Automated detection of corrosion in used nuclear fuel dry storage canisters using residual neural networks

Theodore Papamarkou; Hayley Guy; Bryce Kroencke; Jordan Miller; Preston Robinette; Daniel Schultz; Jacob Hinkle; Laura Pullum; Catherine Schuman; Jeremy Renshaw; Stylianos Chatzidakis

doi:https://doi.org/10.1016/j.net.2020.07.020

추천

검색

자료유형: 학술저널

저자정보: Theodore Papamarkou (Computational Sciences and Engineering Division, Oak Ridge National Laboratory, Oak Ridge) Hayley Guy (Department of Mathematics, North Carolina State University, Raleigh, NC, USA) Bryce Kroencke (Department of Computer Science, University of California, Davis, CA, USA) Jordan Miller (Center for Cognitive Ubiquitous Computing, Arizona State University, Tempe, AZ, USA) Preston Robinette (Presbyterian College, Clinton, SC, USA) Daniel Schultz (Innovative Computing Laboratory, University of Tennessee, Knoxville, TN, USA) Jacob Hinkle (Computational Sciences and Engineering Division, Oak Ridge National Laboratory, Oak Ridge, TN, USA) Laura Pullum (Computer Science and Mathematics Division, Oak Ridge National Laboratory, Oak Ridge, TN, USA h) Catherine Schuman (Computer Science and Mathematics Division, Oak Ridge National Laboratory, Oak Ridge, TN, USA h) Jeremy Renshaw (Electric Power Research Institute, Palo Alto, CA, USA) Stylianos Chatzidakis (Reactor and Nuclear Systems Division, Oak Ridge National Laboratory, Oak Ridge, TN, USA)

저널정보: 한국원자력학회 Nuclear Engineering and Technology Nuclear Engineering and Technology 제53권 제2호

발행연도: 2021.2

수록면: 657 - 665 (9page)

DOI: https://doi.org/10.1016/j.net.2020.07.020

이용수

📌

연구주제

📖

연구배경

🔬

연구방법

🏆

연구결과

초록· 키워드

오류제보하기

Nondestructive evaluation methods play an important role in ensuring component integrity and safety inmany industries. Operator fatigue can play a critical role in the reliability of such methods. This isimportant for inspecting high value assets or assets with a high consequence of failure, such as aerospaceand nuclear components. Recent advances in convolution neural networks can support and automatethese inspection efforts. This paper proposes using residual neural networks (ResNets) for real-timedetection of corrosion, including iron oxide discoloration, pitting and stress corrosion cracking, in drystorage stainless steel canisters housing used nuclear fuel. The proposed approach crops nuclear canisterimages into smaller tiles, trains a ResNet on these tiles, and classifies images as corroded or intact usingthe per-image count of tiles predicted as corroded by the ResNet. The results demonstrate that such adeep learning approach allows to detect the locus of corrosion via smaller tiles, and at the same time toinfer with high accuracy whether an image comes from a corroded canister. Thereby, the proposedapproach holds promise to automate and speed up nuclear fuel canister inspections, to minimize inspectioncosts, and to partially replace human-conducted onsite inspections, thus reducing radiationdoses to personnel.

#Convolutional neural networks #Corrosion #Deep learning #Dry storage canisters #Feature detection #Residual neural networks

참고문헌 (0)

참고문헌 신청

참고문헌이 DBpia에서 서비스 중이라면, [참고문헌 신청]을 통해 등록해보세요

함께 읽어보면 좋을 논문

논문 유사도에 따라 DBpia 가 추천하는 논문입니다. 함께 보면 좋을 연관 논문을 확인해보세요!