Optimal scheduling of DC microgrids integrating deep learning irradiance forecasting and battery degradation dynamics
Renewable Energies · 2026-07 · 4 (2)
DOI: 10.1177/27533735261507771 ↗Kurz gefasst
Fachartikel in der Zeitschrift Renewable Energies, online erschienen am 17.11.2026. Verfasst von Gabriel Narvaez und 5 weiteren Autor:innen. Die Arbeit nennt 44 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Abstract (Originalsprache)
Efficient operational scheduling of battery energy storage systems is a critical requirement for stabilizing DC microgrids heavily reliant on intermittent renewable sources. However, existing energy management strategies often address renewable uncertainty and battery longevity as separate optimization challenges, frequently relying on simplified degradation models or restrictive state-of-charge limits. This paper proposes a holistic methodology that simultaneously confronts solar variability and technical battery health by integrating Long Short-Term Memory neural networks with an innovative degradation cost formulation. Unlike conventional models that impose hard State of Charge constraints (typically 20–80%), our formulation utilizes manufacturer-specific data to correlate energy delivery with depth of discharge, permitting full 0–100% battery utilization while economically penalizing deep discharge states. Validated using real-world photovoltaic and load data from the ADREAM building (Laboratory for Analysis and Architecture of Systems - National Centre for Scientific Research), the integrated approach demonstrates superior performance over standard probabilistic Beta Probability Density Function models. The results indicate a 3.5% reduction in total operational costs compared to traditional cycle-count methods, proving that coupling predictive machine learning with realistic degradation physics provides a more flexible and cost-effective management framework for modern microgrid operators.
Zitieren
APA
Narvaez, G., Giraldo, L. F., Bressan, M., Pantoja, A., Jammes, B., & Alonso, C. (2026). Optimal scheduling of DC microgrids integrating deep learning irradiance forecasting and battery degradation dynamics. Renewable Energies, 4(2). https://doi.org/10.1177/27533735261507771
Vancouver
Narvaez G, Giraldo LF, Bressan M, Pantoja A, Jammes B, Alonso C. Optimal scheduling of DC microgrids integrating deep learning irradiance forecasting and battery degradation dynamics. Renewable Energies. 2026;4(2). doi:10.1177/27533735261507771
BibTeX
@article{Narvaez2026,
author = {Narvaez, Gabriel and Giraldo, Luis Felipe and Bressan, Michael and Pantoja, Andres and Jammes, Bruno and Alonso, Corinne},
title = {Optimal scheduling of DC microgrids integrating deep learning irradiance forecasting and battery degradation dynamics},
journal = {Renewable Energies},
year = {2026},
volume = {4},
number = {2},
doi = {10.1177/27533735261507771}
}
EUVIMED