EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Dynamic Load Balancing for Renewable Integrated Power Systems via Hierarchical Reinforcement Learning

Jiacheng Huang

American Journal Of Big Data · 31.01.2026 · 7 (1) · 127-141

DOI: 10.71465/ajbd3587 ↗

Kurz gefasst

Fachartikel in der Zeitschrift American Journal Of Big Data, online erschienen am 31.01.2026. Verfasst von Jiacheng Huang. Die Arbeit nennt 28 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The accelerating penetration of renewable energy sources (RES) into modern electricity networks presents substantial challenges in maintaining stable power balancing, frequency regulation, and load dispatch efficiency. This paper introduces a hierarchical reinforcement learning (HRL) framework designed to achieve dynamic load balancing (DLB) in renewable integrated power systems (RIPS). The proposed two-tier architecture consists of a macro-level scheduling agent that determines long-horizon dispatch priorities and a micro-level execution agent that performs real-time active power allocation and frequency correction. Bothagents are co-trained within a Markov decision process (MDP) environment built upon a modified IEEE 33-bus test network incorporating photovoltaic (PV) generation, wind turbine (WT) arrays, battery energy storage systems (BESS), and temporally variable load profiles. Comparative experiments against single agent deep Q-network (DQN) and model predictive control (MPC) baselines demonstrate that the proposed framework reduces frequency deviation by 31.4%, improves renewable curtailment efficiency by 18.7%, and yields superior voltage stability metrics under high RES penetration scenarios. The results suggest that hierarchically decomposed multi-timescale decision architectures offer a scalable and adaptive pathway for intelligent management of next generation power grids.

Zitieren

APA

Huang, J. (2026). Dynamic Load Balancing for Renewable Integrated Power Systems via Hierarchical Reinforcement Learning. American Journal Of Big Data, 7(1), 127-141. https://doi.org/10.71465/ajbd3587

Vancouver

Huang J. Dynamic Load Balancing for Renewable Integrated Power Systems via Hierarchical Reinforcement Learning. American Journal Of Big Data. 2026;7(1):127-141. doi:10.71465/ajbd3587

BibTeX
@article{Huang2026,
  author = {Huang, Jiacheng},
  title = {Dynamic Load Balancing for Renewable Integrated Power Systems via Hierarchical Reinforcement Learning},
  journal = {American Journal Of Big Data},
  year = {2026},
  volume = {7},
  number = {1},
  pages = {127--141},
  doi = {10.71465/ajbd3587}
}
Kostenlos starten →