The National Academies Logo
Research Associateship Programs
Fellowships Office
Policy and Global Affairs

Participating Agencies - NIST

  Sign InPrintable View

Opportunity at National Institute of Standards and Technology (NIST)

Passive Monitoring Toolkit for Research Reactor Maintenance and Cyber Security

Location

NIST Center for Neutron Research

RO# Location
50.61.01.C0360 Gaithersburg, MD

Please note: This Agency only participates in the February and August reviews.

Advisers

Name E-mail Phone
Sahin, Dagistan dagistan.sahin@nist.gov 301.975.4611

Description

The majority of the research reactors have their plant condition data available in digital format and stored in relational databases. As part of the NBSR Control Room Upgrade (CRU) phase 1, digitalization for most NBSR plant process parameters is completed. Unfortunately, available data trending systems are not tailored for research reactors and do not provide a cyber security warning. Currently, troubleshooting requires detailed knowledge of reactor operations and nuclear systems for accurate diagnosis of a particular failure and its cause. It is possible to use big-data analysis approaches to minimize such effort. Such volume of trending data through an engine can be analyzed using time-series based anomaly detection methods, and Deep Learning (DL) models, e.g. Recurrent Neural Networks (RNNs), to detect anomalies and predict failures related to reactor rundown and shutdown events. Furthermore, it is possible to continuously verify operator inputs from the console with expected plant behavior to detect and warn of a cyber attack where a single component or multiple digital components are compromised.

Goals

  • Reduce or prevent rundown events from occurring.
  • Increase awareness of & trending changes in reactor systems, such as increasing radiation levels, lowering flow rates, etc. Therefore, improve preventive maintenance.
  • Give insight and list possible sources when failures (or near failures) occur. This would reduce the time spent in troubleshooting.
  • Advance our understanding of long-term reactor operational behavior.
  • Alert for plant condition abnormalities including cyber attacks.

REFERENCES

L. BAYOU et al., “A Prediction-Based Method for False Data Injection Attacks Detection in Industrial Control Systems,” in Risks and Security of Internet and Systems, A. Zemmari et al., Eds., pp. 35–40, Springer International Publishing (2019).

S. LEE and J.-H. HUH, “An effective security measures for nuclear power plant using big data analysis approach,” J Supercomput 75 8, 4267 (2019); https://doi.org/10.1007/s11227-018-2440-4.

K. T. P. NGUYEN and K. MEDJAHER, “A new dynamic predictive maintenance framework using deep learning for failure prognostics,” Reliability Engineering & System Safety 188, 251 (2019); https://doi.org/10.1016/j.ress.2019.03.018.

Keywords:
Deep Learning; Nuclear Reactor; Recurrent Neural Networks; Industrial Control System; Cyber Security

Eligibility

Citizenship:  Open to U.S. citizens
Level:  Open to Postdoctoral applicants

Stipend

Base Stipend Travel Allotment Supplementation
$72,028.00 $3,000.00
Copyright © 2014. National Academy of Sciences. All rights reserved. 500 Fifth St. N.W., Washington, D.C. 20001.
Terms of Use and Privacy Statement.