Semi-supervised learning approach for crack detection and identification in advanced gas-cooled reactor graphite bricks

Research output: Contribution to conferencePaper

Abstract

One of the life-limiting components of an Advanced Gas cooled Reactor (AGR) is its graphite core. The bricks present in the core undergo radiolytic oxidation throughout their lifetime which causes graphite weight loss and irradiation which can result in some of the bricks developing cracks. Understanding the nature and extent of brick cracking within the core is key to ensuring continued and extended operation of the AGR fleet. A semi-supervised machine learning classification algorithm is proposed as a method for improving the detection of cracked graphite bricks, by combining the labels derived from infrequent, detailed inspections of the core, with unlabeled, more frequent monitoring measurements taken during refueling operations. Semi-supervised machine learning, which is an emerging field in nuclear power condition monitoring, is the combination of ideas from both supervised and unsupervised machine learning whereby the data that is used to train the algorithm is a combination of labeled and unlabeled data. This paper introduces the initial research that has been undertaken in creating a semi-supervised self-training algorithm to detect the presence of graphite brick cracks and then proceeds to show that there is an improvement in the classification of graphite bricks using a semi-supervised machine learning classifier compared to supervised machine learning classifiers. This improved classification performance is encouraging as it does not require time consuming and costly human analysis to obtain extra learning information from available data.

Conference

Conference10th International Topical Meeting on Nuclear Plant Instrumentation, Control, and Human-Machine Interface Technologies, NPIC and HMIT 2017
Abbreviated titleNPIC and HMIT 2017
CountryUnited States
CitySan Francisco
Period11/06/1715/06/17
Internet address

Fingerprint

Gas cooled reactors
Crack detection
Supervised learning
Brick
Graphite
Learning systems
Classifiers
Cracks
Condition monitoring
Nuclear energy
Labels
Inspection
Irradiation
Oxidation
Monitoring

Keywords

  • monitoring
  • semi-supervised learning
  • self-training

Cite this

Berry, C., West, G., McArthur, S., & Rudge, A. (Accepted/In press). Semi-supervised learning approach for crack detection and identification in advanced gas-cooled reactor graphite bricks. Paper presented at 10th International Topical Meeting on Nuclear Plant Instrumentation, Control, and Human-Machine Interface Technologies, NPIC and HMIT 2017, San Francisco, United States.
Berry, Craig ; West, Graeme ; McArthur, Stephen ; Rudge, Anna. / Semi-supervised learning approach for crack detection and identification in advanced gas-cooled reactor graphite bricks. Paper presented at 10th International Topical Meeting on Nuclear Plant Instrumentation, Control, and Human-Machine Interface Technologies, NPIC and HMIT 2017, San Francisco, United States.10 p.
@conference{7a73adb965504212af8cf1629bf1baf8,
title = "Semi-supervised learning approach for crack detection and identification in advanced gas-cooled reactor graphite bricks",
abstract = "One of the life-limiting components of an Advanced Gas cooled Reactor (AGR) is its graphite core. The bricks present in the core undergo radiolytic oxidation throughout their lifetime which causes graphite weight loss and irradiation which can result in some of the bricks developing cracks. Understanding the nature and extent of brick cracking within the core is key to ensuring continued and extended operation of the AGR fleet. A semi-supervised machine learning classification algorithm is proposed as a method for improving the detection of cracked graphite bricks, by combining the labels derived from infrequent, detailed inspections of the core, with unlabeled, more frequent monitoring measurements taken during refueling operations. Semi-supervised machine learning, which is an emerging field in nuclear power condition monitoring, is the combination of ideas from both supervised and unsupervised machine learning whereby the data that is used to train the algorithm is a combination of labeled and unlabeled data. This paper introduces the initial research that has been undertaken in creating a semi-supervised self-training algorithm to detect the presence of graphite brick cracks and then proceeds to show that there is an improvement in the classification of graphite bricks using a semi-supervised machine learning classifier compared to supervised machine learning classifiers. This improved classification performance is encouraging as it does not require time consuming and costly human analysis to obtain extra learning information from available data.",
keywords = "monitoring, semi-supervised learning, self-training",
author = "Craig Berry and Graeme West and Stephen McArthur and Anna Rudge",
year = "2017",
month = "4",
day = "4",
language = "English",
note = "10th International Topical Meeting on Nuclear Plant Instrumentation, Control, and Human-Machine Interface Technologies, NPIC and HMIT 2017, NPIC and HMIT 2017 ; Conference date: 11-06-2017 Through 15-06-2017",
url = "http://npic-hmit2017.org/",

}

Berry, C, West, G, McArthur, S & Rudge, A 2017, 'Semi-supervised learning approach for crack detection and identification in advanced gas-cooled reactor graphite bricks' Paper presented at 10th International Topical Meeting on Nuclear Plant Instrumentation, Control, and Human-Machine Interface Technologies, NPIC and HMIT 2017, San Francisco, United States, 11/06/17 - 15/06/17, .

Semi-supervised learning approach for crack detection and identification in advanced gas-cooled reactor graphite bricks. / Berry, Craig; West, Graeme; McArthur, Stephen; Rudge, Anna.

2017. Paper presented at 10th International Topical Meeting on Nuclear Plant Instrumentation, Control, and Human-Machine Interface Technologies, NPIC and HMIT 2017, San Francisco, United States.

Research output: Contribution to conferencePaper

TY - CONF

T1 - Semi-supervised learning approach for crack detection and identification in advanced gas-cooled reactor graphite bricks

AU - Berry, Craig

AU - West, Graeme

AU - McArthur, Stephen

AU - Rudge, Anna

PY - 2017/4/4

Y1 - 2017/4/4

N2 - One of the life-limiting components of an Advanced Gas cooled Reactor (AGR) is its graphite core. The bricks present in the core undergo radiolytic oxidation throughout their lifetime which causes graphite weight loss and irradiation which can result in some of the bricks developing cracks. Understanding the nature and extent of brick cracking within the core is key to ensuring continued and extended operation of the AGR fleet. A semi-supervised machine learning classification algorithm is proposed as a method for improving the detection of cracked graphite bricks, by combining the labels derived from infrequent, detailed inspections of the core, with unlabeled, more frequent monitoring measurements taken during refueling operations. Semi-supervised machine learning, which is an emerging field in nuclear power condition monitoring, is the combination of ideas from both supervised and unsupervised machine learning whereby the data that is used to train the algorithm is a combination of labeled and unlabeled data. This paper introduces the initial research that has been undertaken in creating a semi-supervised self-training algorithm to detect the presence of graphite brick cracks and then proceeds to show that there is an improvement in the classification of graphite bricks using a semi-supervised machine learning classifier compared to supervised machine learning classifiers. This improved classification performance is encouraging as it does not require time consuming and costly human analysis to obtain extra learning information from available data.

AB - One of the life-limiting components of an Advanced Gas cooled Reactor (AGR) is its graphite core. The bricks present in the core undergo radiolytic oxidation throughout their lifetime which causes graphite weight loss and irradiation which can result in some of the bricks developing cracks. Understanding the nature and extent of brick cracking within the core is key to ensuring continued and extended operation of the AGR fleet. A semi-supervised machine learning classification algorithm is proposed as a method for improving the detection of cracked graphite bricks, by combining the labels derived from infrequent, detailed inspections of the core, with unlabeled, more frequent monitoring measurements taken during refueling operations. Semi-supervised machine learning, which is an emerging field in nuclear power condition monitoring, is the combination of ideas from both supervised and unsupervised machine learning whereby the data that is used to train the algorithm is a combination of labeled and unlabeled data. This paper introduces the initial research that has been undertaken in creating a semi-supervised self-training algorithm to detect the presence of graphite brick cracks and then proceeds to show that there is an improvement in the classification of graphite bricks using a semi-supervised machine learning classifier compared to supervised machine learning classifiers. This improved classification performance is encouraging as it does not require time consuming and costly human analysis to obtain extra learning information from available data.

KW - monitoring

KW - semi-supervised learning

KW - self-training

UR - http://npic-hmit2017.org/

M3 - Paper

ER -

Berry C, West G, McArthur S, Rudge A. Semi-supervised learning approach for crack detection and identification in advanced gas-cooled reactor graphite bricks. 2017. Paper presented at 10th International Topical Meeting on Nuclear Plant Instrumentation, Control, and Human-Machine Interface Technologies, NPIC and HMIT 2017, San Francisco, United States.