基于自适应卡尔曼滤波和深度前馈神经网络的氚源项反演

张金龙; 崔威杰; 栗再新

doi:10.11889/j.1000-3436.2022-0104

您当前的位置：

首页 >

文章列表页 >

基于自适应卡尔曼滤波和深度前馈神经网络的氚源项反演

辐射防护 | 更新时间：2023-12-18

- 基于自适应卡尔曼滤波和深度前馈神经网络的氚源项反演
- Inversion of tritium source term based on adaptive Kalman filter and deep feedforward neural network
- 辐射研究与辐射工艺学报 2023年41卷第6期页码：60602
- 作者机构：
  
  1.核工业西南物理研究院成都 610000
- 作者简介：
  
  张金龙，男，1995年3月出生，现为核工业西南物理研究院硕士研究生，核能科学与工程专业
  栗再新，博士，研究员， E-mail: lizx@swip.ac.cn
- 基金信息：
- DOI：10.11889/j.1000-3436.2022-0104
  中图分类号：
扫描看全文
张金龙, 崔威杰, 栗再新. 基于自适应卡尔曼滤波和深度前馈神经网络的氚源项反演[J]. 辐射研究与辐射工艺学报, 2023, 41(06): 060602.

ZHANG Jinlong, CUI Weijie, LI Zaixin. Inversion of tritium source term based on adaptive Kalman filter and deep feedforward neural network[J]. Journal of Radiation Research and Radiation Processing, 2023, 41(6): 060602.
张金龙, 崔威杰, 栗再新. 基于自适应卡尔曼滤波和深度前馈神经网络的氚源项反演[J]. 辐射研究与辐射工艺学报, 2023, 41(06): 060602. DOI： 10.11889/j.1000-3436.2022-0104.

ZHANG Jinlong, CUI Weijie, LI Zaixin. Inversion of tritium source term based on adaptive Kalman filter and deep feedforward neural network[J]. Journal of Radiation Research and Radiation Processing, 2023, 41(6): 060602. DOI： 10.11889/j.1000-3436.2022-0104.

摘要

氘氚聚变反应被认为是能够最先实现商业发电的聚变反应，但氚的使用也带来了放射性安全问题。为探究适用于聚变堆事故后的大气释放氚源项反演的计算方法，本研究将自适应卡尔曼滤波与深度前馈神经网络相结合，建立聚变堆事故后的氚释放源项估计算法，对氚的释放高度及释放率进行反演。对神经网络使用滤波前后的观测值作为输入数据时的预测源强进行分析。结果表明，滤波能有效降低神经网络的预测误差。当监测数据误差为20%时，释放高度反演相对误差均值约为3%，释放率反演相对误差均值约为4%。

Abstract

Deuterium （D） and tritium （T） have been regarded as the first-generation fuels for achieving commercial fusion energy. However， the utilization of the radionuclide tritium introduces concerns related to radioactive safety. This study sought to investigate methods for estimating airborne tritium sources following a fusion reactor incident. An algorithm that combines an adaptive Kalman filter with a deep feedforward neural network was developed to determine the tritium release height and rate. By utilizing observed data both pre- and post-filtering as inputs， the neural network's predictions for the tritium release rate were analyzed. The findings indicate that filtering significantly lowers the prediction errors. Considering a 20% monitoring error， the average relative error for the estimated release height is approximately 3% and that for the release rate is approximately 4%.

关键词

自适应卡尔曼滤波深度前馈神经网络氚源项反演

Keywords

Adaptive Kalman filterDeep feedforward neural networkTritium source inversionCLC TL732

references

马建文, 秦思娴. 数据同化算法研究现状综述[J]. 地球科学进展, 2012, 27(7): 747-757. 10.1007/s11783-011-0280-zhttp://dx.doi.org/10.1007/s11783-011-0280-z

MA Jianwen, QIN Sixian. Recent advances and development of data assimilation algorithms[J]. Advances in Earth Sciences, 2012, 27(7): 747-757. 10.1007/s11783-011-0280-zhttp://dx.doi.org/10.1007/s11783-011-0280-z

刘蕴, 方晟, 李红, 等. 基于四维变分资料同化的核事故源项反演[J]. 清华大学学报(自然科学版), 2015, 55(1): 98-104. DOI: 10.16511/j.cnki.qhdxxb.2015.01.017http://dx.doi.org/10.16511/j.cnki.qhdxxb.2015.01.017.

LIU Yun, FANG Sheng, LI Hong, et al. Source inversion in nuclear accidents based on 4D variational data assimilation[J]. Journal of Tsinghua University (Science and Technology), 2015, 55(1): 98-104. DOI: 10.16511/j.cnki.qhdxxb.2015.01.017http://dx.doi.org/10.16511/j.cnki.qhdxxb.2015.01.017.

郑子豪, 陈春花, 朱婧娴, 等. 基于粒子滤波算法的源项反演及不确定性分析[J]. 辐射研究与辐射工艺学报, 2022, 40(4): 040602. DOI: 10.11889/j.1000-3436.2022-0024http://dx.doi.org/10.11889/j.1000-3436.2022-0024.

ZHENG Zihao, CHEN Chunhua, ZHU Jingxian, et al. Research on source term inversion and uncertainty analysis based on particle filter algorithm[J]. Journal of Radiation Research and Radiation Processing, 2022, 40(4): 040602. DOI: 10.11889/j.1000-3436.2022-0024http://dx.doi.org/10.11889/j.1000-3436.2022-0024.

Astrup P, Turcanu C, Puch R O, et al. Data assimilation in the early phase: Kalman filtering RIMPUFF[R]. Roskilde, Denmark: Riso National Laboratory, 2004: 78-89.

唐秀欢, 李华, 包利红. 核事故实时释放量集合卡尔曼滤波反演算法研究[J]. 原子能科学技术, 2014, 48(S1): 415-420. 10.7538/yzk.2014.48.S0.0415http://dx.doi.org/10.7538/yzk.2014.48.S0.0415

TANG Xiuhuan, LI Hua, BAO Lihong. Inversion algorithm for nuclear accident instantaneous release quantity based on ensemble Kalman filter[J]. Atomic Energy Science and Technology, 2014, 48(S1): 415-420. 10.7538/yzk.2014.48.S0.0415http://dx.doi.org/10.7538/yzk.2014.48.S0.0415

侯闻宇, 凌永生, 赵丹, 等. BP神经网络反演核事故源项中重要参数的研究[J]. 南京航空航天大学学报, 2015, 47(5): 778-784. DOI: 10.16356/j.1005-2615.2015. 05.024http://dx.doi.org/10.16356/j.1005-2615.2015.05.024.

HOU Wenyu, LING Yongsheng, ZHAO Dan, et al. Important parameters in inversion of nuclear accident source term based on BP neural network[J]. Journal of Nanjing University of Aeronautics & Astronautics, 2015, 47(5): 778-784. DOI: 10.16356/j.1005-2615.2015.05.024http://dx.doi.org/10.16356/j.1005-2615.2015.05.024.

Pati Y C, Krishnaprasad P S. Analysis and synthesis of feedforward neural networks using discrete affine wavelet transformations[J]. IEEE Transactions on Neural Networks, 1993, 4(1): 73-85. DOI: 10.1109/72.182697http://dx.doi.org/10.1109/72.182697.

Blum A L, Rivest R L. Training a 3-dose neural networks is NP-complete[J]. Neural Networks, 1992, 5(1): 117-127. 10.1016/s0893-6080(05)80010-3http://dx.doi.org/10.1016/s0893-6080(05)80010-3

Yang Y, Gao W. An optimal adaptive Kalman filter[J]. Journal of Geodesy, 2006, 80: 177-183. 10.1007/s00190-006-0041-0http://dx.doi.org/10.1007/s00190-006-0041-0

国家环境保护总局环境工程评估中心. 环境影响评价技术导则与标准汇编[M]. 北京: 中国环境科学出版社, 2005.

Environmental Engineering Assessment Center of the State Environmental Protection. Administration compilation of technical guidelines and standards for environmental impact assessment[M] Beijing: China Environmental Science Press, 2005.

Eckerman K, Harrison J, Menzel H G, et al. ICRP publication 119: compendium of dose coefficients based on ICRP publication 60[J]. Annals of the ICRP, 2013, 42(4): e1-e130. DOI: 10.1016/j.icrp.2013.05.003http://dx.doi.org/10.1016/j.icrp.2013.05.003.

Raskob W. UFOTRI: program for assessing the off-site consequences from accidental tritium releases[R]. Germany: Institute of Technology Karlruhe, 1990.

Raskob W. Description of the new version 4.0 of the tritium model UFOTRI including user guide[R]. Germany: Institute of Technology Karlruhe, 1993.

聂保杰. 聚变堆放射性核素的环境迁移与公众后果研究[D]. 合肥: 中国科学技术大学, 2017.

NIE Baojie. Research on radionuclides environmental migration and public consequences of fusion reactor[D].Hefei: University of Science and Technology of China, 2017.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

暂无数据