A Standardized Framework for Machine Learning in Power System Protection

Oelhaf J, Kordowich G, Perez Toro PA, Bergler C, Jäger J, Maier A, Bayer S (2026)

Publication Language: English

Publication Type: Journal article

Publication year: 2026

Journal

Original Authors: Julian Oelhaf, Georg Kordowich, Paula Andrea Pérez-Toro, Christian Bergler, Johann Jäger, Andreas Maier, Siming Bayer

Book Volume: 181

Article Number: 112169

URI: https://www.sciencedirect.com/science/article/pii/S0142061526006113

DOI: 10.1016/j.ijepes.2026.112169

Open Access Link: https://www.sciencedirect.com/science/article/pii/S0142061526006113

Abstract

Studies of machine-learning-based power-system protection increasingly report near-perfect scores, yet the meaning of those scores depends strongly on the evaluation setting. Protection task, physical scope, measurements, timing, targets, preprocessing, and validation often vary jointly and remain incompletely specified. This paper proposes a standardization-oriented framework that treats evaluation design as part of the scientific contribution. It defines seven required study dimensions: protection objective, physical system scope, observability and measurements, timing and decision windows, targets and valid samples, training and validation protocol, and evaluation outputs. The framework is instantiated in a bounded case study on the public PROTECT-90 electromagnetic-transient benchmark, comprising 9022 simulated episodes from one 90 kV double-line topology for onset-conditioned fault classification and fault localization. Under centralized sensing, simulation-metadata-aligned 20 ms windows, and episode-grouped validation, a multi-layer perceptron (MLP) achieved a five-fold mean macro-averaged F1 score of 0.991 ± 0.001 for classification and a localization mean absolute error of 10.20 ± 0.25 % of line length, where the standard deviations describe variation across the episode-grouped folds. Extending the decision horizon to 50 ms preserved this task-dependent performance asymmetry, while reduced observability approximately doubled the MLP localization error but had little effect on classification. A synchronized two-ended conventional locator outperformed the learning locators under its richer clean information set, and measurement degradation showed that clean predictive performance did not determine robustness. The framework turns evaluation assumptions into explicit and reproducible evidence and provides a basis for more comparable, auditable research evaluation and future certification-oriented assessment of machine-learning protection functions.

Authors with CRIS profile

Related research project(s)

Involved external institutions

Ostbayerische Technische Hochschule (OTH) Amberg-Weiden Germany (DE)

How to cite

APA:

Oelhaf, J., Kordowich, G., Perez Toro, P.A., Bergler, C., Jäger, J., Maier, A., & Bayer, S. (2026). A Standardized Framework for Machine Learning in Power System Protection. International Journal of Electrical Power & Energy Systems, 181. https://doi.org/10.1016/j.ijepes.2026.112169

MLA:

Oelhaf, Julian, et al. "A Standardized Framework for Machine Learning in Power System Protection." International Journal of Electrical Power & Energy Systems 181 (2026).

BibTeX: Download