Evaluación del desempeño de modelos LSTM y XGBoost en la predicción de la demanda eléctrica del sistema ecuatoriano

Wilson Brito; Wilson Sánchez

doi:10.37116/revistaenergia.v22.n2.2026.729

Performance Evaluation of LSTM and XGBoost Models for Electric Demand Forecasting in the Ecuadorian Power System

Evaluación del desempeño de modelos LSTM y XGBoost en la predicción de la demanda eléctrica del sistema ecuatoriano

Published 2026-01-30

https://doi.org/10.37116/revistaenergia.v22.n2.2026.729

Open | Download

PDF (Spanish) FLIP HTML (Spanish) XML (Spanish) EPUB (Spanish)

Issue

Vol. 22 No. 2 (2026): Revista Técnica "energía", Edición No. 22 ISSUE II

Section

SISTEMAS ELÉCTRICOS DE POTENCIA

How to Cite

Performance Evaluation of LSTM and XGBoost Models for Electric Demand Forecasting in the Ecuadorian Power System. (2026). Revista Técnica "energía", 22(2), PP. 24-31. https://doi.org/10.37116/revistaenergia.v22.n2.2026.729

DOI

https://doi.org/10.37116/revistaenergia.v22.n2.2026.729

Dimensions

PlumX

How to Cite

Performance Evaluation of LSTM and XGBoost Models for Electric Demand Forecasting in the Ecuadorian Power System. (2026). Revista Técnica "energía", 22(2), PP. 24-31. https://doi.org/10.37116/revistaenergia.v22.n2.2026.729

Download Citation

license

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

Wilson Brito

Operador Nacional de Electricidad - CENACE

Perfil Google Scholar

https://orcid.org/0009-0004-3002-2427

Wilson Sánchez

CELEC EP, Transelectric

Perfil Google Scholar

Show authors biography

Most read articles by the same author(s)

Carlos Lozada, David Panchi, Wilson Sánchez, Andrés Jacho, Linear Regression for the Identification of the Maximum Power Point in Hybrid Microgrids Implemented in HYPERSIM , Revista Técnica "energía": Vol. 20 No. 2 (2024): Revista Técnica "energía", Edición No. 20, ISSUE II
Andrés Jacho, Diego Echeverría, Santiago Chamba, Carlos Lozada, Wilson Sánchez, Application of Grid Forming Control in Energy Storage Systems for Primary Frequency Regulation Case Study: Galapagos Islands , Revista Técnica "energía": Vol. 21 No. 1 (2024): Revista Técnica "energía", Edición No. 21, ISSUE I
Wilson Sánchez, Diego Echeverría, Santiago Chamba, Andrés Jacho, Carlos Lozada, Economic Energy Dispatch of the Micro-Grid in the Galapagos Islands Using the Simsee Platform , Revista Técnica "energía": Vol. 21 No. 1 (2024): Revista Técnica "energía", Edición No. 21, ISSUE I
Wilson Brito, Santiago Chamba, Diego Echeverría, Aharon De La Torre, David Panchi, Parameter Identification, Validation and Tunning of Speed Regulator Tool Using Heuristic Optimization Algorithms , Revista Técnica "energía": Vol. 20 No. 2 (2024): Revista Técnica "energía", Edición No. 20, ISSUE II

LSTM

XGBoost

short-term load forecasting

Temperature

Ecuador

CENACE

Accurate short-term electricity demand forecasting is essential for the technical and economic operation of the Ecuadorian power system. This paper presents a comparison between Long Short-Term Memory (LSTM) neural networks and the XGBoost algorithm for short-term load forecasting, incorporating exogenous variables such as apparent temperature and national holidays. Hourly demand data were obtained from the CENACE database starting in 2021, and meteorological data were sourced from the Open-Meteo satellite platform. A recursive single-step forecasting strategy was implemented for a 24-hour prediction horizon. Results show that the LSTM model achieved the highest accuracy, significantly outperforming XGBoost. The study concludes that incorporating exogenous variables improves forecasting performance and that LSTM provides a reliable approach for short-term load prediction to support national power system planning.

Article visits 10 | PDF visits 1

Downloads

Download data is not yet available.

A. Kumar Dubey, A. Kumar, V. García-Díaz, A. Kumar Sharma, y K. Kanhaiya, «Study and analysis of SARIMA and LSTM in forecasting time series data», Sustain. Energy Technol. Assess., vol. 47, p. 101474, oct. 2021, doi: 10.1016/j.seta.2021.101474.
H. Abbasimehr, M. Shabani, y M. Yousefi, «An optimized model using LSTM network for demand forecasting», Comput. Ind. Eng., vol. 143, p. 106435, may 2020, doi: 10.1016/j.cie.2020.106435.
M. Tan, S. Yuan, S. Li, Y. Su, H. Li, y F. He, «Ultra-Short-Term Industrial Power Demand Forecasting Using LSTM Based Hybrid Ensemble Learning», IEEE Trans. Power Syst., vol. 35, n.o 4, pp. 2937-2948, jul. 2020, doi: 10.1109/TPWRS.2019.2963109.
B. ul Islam y S. F. Ahmed, «Short-Term Electrical Load Demand Forecasting Based on LSTM and RNN Deep Neural Networks», Math. Probl. Eng., vol. 2022, n.o 1, p. 2316474, 2022, doi: https://doi.org/10.1155/2022/2316474.
M. Mohamed, F. E. Mahmood, M. A. Abd, M. Rezkallah, A. Hamadi, y A. Chandra, «Load Demand Forecasting Using eXtreme Gradient Boosting (XGboost)», en 2023 IEEE Industry Applications Society Annual Meeting (IAS), oct. 2023, pp. 1-7. doi: 10.1109/IAS54024.2023.10406613.
T. Mazibuko y K. Akindeji, «Hybrid Forecasting for Energy Consumption in South Africa: LSTM and XGBoost Approach», Energies, vol. 18, n.o 16, p. 4285, ene. 2025, doi: 10.3390/en18164285.
Ministerio de Turismo de Ecuador, «Calendario Oficial de Feriados Nacionales 2023-2025». diciembre de 2022. [En línea]. Disponible en: https://www.turismo.gob.ec/wp-content/uploads/2023/12/CALENDARIO-FERIADOS-2023-2025-06-12-2022-.pdf
R. G. Steadman, «A Universal Scale of Apparent Temperature», J. Appl. Meteorol. Climatol., vol. 23, n.o 12, pp. 1674-1687, dic. 1984, doi: 10.1175/1520-0450(1984)023%3C1674:AUSOAT%3E2.0.CO;2.
J. Kang y D. M. Reiner, «What is the effect of weather on household electricity consumption? Empirical evidence from Ireland», Energy Econ., vol. 111, p. 106023, jul. 2022, doi: 10.1016/j.eneco.2022.106023.
S. Ozdemir, Feature Engineering Bookcamp. Simon and Schuster, 2022.
E. Lewinson, Python for Finance Cookbook: Over 80 powerful recipes for effective financial data analysis. Packt Publishing Ltd, 2022.
X. Vasques, Machine Learning Theory and Applications: Hands-on Use Cases with Python on Classical and Quantum Machines. John Wiley & Sons, 2024.
V. Babuskhin y A. Kravchenko, Machine Learning System Design: With end-to-end examples. Simon and Schuster, 2025.
«Writing a training loop from scratch | TensorFlow Core», TensorFlow. Accedido: 4 de noviembre de 2025. [En línea]. Disponible en: https://www.tensorflow.org/guide/keras/writing_a_training_loop_from_scratch
«XGBoost Documentation — xgboost 3.1.1 documentation». Accedido: 4 de noviembre de 2025. [En línea]. Disponible en: https://xgboost.readthedocs.io/en/stable/index.html
A. van Wyk, Machine Learning with LightGBM and Python: A practitioner’s guide to developing production-ready machine learning systems. Packt Publishing Ltd, 2023.
T. Chen y C. Guestrin, «XGBoost: A Scalable Tree Boosting System», en Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, en KDD ’16. New York, NY, USA: Association for Computing Machinery, ago. 2016, pp. 785-794. doi: 10.1145/2939672.2939785.
S. S. M Paulraj, Introduction to Artificial Neural Networks. Vikas Publishing House, 2009.
R. Rojas, Neural Networks: A Systematic Introduction. Springer Science & Business Media, 2013.
S. Hochreiter y J. Schmidhuber, «Long Short-Term Memory», Neural Comput., vol. 9, n.o 8, pp. 1735-1780, nov. 1997, doi: 10.1162/neco.1997.9.8.1735.
J. Brownlee, Deep Learning for Time Series Forecasting: Predict the Future with MLPs, CNNs and LSTMs in Python. Machine Learning Mastery, 2018.
R. A. Yaffee y M. McGee, An Introduction to Time Series Analysis and Forecasting: With Applications of SAS® and SPSS®. Academic Press, 2000.
P. Cichosz, Data Mining Algorithms: Explained Using R. John Wiley & Sons, 2015.
I. Gridin, Time Series Forecasting using Deep Learning: Combining PyTorch, RNN, TCN, and Deep Neural Network Models to Provide Production-Ready Prediction Solutions (English Edition). BPB Publications, 2021.