Publications

• Sorted by Date • Classified by Publication Type • Classified by Research Category •

Physics-Informed Reinforcement Learning for Large-Scale EV Smart Charging Considering Distribution Network Voltage Constraints

Stavros Orfanoudakis, Frans A. Oliehoek, Peter Palensky, and Pedro P. Vergara. Physics-Informed Reinforcement Learning for Large-Scale EV Smart Charging Considering Distribution Network Voltage Constraints. Applied Energy, 421:128224, 2026.

Download

pdf [933.1kB]  ps.gz ps HTML 

Abstract

Electric Vehicles (EVs) offer substantial flexibility for grid services, yet large-scale, uncoordinated charging can threaten voltage stability in distribution networks. Existing Reinforcement Learning (RL) approaches for smart charging often disregard physical grid constraints or have degraded performance for complex, large-scale tasks, limiting their scalability and real-world applicability. This paper introduces a physics-informed (PI) RL algorithm that integrates a differentiable power flow model and voltage-based reward design into the Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithm, enabling RL to better scale to a large number of EVs and deliver real-time voltage support while meeting EV user demands. The resulting PI-TD3 algorithm achieves faster convergence, improved sample efficiency, and reliable voltage magnitude regulation under uncertain and overloaded conditions. Benchmarks on the IEEE 34-bus and 123-bus networks show that the proposed PI-TD3 outperforms both model-free RL and optimization-based baselines in grid constraint management, user satisfaction, and economic metrics, even as the system scales to hundreds of EVs. These advances enable robust, scalable, and practical EV charging strategies that enhance grid resilience and support the operation of distribution networks.

BibTeX Entry

@ARTICLE{Orfanoudakis26APEN,
    author =    {Stavros Orfanoudakis and Frans A. Oliehoek and Peter Palensky 
                 and Pedro P. Vergara},
    title =     {Physics-Informed Reinforcement Learning for Large-Scale {EV} 
                 Smart Charging Considering Distribution Network Voltage Constraints},
    journal =   {Applied Energy},
    volume =    {421},
    pages =     {128224},
    year =      {2026},
    issn =      {0306-2619},
    doi =       {https://doi.org/10.1016/j.apenergy.2026.128224},
    url =       {https://www.sciencedirect.com/science/article/pii/S0306261926008780},
    keywords =   {refereed},
    abstract = {Electric Vehicles (EVs) offer substantial flexibility for grid services, yet large-scale, uncoordinated charging can threaten voltage stability in distribution networks. Existing Reinforcement Learning (RL) approaches for smart charging often disregard physical grid constraints or have degraded performance for complex, large-scale tasks, limiting their scalability and real-world applicability. This paper introduces a physics-informed (PI) RL algorithm that integrates a differentiable power flow model and voltage-based reward design into the Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithm, enabling RL to better scale to a large number of EVs and deliver real-time voltage support while meeting EV user demands. The resulting PI-TD3 algorithm achieves faster convergence, improved sample efficiency, and reliable voltage magnitude regulation under uncertain and overloaded conditions. Benchmarks on the IEEE 34-bus and 123-bus networks show that the proposed PI-TD3 outperforms both model-free RL and optimization-based baselines in grid constraint management, user satisfaction, and economic metrics, even as the system scales to hundreds of EVs. These advances enable robust, scalable, and practical EV charging strategies that enhance grid resilience and support the operation of distribution networks.}
}

Generated by bib2html.pl (written by Patrick Riley) on Wed Jun 17, 2026 13:52:25 UTC