Low-Carbon Scheduling for Power-Hydrogen Integrated Energy System Using Large Language Model Enhanced Deep Reinforcement Learning
IEEE Transactions on Sustainable Energy · 2026-10 · 17 (4) · 3480-3495
DOI: 10.1109/tste.2026.3677475 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Sustainable Energy, 2026-10. Verfasst von Mingde Lin und 2 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 36 Literaturangaben. Gefördert von: National Natural Science Foundation of China.
Abstract (Originalsprache)
Power-hydrogen integrated energy systems (PH-IESs) are emerging as a promising solution to achieve net-zero carbon emissions. However, the large-scale integration of uncertain renewable energy sources (RESs) and flexible demand-side resources (e.g., hydrogen electrolysis) pose great challenges in the real-time low-carbon scheduling of PH-IESs. This paper proposes a novel source-demand coordinated low-carbon scheduling framework for PH-IESs, empowered by deep reinforcement learning (DRL) and large language model (LLM) techniques. The proposed framework systematically integrates source-side generation with demand-side flexible hydrogen loads to improve carbon reduction and economic efficiency. An LLM-enhanced carbon flow pattern recognition and inference mechanism is developed to enable accurate predictions of real-time carbon intensity under complex, high-dimensional operating conditions. An LLM-enhanced DRL algorithm is then proposed to directly generate optimal scheduling strategies for the flexible demand-side resources, which can overcome computational bottlenecks in real-time decision-making. Numerical simulations on two PH-IESs are conducted to demonstrate the effectiveness of integrating DRL and LLM techniques for intelligent, efficient, and low-carbon operation of PH-IESs.
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Lin, M., Wu, T., & Xie, K. (2026). Low-Carbon Scheduling for Power-Hydrogen Integrated Energy System Using Large Language Model Enhanced Deep Reinforcement Learning. IEEE Transactions on Sustainable Energy, 17(4), 3480-3495. https://doi.org/10.1109/tste.2026.3677475
Vancouver
Lin M, Wu T, Xie K. Low-Carbon Scheduling for Power-Hydrogen Integrated Energy System Using Large Language Model Enhanced Deep Reinforcement Learning. IEEE Transactions on Sustainable Energy. 2026;17(4):3480-3495. doi:10.1109/tste.2026.3677475
BibTeX
@article{Lin2026,
author = {Lin, Mingde and Wu, Tao and Xie, Kaigui},
title = {Low-Carbon Scheduling for Power-Hydrogen Integrated Energy System Using Large Language Model Enhanced Deep Reinforcement Learning},
journal = {IEEE Transactions on Sustainable Energy},
year = {2026},
volume = {17},
number = {4},
pages = {3480--3495},
doi = {10.1109/tste.2026.3677475}
}
EUVIMED