An adaptive deviation-resistant neutron spectrum unfolding method based on transfer learning

Chenglong Cao; Quan Gan; Jing Song; Qi Yang; Liqin Hu; Fang Wang; Tao Zhou

doi:https://doi.org/10.1016/j.net.2020.04.028

추천

검색

자료유형: 학술저널

저자정보: Chenglong Cao (Chinese Academy of Sciences) Quan Gan (Chinese Academy of Sciences) Jing Song (Chinese Academy of Sciences) Qi Yang (Chinese Academy of Sciences) Liqin Hu (Chinese Academy of Sciences) Fang Wang (Chinese Academy of Sciences) Tao Zhou (Chinese Academy of Sciences)

저널정보: 한국원자력학회 Nuclear Engineering and Technology Nuclear Engineering and Technology 제52권 제11호

발행연도: 2020.11

수록면: 2,452 - 2,459 (8page)

DOI: https://doi.org/10.1016/j.net.2020.04.028

이용수

📌

연구주제

📖

연구배경

🔬

연구방법

🏆

연구결과

초록· 키워드

오류제보하기

Neutron spectrum is essential to the safe operation of reactors. Traditional online neutron spectrummeasurement methods still have room to improve accuracy for the application cases of wide energyrange. From the application of artificial neural network (ANN) algorithm in spectrum unfolding, its accuracy is difficult to be improved for lacking of enough effective training data. In this paper, an adaptivedeviation-resistant neutron spectrum unfolding method based on transfer learning was developed. Themodel of ANN was trained with thousands of neutron spectra generated with Monte Carlo transportcalculation to construct a coarse-grained unfolded spectrum. In order to improve the accuracy of theunfolded spectrum, results of the previous ANN model combined with some specific eigenvalues of thecurrent system were put into the dataset for training the deeper ANN model, and fine-grained unfoldedspectrum could be achieved through the deeper ANN model. The method could realize accurate spectrum unfolding while maintaining universality, combined with detectors covering wide energy range, itcould improve the accuracy of spectrum measurement methods for wide energy range. This method wasverified with a fast neutron reactor BN-600. The mean square error (MSE), average relative deviation(ARD) and spectrum quality (Qs) were selected to evaluate the final results and they all demonstratedthat the developed method was much more precise than traditional spectrum unfolding methods

#Neutron spectrum unfolding #Artificial neural network #Adaptive deviation-resistant # #Transfer learning

참고문헌 (29)

참고문헌 신청

J. M. Ortiz-Rodríguez / 2014 / A neutron spectrum unfolding computer code based on artificial neural networks / Radiat. Phys. Chem 95(4):428~431

J. Wang / 2019 / Neutron spectrum unfolding using three artificial intelligence optimization methods / Appl. Radiat. Isot 147:136~143

H. Xiong / 2018 / Research of On-Line Neutron Spectrum Measurement Method for Nuclear Reactor

R. Bedogni / 2007 / FRUIT : an operational tool for multisphere neutron spectrometry in workplaces / Nucl. Instrum. Methods Phys. Res. A 580(3):1301~1309

J. Shore / 1980 / Axiomatic derivation of the principle of maximum entropy and the principle of minimum cross-entropy / IEEE Trans. Inf. Theor 26(1):26~37

함께 읽어보면 좋을 논문

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