EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Bridging Swing‐Equation Dynamics and Deep Neural Networks for Cyber‐Physical Anomaly Differentiation in Power Systems

HyungSeon Oh, Daniel F. Opila

Journal of Control Science and Engineering · 2026-01 · 2026 (1)

DOI: 10.1155/jcse/7161510 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Control Science and Engineering, online erschienen am 29.06.2026. Verfasst von HyungSeon Oh und 1 weiteren Autor:innen. Die Arbeit nennt 44 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Office of Naval Research.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This paper presents an advanced method for detecting physical anomalies in power systems using convolutional neural networks (CNNs) applied to time‐series voltage measurements and their corresponding finite Fourier transforms (FFTs). The approach utilizes data generated by phasor measurement units (PMUs) in real time, focusing on the detection of anomalies caused by contingencies, such as faults, generator outages, or power line trips. We derive an analytical solution to the swing equations that models the behavior of voltages during disturbances and illustrate how these dynamics are represented in both the time and frequency domains. By combining voltage time‐series data with FFTs in a structured format, we train a CNN model to distinguish between normal operation and various types of physical anomalies. The model captures spatial and temporal patterns in the data and utilizes the rank values of the voltage and FFT matrices as features for classification. The results show that the CNN model is effective at detecting anomalies in power systems, even in the presence of high‐frequency data and complex system dynamics. This methodology provides a robust tool for real‐time monitoring and can be used to prevent cascading failures or large‐scale blackouts in power grids.

Förderung

  • Office of Naval Research

Zitieren

APA

Oh, H., & Opila, D. F. (2026). Bridging Swing‐Equation Dynamics and Deep Neural Networks for Cyber‐Physical Anomaly Differentiation in Power Systems. Journal of Control Science and Engineering, 2026(1). https://doi.org/10.1155/jcse/7161510

Vancouver

Oh H, Opila DF. Bridging Swing‐Equation Dynamics and Deep Neural Networks for Cyber‐Physical Anomaly Differentiation in Power Systems. Journal of Control Science and Engineering. 2026;2026(1). doi:10.1155/jcse/7161510

BibTeX
@article{Oh2026,
  author = {Oh, HyungSeon and Opila, Daniel F.},
  title = {Bridging Swing‐Equation Dynamics and Deep Neural Networks for Cyber‐Physical Anomaly Differentiation in Power Systems},
  journal = {Journal of Control Science and Engineering},
  year = {2026},
  volume = {2026},
  number = {1},
  doi = {10.1155/jcse/7161510}
}
Kostenlos starten →