Skip to content
Back to outputs

A method for measuring the robustness of diagnostic models for predicting the break size during LOCA

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Standard

A method for measuring the robustness of diagnostic models for predicting the break size during LOCA. / Tian, Xiange; Becerra, Victor; Bausch, Nils; Vinod, Gopika; Santhosh, T.V.

Annual Conference of the Prognostics and Health Management Society 2017. Vol. 8 Tampa, Floria, US : Prognostics and Health Management Society, 2017.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Harvard

Tian, X, Becerra, V, Bausch, N, Vinod, G & Santhosh, TV 2017, A method for measuring the robustness of diagnostic models for predicting the break size during LOCA. in Annual Conference of the Prognostics and Health Management Society 2017. vol. 8, Prognostics and Health Management Society, Tampa, Floria, US, Annual Conference of the Prognostics and Health Management Society 2017, St. Petersburg, United States, 2/10/17. <http://www.phmsociety.org/node/2279/>

APA

Tian, X., Becerra, V., Bausch, N., Vinod, G., & Santhosh, T. V. (2017). A method for measuring the robustness of diagnostic models for predicting the break size during LOCA. In Annual Conference of the Prognostics and Health Management Society 2017 (Vol. 8). Prognostics and Health Management Society. http://www.phmsociety.org/node/2279/

Vancouver

Tian X, Becerra V, Bausch N, Vinod G, Santhosh TV. A method for measuring the robustness of diagnostic models for predicting the break size during LOCA. In Annual Conference of the Prognostics and Health Management Society 2017. Vol. 8. Tampa, Floria, US: Prognostics and Health Management Society. 2017

Author

Tian, Xiange ; Becerra, Victor ; Bausch, Nils ; Vinod, Gopika ; Santhosh, T.V. / A method for measuring the robustness of diagnostic models for predicting the break size during LOCA. Annual Conference of the Prognostics and Health Management Society 2017. Vol. 8 Tampa, Floria, US : Prognostics and Health Management Society, 2017.

Bibtex

@inproceedings{6cce425c493f4570a61a15fb2076fec2,
title = "A method for measuring the robustness of diagnostic models for predicting the break size during LOCA",
abstract = "The diagnosis of loss of coolant accidents (LOCA) in nuclear reactors has attracted a great deal of attention in condition monitoring of nuclear power plants (NPPs) because the health of cooling system is crucial to the stability of the nuclear reactor. Multi-layer perceptron (MLP) neural networks have commonly been applied to LOCA diagnosis. The data used for training these models consists of a number of time-series data sets, each for a different break size, with the transient behavior of different measurable variables in the coolant system of the reactor following a LOCA. It is important to select a suitable architecture for the neural network that delivers robust results, in that the predicted break size is deemed to be accurate even for a break size that is not included in the training data sets. The objective of this paper is to present a simple method for measuring the robustness of diagnostic models for predicting the break size during the loss of coolant accidents. A robustness metric is proposed based on the leave-one-out approach and the mean squared error resulting from a diagnostics model. Using this metric it becomes possible to compare the robustness of different diagnostic models. Given data obtained from a high fidelity simulation of the coolant system of a nuclear reactor, four different diagnostic models are obtained and their properties compared and discussed. These models include a fully connected multi-layer perceptron with one hidden layer, a fully connected multi-layer perceptron with two hidden layers, a multi-layer perceptron with one hidden layer that is pruned using the optimal brain surgeon algorithm, a group method of data handling (GMDH) neural network, and an adaptive network based fuzzy inference system (ANFIS).",
keywords = "RCUK, EPSRC, EP/M018709/1",
author = "Xiange Tian and Victor Becerra and Nils Bausch and Gopika Vinod and T.V. Santhosh",
year = "2017",
month = oct,
day = "5",
language = "English",
volume = "8",
booktitle = "Annual Conference of the Prognostics and Health Management Society 2017",
publisher = "Prognostics and Health Management Society",
address = "United States",
note = "Annual Conference of the Prognostics and Health Management Society 2017 ; Conference date: 02-10-2017 Through 05-10-2017",
url = "https://www.phmsociety.org/events/conference/phm/17",

}

RIS

TY - GEN

T1 - A method for measuring the robustness of diagnostic models for predicting the break size during LOCA

AU - Tian, Xiange

AU - Becerra, Victor

AU - Bausch, Nils

AU - Vinod, Gopika

AU - Santhosh, T.V.

PY - 2017/10/5

Y1 - 2017/10/5

N2 - The diagnosis of loss of coolant accidents (LOCA) in nuclear reactors has attracted a great deal of attention in condition monitoring of nuclear power plants (NPPs) because the health of cooling system is crucial to the stability of the nuclear reactor. Multi-layer perceptron (MLP) neural networks have commonly been applied to LOCA diagnosis. The data used for training these models consists of a number of time-series data sets, each for a different break size, with the transient behavior of different measurable variables in the coolant system of the reactor following a LOCA. It is important to select a suitable architecture for the neural network that delivers robust results, in that the predicted break size is deemed to be accurate even for a break size that is not included in the training data sets. The objective of this paper is to present a simple method for measuring the robustness of diagnostic models for predicting the break size during the loss of coolant accidents. A robustness metric is proposed based on the leave-one-out approach and the mean squared error resulting from a diagnostics model. Using this metric it becomes possible to compare the robustness of different diagnostic models. Given data obtained from a high fidelity simulation of the coolant system of a nuclear reactor, four different diagnostic models are obtained and their properties compared and discussed. These models include a fully connected multi-layer perceptron with one hidden layer, a fully connected multi-layer perceptron with two hidden layers, a multi-layer perceptron with one hidden layer that is pruned using the optimal brain surgeon algorithm, a group method of data handling (GMDH) neural network, and an adaptive network based fuzzy inference system (ANFIS).

AB - The diagnosis of loss of coolant accidents (LOCA) in nuclear reactors has attracted a great deal of attention in condition monitoring of nuclear power plants (NPPs) because the health of cooling system is crucial to the stability of the nuclear reactor. Multi-layer perceptron (MLP) neural networks have commonly been applied to LOCA diagnosis. The data used for training these models consists of a number of time-series data sets, each for a different break size, with the transient behavior of different measurable variables in the coolant system of the reactor following a LOCA. It is important to select a suitable architecture for the neural network that delivers robust results, in that the predicted break size is deemed to be accurate even for a break size that is not included in the training data sets. The objective of this paper is to present a simple method for measuring the robustness of diagnostic models for predicting the break size during the loss of coolant accidents. A robustness metric is proposed based on the leave-one-out approach and the mean squared error resulting from a diagnostics model. Using this metric it becomes possible to compare the robustness of different diagnostic models. Given data obtained from a high fidelity simulation of the coolant system of a nuclear reactor, four different diagnostic models are obtained and their properties compared and discussed. These models include a fully connected multi-layer perceptron with one hidden layer, a fully connected multi-layer perceptron with two hidden layers, a multi-layer perceptron with one hidden layer that is pruned using the optimal brain surgeon algorithm, a group method of data handling (GMDH) neural network, and an adaptive network based fuzzy inference system (ANFIS).

KW - RCUK

KW - EPSRC

KW - EP/M018709/1

UR - http://www.phmsociety.org/node/2279/

M3 - Conference contribution

VL - 8

BT - Annual Conference of the Prognostics and Health Management Society 2017

PB - Prognostics and Health Management Society

CY - Tampa, Floria, US

T2 - Annual Conference of the Prognostics and Health Management Society 2017

Y2 - 2 October 2017 through 5 October 2017

ER -

ID: 10967542