Proximal Policy Optimization with Graph Neural Networks for Optimal Power Flow

Ángela López-Cardona, Guillermo Bernárdez, Pere Barlet-Rose, Pere Barlet-Rose, Albert Cabellos-Aparicio, Albert Cabellos-Aparicio

2025

Abstract

Optimal Power Flow (OPF) is a key research area within the power systems field that seeks the optimal operation point of electric power plants, and which needs to be solved every few minutes in real-world scenarios. However, due to the non-convex nature of power generation systems, there is not yet a fast, robust solution for the full Alternating Current Optimal Power Flow (ACOPF). In the last decades, power grids have evolved into a typical dynamic, non-linear and large-scale control system —known as the power system—, so searching for better and faster ACOPF solutions is becoming crucial. The appearance of Graph Neural Networks (GNN) has allowed the use of Machine Learning (ML) algorithms on graph data, such as power networks. On the other hand, Deep Reinforcement Learning (DRL) is known for its proven ability to solve complex decision-making problems. Although solutions that use these two methods separately are beginning to appear in the literature, none has yet combined the advantages of both. We propose a novel architecture based on the Proximal Policy Optimization (PPO) algorithm with Graph Neural Networks to solve the Optimal Power Flow. The objective is to design an architecture that learns how to solve the optimization problem and, at the same time, is able to generalize to unseen scenarios. We compare our solution with the Direct Current Optimal Power Flow approximation (DCOPF) in terms of cost. We first trained our DRL agent on the IEEE 30 bus system and with it, we computed the OPF on that base network with topology changes.

Paper Citation

in Harvard Style

López-Cardona Á., Bernárdez G., Barlet-Rose P. and Cabellos-Aparicio A. (2025). Proximal Policy Optimization with Graph Neural Networks for Optimal Power Flow. In Proceedings of the 14th International Conference on Data Science, Technology and Applications - Volume 1: DATA; ISBN 978-989-758-758-0, SciTePress, pages 347-354. DOI: 10.5220/0013462700003967

in Bibtex Style

@conference{data25,
author={Ángela López-Cardona and Guillermo Bernárdez and Pere Barlet-Rose and Albert Cabellos-Aparicio},
title={Proximal Policy Optimization with Graph Neural Networks for Optimal Power Flow},
booktitle={Proceedings of the 14th International Conference on Data Science, Technology and Applications - Volume 1: DATA},
year={2025},
pages={347-354},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013462700003967},
isbn={978-989-758-758-0},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 14th International Conference on Data Science, Technology and Applications - Volume 1: DATA
TI - Proximal Policy Optimization with Graph Neural Networks for Optimal Power Flow
SN - 978-989-758-758-0
AU - López-Cardona Á.
AU - Bernárdez G.
AU - Barlet-Rose P.
AU - Cabellos-Aparicio A.
PY - 2025
SP - 347
EP - 354
DO - 10.5220/0013462700003967
PB - SciTePress