EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Chi-Squared Detection of Physical Accident Signatures in Nuclear Power Plant Data, with Synthetic Adversarial Stress Test

Mordecai Opoku Ohemeng iD, Frederick T Sheldon iD

Sensors · 28.10.2026 · 26 (21) · 6812

DOI: 10.3390/s26216812 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 28.10.2026. Verfasst von Mordecai Opoku Ohemeng und 1 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science, University of Idaho, Moscow, ID 83844, USA. Hauptthemen laut MEDLINE: Nuclear Power Plants, Radioactive Hazard Release. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Nuclear-Renewable Hybrid Energy Systems (N-RHES) integrate small modular reactors with intermittent renewable generation to deliver flexible, zero-carbon grid balancing. However, their interconnected cyber-physical architecture remains susceptible to adversarial manipulation of digital instrumentation and control pathways. Advanced consequence-driven methodologies, such as Consequence-driven Cyber-Informed Engineering (CCE), prioritize engineering out high-consequence cyber risks rather than relying solely on post hoc detection. This paper presents a benchmark of lightweight detectors motivated by the CCE analytical workflow, instantiated as an anomaly detection pipeline, evaluated on the Nuclear Power Plant Accident Data (NPPAD) dataset. We compare a classical chi-squared (Mahalanobis distance) detector against an Isolation Forest detector, alongside per-feature z-score and PCA reconstruction baselines. All detectors are evaluated at a matched 1% target false alarm rate (FAR) on a random split of the 302 normal samples, yielding an achieved FAR of 0.98%. Chi-squared achieves an AUC of 0.9989 (95% CI [0.9975, 0.9997]) and a 96.61% sample-level detection rate, significantly outperforming Isolation Forest (AUC 0.9762, 81.79% detection rate; DeLong p

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Nuclear Power PlantsRadioactive Hazard ReleaseAlgorithms

Stichwörter der Autor:innen

Isolation ForestNPPAD benchmarkadversarial stress testanomaly detectionchi-squared residualcyber-informed engineeringnuclear power plant accident diagnosissynthetic false-data injection

Autor:innen und Institutionen

  • Department of Computer Science, University of Idaho, Moscow, ID 83844, USA.

ORCID-Profile: Mordecai Opoku Ohemeng, Frederick T Sheldon

Zitieren

APA

Ohemeng, M. O., & Sheldon, F. T. (2026). Chi-Squared Detection of Physical Accident Signatures in Nuclear Power Plant Data, with Synthetic Adversarial Stress Test. Sensors, 26(21), 6812. https://doi.org/10.3390/s26216812

Vancouver

Ohemeng MO, Sheldon FT. Chi-Squared Detection of Physical Accident Signatures in Nuclear Power Plant Data, with Synthetic Adversarial Stress Test. Sensors. 2026;26(21):6812. doi:10.3390/s26216812

BibTeX
@article{Ohemeng2026,
  author = {Ohemeng, Mordecai Opoku and Sheldon, Frederick T},
  title = {Chi-Squared Detection of Physical Accident Signatures in Nuclear Power Plant Data, with Synthetic Adversarial Stress Test},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {21},
  pages = {6812},
  doi = {10.3390/s26216812}
}
Kostenlos starten →