Application of dimensionality reduction to BPNN-based assessment of nuclear accident source terms

doi:10.16265/j.cnki.issn1003-3033.2018.09.021

Abstract

Abstract: In order to accurately estimate the radioactive source terms in the nuclear accident consequence assessment, the BPNN inversion model of nuclear accident source terms was optimized. A PCA-BPNN assessment model was built after reducing the number of factors influencing the terms from 10 to 6. The cumulative contribution rate of the 6 factors was greater than 85%. Random forest algorithm was used to calculate values of importance of the influence factors to remove the wind direction and the height of the mixed layer, and to build an RF-BPNN assessment model. The estimation results by the above three models were analyzed. The results show that compared with BPNN model, PCA-BPNN model and RF-BPNN model have shorter estimation time and smaller error, the two can reflect the source information of the accident truthfully, and that RF-BPNN model has better accuracy and stability than PCA-BPNN model.

Key words: nuclear accident, source term assessment, back propagation neural network(BPNN), dimensionality reduction, principal component analysis(PCA), random forests(RF)

CLC Number:

X945

CHAI Chaojun, LING Yongsheng, YUE Qi, JIA Wenbao. Application of dimensionality reduction to BPNN-based assessment of nuclear accident source terms[J]. China Safety Science Journal, 2018, 28(9): 122-127.

References

[1] 唐秀欢,包利红,李华,等.卡尔曼线性滤波反演核设施核事故中核素释放率的研究[J].原子能科学技术, 2014, 48(10): 1 847-1 852.
TANG Xiuhuan, BAO Lihong, LI Hua, et, al. Radionuclide release rate inversion of nuclear accidents in nuclear facility based on Kalman filter[J].Atomic Energy Science and Technology, 2014, 48(10): 1 847-1 852.
[2] CHOI G P, KIM D Y, YOO K H, et al. Prediction of hydrogen concentration in nuclear power plant containment under severe accidents using cascaded fuzzy neural networks[J]. Nuclear Engineering and Design, 2016, 300: 393-402.
[3] REGE M A, TOCK R W. A simple neural network for estimating emission rates of hydrogen sulfide and ammonia from single point sources [J]. Journal of the Air & Waste Management Association, 1996, 46(10): 953-962.
[4] 凌永生, 侯闻宇, 贾文宝,等. 基于BP神经网络的核事故源项估算方法研究[J]. 中国安全科学学报, 2014, 24(8): 21-25.
LING Yongsheng, HOU Wenyu, JIA Wenbao, et, al. Study on nuclear accident source term assessment based on BP neural network[J]. China Safety Science Journal, 2014, 24(8): 21-25.
[5] DASTABAZ M. Technology for smart futures [M]. New York: Springer Press, 2018: 345-354.
[6] HOTELLING H. Analysis of a complex of statistical variables into principal components[J]. Journal of Educational Psychology, 1933,24: 417-441.
[7] 谢晓峰, 李夕兵, 尚雪义,等. PCA-BP神经网络模型估算导水裂隙带高度[J].中国安全科学学报, 2017, 27 (3): 100-105.
XIE Xiaofeng, LI Xibing, SHANG Xueyi, et, al. Prediction of height of water flowing fractured zone based on PCA-BP neural networks model [J]. China Safety Science Journal, 2017, 27 (3): 100-105.
[8] 施仲齐. 核或辐射应急的准备与响应[M]. 北京: 原子能出版社, 2010: 122-135.
[9] KATATA G, TERADA H, NAGAI H, et al. Numerical reconstruction of high dose rate zones due to the Fukushima dai-ichi nuclear power plant accident[J]. Journal of Environmental Radioactivity, 2011, 111(10): 2-12.
[10] International Atomic Energy Agency. Generic assessment procedures for determining protective actions during a reactor accident (955) [M]. Vienna: Radiation Safety Section International Atomic Energy Agency, 1997: 205-209.