Abstract
This paper explores the application of Graph Neural Networks (GNNs) to power flow problems, comparing several spectral and spatial methods. The research reveals that spatial methods generally outperform their spectral counterparts, which do not rely on spectral theory, eigenvalues, or eigenvectors. GraphSAGE [9] demonstrates the best performance among the spatial methods tested, achieving a Mean Absolute Percentage Error (MAPE) of 0.79% on the test set in an experiment with 14-buses and 0.53% in the experiment with 30-buses. These findings suggest that for power flow problems, it is beneficial to consider at least hybrid or predominantly spatial models that leverage information from non-immediate neighbors. This research highlights the potential of spatial GNN methods in accurately capturing the complexities of power systems, paving the way for more robust and efficient solutions in the domain.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
pandapower. https://www.pandapower.org/
PyG documentation - pytorch_geometric documentation. https://pytorch-geometric.readthedocs.io/en/latest/
PyTorch geometric temporal documentation - PyTorch geometric temporal documentation. https://pytorch-geometric-temporal.readthedocs.io/en/latest/
Tinney: Power flow solution by newton’s method - google académico. https://scholar.google.com/scholar_lookup?title=Power%20flow%20solution%20by%20newtons%20method&publication_year=1967 &author=William%C2%A0F.%20Tinney &author=Clifford%C2%A0E.%20Hart
Böttcher, L., et al.: Solving AC power flow with graph neural networks under realistic constraints. In: 2023 IEEE Belgrade PowerTech, pp. 1–7. https://doi.org/10.1109/PowerTech55446.2023.10202246. http://arxiv.org/abs/2204.07000
Donon, B., Donnot, B., Guyon, I., Marot, A.: Graph neural solver for power systems. In: 2019 International Joint Conference on Neural Networks (IJCNN), pp. 1–8. IEEE. https://doi.org/10.1109/IJCNN.2019.8851855. https://ieeexplore.ieee.org/document/8851855/
Fey, M., Lenssen, J.E.: Fast graph representation learning with PyTorch geometric. http://arxiv.org/abs/1903.02428
Gilmer, J., Schoenholz, S.S., Riley, P.F., Vinyals, O., Dahl, G.E.: Neural message passing for quantum chemistry. https://doi.org/10.48550/ARXIV.1704.01212, https://arxiv.org/abs/1704.01212, publisher: arXiv Version Number: 2
Hamilton, W., Ying, Z., Leskovec, J.: Inductive representation learning on large graphs. In: Guyon, I., et al. (eds.) Advances in Neural Information Processing Systems, vol. 30. Curran Associates, Inc
Hammond, D.K., Vandergheynst, P., Gribonval, R.: Wavelets on graphs via spectral graph theory 30(2), 129–150. https://doi.org/10.1016/j.acha.2010.04.005. https://linkinghub.elsevier.com/retrieve/pii/S1063520310000552
Hansen, J.B., Anfinsen, S.N., Bianchi, F.M.: Power flow balancing with decentralized graph neural networks 38(3), 2423–2433. https://doi.org/10.1109/TPWRS.2022.3195301. https://ieeexplore.ieee.org/document/9847037/
Hernandez, C., Sánchez Huertas, W., Gómez, V.: Optimal power flow through artificial intelligence techniques 25(69), 150–170. https://doi.org/10.14483/22487638.18245. https://revistas.udistrital.edu.co/index.php/Tecnura/article/view/18245
Kipf, T.N., Welling, M.: Semi-supervised classification with graph convolutional networks. http://arxiv.org/abs/1609.02907
Liao, W., Bak-Jensen, B., Radhakrishna Pillai, J., Wang, Y., Wang, Y.: A review of graph neural networks and their applications in power systems 10(2), 345–36. https://doi.org/10.35833/MPCE.2021.000058. https://ieeexplore.ieee.org/document/9520300/
Liu, S., Wu, C., Zhu, H.: Topology-aware graph neural networks for learning feasible and adaptive AC-OPF solutions 38(6), 5660–5670. https://doi.org/10.1109/TPWRS.2022.3230555. https://ieeexplore.ieee.org/document/9992121/
Lopez-Garcia, T.B., Domínguez-Navarro, J.A.: Power flow analysis via typed graph neural networks 117, 105567. https://doi.org/10.1016/j.engappai.2022.105567https://linkinghub.elsevier.com/retrieve/pii/S0952197622005577
Marković, M., Bossart, M., Hodge, B.M.: Machine learning for modern power distribution systems: Progress and perspectives 15(3), 032301. https://doi.org/10.1063/5.0147592. https://pubs.aip.org/jrse/article/15/3/032301/2900695/Machine-learning-for-modern-power-distribution
Owerko, D., Gama, F., Ribeiro, A.: Optimal power flow using graph neural networks. In: ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 5930–5934. IEEE. https://doi.org/10.1109/ICASSP40776.2020.9053140. https://ieeexplore.ieee.org/document/9053140/
Rozemberczki, B., et al.: PyTorch geometric temporal: Spatiotemporal signal processing with neural machine learning models. http://arxiv.org/abs/2104.07788
Sereeter, B., Vuik, C., Witteveen, C.: On a comparison of newton-raphson solvers for power flow problems 360, 157–169. https://doi.org/10.1016/j.cam.2019.04.007. https://www.sciencedirect.com/science/article/pii/S0377042719301876
Stock, S., Babazadeh, D., Becker, C.: Applications of artificial intelligence in distribution power system operation 9, 150098–150119. https://doi.org/10.1109/ACCESS.2021.3125102. https://ieeexplore.ieee.org/document/9599712/
Thurner, L., et al.: Pandapower-an open-source python tool for convenient modeling, analysis, and optimization of electric power systems 33(6), 6510–652. https://doi.org/10.1109/TPWRS.2018.2829021. https://ieeexplore.ieee.org/document/8344496/
Tuo, M., Li, X., Zhao, T.: Graph neural network-based power flow model
Veličković, P., Cucurull, G., Casanova, A., Romero, A., Liò, P., Bengio, Y.: Graph attention networks. https://doi.org/10.48550/ARXIV.1710.10903. https://arxiv.org/abs/1710.10903, publisher: arXiv Version Number: 3
Wu, Z., Pan, S., Chen, F., Long, G., Zhang, C., Yu, P.S.: A comprehensive survey on graph neural networks 32(1), 4–24. https://doi.org/10.1109/TNNLS.2020.2978386. http://arxiv.org/abs/1901.00596
Yaniv, A., Kumar, P., Beck, Y.: Towards adoption of GNNs for power flow applications in distribution systems 216, 109005. https://doi.org/10.1016/j.epsr.2022.109005. https://linkinghub.elsevier.com/retrieve/pii/S0378779622010549
Zhou, J., et al.: Graph neural networks: a review of methods and applications 1, 57–8. https://doi.org/10.1016/j.aiopen.2021.01.001. https://linkinghub.elsevier.com/retrieve/pii/S2666651021000012
Acknowledgments
G.A.R. thanks ANID FONDECYT 1230315 and ANID PIA/BASAL FB0002. P.A.E. thanks Enel Distribucion Chile for the support provided in the development of this research.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2025 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Espinoza, P.A., Ruz, G.A. (2025). Comparative Analysis of Spatial and Spectral Methods in GNN for Power Flow in Electrical Power Systems. In: Hernández-García, R., Barrientos, R.J., Velastin, S.A. (eds) Progress in Pattern Recognition, Image Analysis, Computer Vision, and Applications. CIARP 2024. Lecture Notes in Computer Science, vol 15369. Springer, Cham. https://doi.org/10.1007/978-3-031-76604-6_2
Download citation
DOI: https://doi.org/10.1007/978-3-031-76604-6_2
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-76603-9
Online ISBN: 978-3-031-76604-6
eBook Packages: Computer ScienceComputer Science (R0)
