-S., Hwang, Y., & Schlüter, S. (2021). Performance Evaluation of Neural Network-Based Short-Term Solar Irradiation Forecasts.Energies. https://doi.org/10.3390/EN14113030




DOI: https://doi.org/10.23857/pc.v10i4.9440

Enlaces de Referencia

  • Por el momento, no existen enlaces de referencia
';

Bookmark and Share




Polo del Conocimiento              

Revista Científico-Académica Multidisciplinaria

ISSN: 2550-682X

Casa Editora del Polo                                                 

Manta - Ecuador       

Dirección: Ciudadela El Palmar, II Etapa,  Manta - Manabí - Ecuador.

Código Postal: 130801

Teléfonos: 056051775/0991871420

Email: polodelconocimientorevista@gmail.com / director@polodelconocimiento.com

URL: https://www.polodelconocimiento.com/