Modeling nuclear reactor core dynamics with recurrent neural networks

Tülay Adali, Bora Bakal, Mustafa (Kemal) Sonmez, Reza Fakory, C. Oliver Tsaoi

Research output: Contribution to journalArticle

17 Citations (Scopus)

Abstract

A recurrent multilayer perceptron (RMLP) model is designed and developed for simulation of core neutronic phenomena in a nuclear power plant, which constitute a non-linear, complex dynamic system characterized by a large number of state variables. Training and testing data are generated by REMARK, a first principles neutronic core model [16]. A modified backpropagation learning algorithm with an adaptive steepness factor is employed to speed up the training of the RMLP. The test results presented exhibit the capability of the recurrent neural network model to capture the complex dynamics of the system, yielding accurate predictions of the system response. The performance of the network is also demonstrated for interpolation, extrapolation, fault tolerance due to incomplete data, and for operation in the presence of noise.

Original languageEnglish (US)
Pages (from-to)363-381
Number of pages19
JournalNeurocomputing
Volume15
Issue number3-4
DOIs
StatePublished - Jun 1997
Externally publishedYes

Fingerprint

Nuclear Reactors
Neural Networks (Computer)
Reactor cores
Recurrent neural networks
Multilayer neural networks
Nuclear Power Plants
Nonlinear Dynamics
Backpropagation algorithms
Fault tolerance
Extrapolation
Learning algorithms
Nuclear power plants
Noise
Interpolation
Dynamical systems
Learning
Testing

Keywords

  • Nuclear reactor core dynamics
  • Recurrent neural networks

ASJC Scopus subject areas

  • Artificial Intelligence
  • Cellular and Molecular Neuroscience

Cite this

Adali, T., Bakal, B., Sonmez, M. K., Fakory, R., & Tsaoi, C. O. (1997). Modeling nuclear reactor core dynamics with recurrent neural networks. Neurocomputing, 15(3-4), 363-381. https://doi.org/10.1016/S0925-2312(97)00018-0

Modeling nuclear reactor core dynamics with recurrent neural networks. / Adali, Tülay; Bakal, Bora; Sonmez, Mustafa (Kemal); Fakory, Reza; Tsaoi, C. Oliver.

In: Neurocomputing, Vol. 15, No. 3-4, 06.1997, p. 363-381.

Research output: Contribution to journalArticle

Adali, T, Bakal, B, Sonmez, MK, Fakory, R & Tsaoi, CO 1997, 'Modeling nuclear reactor core dynamics with recurrent neural networks', Neurocomputing, vol. 15, no. 3-4, pp. 363-381. https://doi.org/10.1016/S0925-2312(97)00018-0
Adali, Tülay ; Bakal, Bora ; Sonmez, Mustafa (Kemal) ; Fakory, Reza ; Tsaoi, C. Oliver. / Modeling nuclear reactor core dynamics with recurrent neural networks. In: Neurocomputing. 1997 ; Vol. 15, No. 3-4. pp. 363-381.
@article{56878777dbbd426db15b3faefc012869,
title = "Modeling nuclear reactor core dynamics with recurrent neural networks",
abstract = "A recurrent multilayer perceptron (RMLP) model is designed and developed for simulation of core neutronic phenomena in a nuclear power plant, which constitute a non-linear, complex dynamic system characterized by a large number of state variables. Training and testing data are generated by REMARK, a first principles neutronic core model [16]. A modified backpropagation learning algorithm with an adaptive steepness factor is employed to speed up the training of the RMLP. The test results presented exhibit the capability of the recurrent neural network model to capture the complex dynamics of the system, yielding accurate predictions of the system response. The performance of the network is also demonstrated for interpolation, extrapolation, fault tolerance due to incomplete data, and for operation in the presence of noise.",
keywords = "Nuclear reactor core dynamics, Recurrent neural networks",
author = "T{\"u}lay Adali and Bora Bakal and Sonmez, {Mustafa (Kemal)} and Reza Fakory and Tsaoi, {C. Oliver}",
year = "1997",
month = "6",
doi = "10.1016/S0925-2312(97)00018-0",
language = "English (US)",
volume = "15",
pages = "363--381",
journal = "Neurocomputing",
issn = "0925-2312",
publisher = "Elsevier",
number = "3-4",

}

TY - JOUR

T1 - Modeling nuclear reactor core dynamics with recurrent neural networks

AU - Adali, Tülay

AU - Bakal, Bora

AU - Sonmez, Mustafa (Kemal)

AU - Fakory, Reza

AU - Tsaoi, C. Oliver

PY - 1997/6

Y1 - 1997/6

N2 - A recurrent multilayer perceptron (RMLP) model is designed and developed for simulation of core neutronic phenomena in a nuclear power plant, which constitute a non-linear, complex dynamic system characterized by a large number of state variables. Training and testing data are generated by REMARK, a first principles neutronic core model [16]. A modified backpropagation learning algorithm with an adaptive steepness factor is employed to speed up the training of the RMLP. The test results presented exhibit the capability of the recurrent neural network model to capture the complex dynamics of the system, yielding accurate predictions of the system response. The performance of the network is also demonstrated for interpolation, extrapolation, fault tolerance due to incomplete data, and for operation in the presence of noise.

AB - A recurrent multilayer perceptron (RMLP) model is designed and developed for simulation of core neutronic phenomena in a nuclear power plant, which constitute a non-linear, complex dynamic system characterized by a large number of state variables. Training and testing data are generated by REMARK, a first principles neutronic core model [16]. A modified backpropagation learning algorithm with an adaptive steepness factor is employed to speed up the training of the RMLP. The test results presented exhibit the capability of the recurrent neural network model to capture the complex dynamics of the system, yielding accurate predictions of the system response. The performance of the network is also demonstrated for interpolation, extrapolation, fault tolerance due to incomplete data, and for operation in the presence of noise.

KW - Nuclear reactor core dynamics

KW - Recurrent neural networks

UR - http://www.scopus.com/inward/record.url?scp=0031171376&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031171376&partnerID=8YFLogxK

U2 - 10.1016/S0925-2312(97)00018-0

DO - 10.1016/S0925-2312(97)00018-0

M3 - Article

AN - SCOPUS:0031171376

VL - 15

SP - 363

EP - 381

JO - Neurocomputing

JF - Neurocomputing

SN - 0925-2312

IS - 3-4

ER -