|本期目录/Table of Contents|

[1]何志平,洪瑶,陈佟越,等.基于转置卷积神经网络的隧道火灾烟气层高度和逆流长度快速预测模型*[J].中国安全生产科学技术,2025,21(9):63-68.[doi:10.11731/j.issn.1673-193x.2025.09.008]
　HE Zhiping,HONG Yao,CHEN Tongyue,et al.A fast prediction model for smoke layer height and back-layering length in tunnel fire based on transposed convolutional neural network[J].Journal of Safety Science and Technology,2025,21(9):63-68.[doi:10.11731/j.issn.1673-193x.2025.09.008]
点击复制

基于转置卷积神经网络的隧道火灾烟气层高度和逆流长度快速预测模型*

《中国安全生产科学技术》[ISSN:1673-193X/CN:11-5335/TB]

卷:
21
期数:
2025年9期
页码:
63-68
栏目:
职业安全卫生管理与技术
出版日期:
2025-09-30

文章信息/Info

Title:
A fast prediction model for smoke layer height and back-layering length in tunnel fire based on transposed convolutional neural network
文章编号:
1673-193X(2025)-09-0063-06
作者:
何志平; 洪瑶; 陈佟越; 李建
（1.广州地铁集团有限公司，广东 广州 510330；
2.中国安全生产科学研究院，北京 100012）
Author(s):
HE Zhiping;  HONG Yao;  CHEN Tongyue;  LI Jian
（1.Guangzhou Metro Group Co.,Ltd.,Guangzhou Guangdong 510330,China；
2.China Academy of Safety Science and Technology,Beijing 100012,China）
关键词:
隧道火灾; 转置神经网络; 烟气层高度; 逆流长度
Keywords:
tunnel fire;  transposed neural network;  smoke layer height;  back-layering length
分类号:
X932
DOI:
10.11731/j.issn.1673-193x.2025.09.008
文献标志码:
A
摘要:
为解决隧道火灾中所产生有害烟气的高度分布和逆流长度的预测问题，基于转置卷积神经网络模型构建用于瞬时快速预测烟气分布和逆流长度的计算模型，给出数据处理的细节方法和损失函数，并通过数值数据库训练评估所提模型。研究结果表明：经过400个训练周期，训练集和测试集中的损失函数均趋于收敛，所提模型较好地捕捉到了烟气可见度物理场中的主要特征和细节；模型训练完毕后，在测试集的验证过程中取得了较好的结果，烟气层高度沿纵向分布的预测结果误差±0.5 m的频率不超过20%，烟气逆流长度的预测基本在20%相对误差限内。研究结果可为隧道火灾烟气层分布规律的快速预测提供参考。
Abstract:
In order to address the prediction problem of harmful smoke layer height distribution and back-layering length in tunnel fires,a computational model based on a transposed convolutional neural network was developed for rapid instantaneous prediction of smoke dispersion and back-layering length.Detailed methods for data processing and loss functions were provided,and the proposed model was trained and evaluated using a numerical database.The results demonstrate that after 400 training epochs,the loss functions in both the training and testing sets tend to converge.The proposed model effectively captures the main features and details in the physical field of smoke visibility well.After model training was completed,the model achieved satisfactory performance during the validation process on the testing set.The error in predicting the longitudinal distribution of smoke layer height is within ±0.5 m for no more than 20% of the cases,and the predictions for smoke back-layering length is mostly within a 20% relative error limit.These findings provide a valuable reference for the rapid prediction of smoke layer distribution patterns in tunnel fires.

参考文献/References:

［1］陈长坤,王玮玉,康恒,等.不同火源面积下隧道火灾温度场试验与数值模拟分析［J］.中国公路学报,2018,31(6):235-243. CHEN Changkun，WANG Weiyu，KANG Heng，et al.Experimental and numerical simulation analysis of temperature field of tunnel fire with different fire source areas［J］.China Journal of Highway and Transport，2018，31（6）：235-243.
［2］胡隆华.隧道火灾烟气蔓延的热物理特性研究［D］.合肥：中国科学技术大学,2006.
［3］ALAN B,RICHARD C.Handbook of tunnel fire safety［M］.ICE publishing，2012.
［4］CARVEL R.A review of tunnel fire research from Edinburgh［J］.Fire Safety Journal,2019,105:300-306.
［5］HONG Y,FU C.A full-range analytical solution of the critical velocity for smoke control in tunnel fires［J］.Case Studies in Thermal Engineering,2022，40:102531.
［6］LBAREZ C D,AYALA P,CANTIZANO A,et al.A coupled hybrid numerical study of tunnel longitudinal ventilation under fire conditions［J］.Case Studies in Thermal Engineering,2022,36:102202.
［7］CHEN K,ZHOU F,XIA T,et al.Ventilation network solution method based on coupling iteration of air state parameters and air quantity［J］.Journal of China University of Mining & Technology,2021,50(4):613-623.
［8］CHOW W K,WONG K Y,CHUNG W Y.Longitudinal ventilation for smoke control in a tilted tunnel by scale modeling［J］.Tunnelling and Underground Space Technology,2010,25(2):122-128.
［9］FAN C,ZHANG L,JIAO S,et al.Smoke spread characteristics inside a tunnel with natural ventilation under a strong environmental wind［J］.Tunnelling and Underground Space Technology,2018，82:99-110.
［10］CHOW W K.On smoke control for tunnels by longitudinal ventilation［J］.Tunnelling and Underground Space Technology,1998,13(3):271-275.
［11］LI J,LI Y,CHENG C,et al.A study on the effects of the slope on the critical velocity for longitudinal ventilation in tilted tunnels［J］.Tunnelling and Underground Space Technology,2019，89:262-267.
［12］JI J,HAN J Y,FAN C G,et al.Influence of cross-sectional area and aspect ratio of shaft on natural ventilation in urban road tunnel［J］.International Journal of Heat and Mass Transfer,2013，67:420-431.
［13］ZHANG T,WANG G,LI J,et al.Experimental study of back-layering length and critical velocity in longitudinally ventilated tunnel fire with various rectangular cross-sections［J］.Fire Safety Journal,2021，126:103483.
［14］WENG M C,LU X,LIU F,et al.Prediction of backlayering length and critical velocity in metro tunnel fires［J］.Tunnelling and Underground Space Technology,2015，47:64-72.
［15］CHEN L,HU L,ZHANG X,et al.Thermal buoyant smoke back-layering flow length in a longitudinal ventilated tunnel with ceiling extraction at difference distance from heat source［J］.Applied Thermal Engineering,2015，78:129-135.
［16］HU L,HUO R,CHOW W K,et al.Studies on buoyancy-driven back-layering flow in tunnel fires［J］.Experimental Thermal and Fluid Science,2008，32(8):1468-1483.
［17］ZHANG X,WU X,PARK Y,et al.Perspectives of big experimental database and artificial intelligence in tunnel fire research［J］.Tunnelling and Underground Space Technology,2021，108:103691.
［18］HONG Y,KANG J,FU C.Rapid prediction of mine tunnel fire smoke movement with machine learning and supercomputing techniques［J］.Fire Safety Journal,2021,127:103492.
［19］MCGRATTAN K,HOSTIKKA S,FLOYD J,et al.Fire dynamics simulator technical reference guide volume 1:mathematical model［M］.NIST Special Publication,2021.
［20］PASZKE A,GROSS S,MASSA F,et al.PyTorch:an imperative style,high-performance deep learning library［C］//32nd Conferenceon Neural Information Processing Systems,2019.
［21］ZHANG K,SUN M,HAN T X,et al.Residual networks of residual networks:multilevel residual networks［J］.IEEE Transactions on Circuits and Systems for Video Technology,2017，28(6):1303-1314.
［22］XU J,LI Z,DU B,et al.Reluplex made more practical:Leaky ReLU［C］.Proceedings-IEEE Symposium on Computers and communications，2020.

相似文献/References:

[1]王婉娣,张靖岩,冯炼,等.纵向通风对长大隧道火灾烟流控制分析[J].中国安全生产科学技术,2010,6(3):28.
　WANG Wan-di,ZHANG Jing-yan,FENG Lian,et al.Three-dimensional simulation analysis of smoke flow controlled by longitudinal ventilation in long tunnel fire[J].Journal of Safety Science and Technology,2010,6(9):28.
[2]刘晓阳,李炎锋,李俊梅,等.纵向通风条件下对隧道内烟气运动影响 因素的实验研究[J].中国安全生产科学技术,2012,8(4):24.
　LIU Xiao yang,LI Yan feng,LI Jun mei,et al.Experimental study on the effect of longitudinal ventilation and factors on smoke movement in tunnels[J].Journal of Safety Science and Technology,2012,8(9):24.
[3]陈震,姚斌,王文伟,等.隧道型细水雾灭火系统在公路隧道的设计与应用[J].中国安全生产科学技术,2013,9(7):144.[doi:10.11731/j.issn.1673-193x.2013.07.026]
　CHEN Zhen,YAO Bin,WANG Wen wei,et al.Design and application of tunneltype water mist fire suppression system[J].Journal of Safety Science and Technology,2013,9(9):144.[doi:10.11731/j.issn.1673-193x.2013.07.026]
[4]崔铁军,马云东.均匀火灾温场下隧道管片受力分析[J].中国安全生产科学技术,2014,10(5):37.[doi:10.11731/j.issn.1673-193x.2014.05.006]
　CUI Tiejun,MA Yundong.Internal forces analysis for tunnel segment under fire based on the coupled fluidmechanical-thermal model in uniform temperature field[J].Journal of Safety Science and Technology,2014,10(9):37.[doi:10.11731/j.issn.1673-193x.2014.05.006]
[5]黄有波,吕淑然,杨凯.火源功率与隧道阻塞比对临界风速变化规律影响研究[J].中国安全生产科学技术,2015,11(8):10.[doi:10.11731/j.issn.1673-193x.2015.08.002]
　HUANG You-bo,LV Shu-ran,YANG Kai.Study on influence law of fire power and blockage ratio on critical wind velocity in tunnel fire[J].Journal of Safety Science and Technology,2015,11(9):10.[doi:10.11731/j.issn.1673-193x.2015.08.002]
[6]李垣志,牛国庆,张轩轩.城市公路隧道火灾近火源区长度的研究[J].中国安全生产科学技术,2017,13(12):43.[doi:10.11731/j.issn.1673-193x.2017.12.007]
　LI Yuanzhi,NIU Guoqing,ZHANG Xuanxuan.Research on length of area near fire source in urban highway tunnel fire[J].Journal of Safety Science and Technology,2017,13(9):43.[doi:10.11731/j.issn.1673-193x.2017.12.007]
[7]胡嘉伟,毛军,郗艳红,等.隧道内偏置火源顶棚近壁面温度边界层效应研究[J].中国安全生产科学技术,2019,15(12):36.[doi:10.11731/j.issn.1673-193x.2019.12.006]
　HU Jiawei,MAO Jun,XI Yanhong,et al.Research on temperature boundary layer effect at near wall of ceiling for offset fire source in tunnel[J].Journal of Safety Science and Technology,2019,15(9):36.[doi:10.11731/j.issn.1673-193x.2019.12.006]
[8]张洪杰,张文鹏,范资悦,等.不同隧道坡度下射流风机临界风速研究[J].中国安全生产科学技术,2020,16(1):99.[doi:10.11731/j.issn.1673-193x.2020.01.016]
　ZHANG Hongjie,ZHANG Wenpeng,FAN Ziyue,et al.Study on critical wind velocity of jet fan under different tunnel slopes[J].Journal of Safety Science and Technology,2020,16(9):99.[doi:10.11731/j.issn.1673-193x.2020.01.016]
[9]张轩轩,万祥云,彭荣富,等.水幕与排烟系统协作下隧道火灾烟气防控研究[J].中国安全生产科学技术,2020,16(7):106.[doi:10.11731/j.issn.1673-193x.2020.07.017]
　ZHANG Xuanxuan,WAN Xiangyun,PENG Rongfu,et al.Research on smoke prevention and control of tunnel fire under cooperation of water curtain and smoke exhaust system[J].Journal of Safety Science and Technology,2020,16(9):106.[doi:10.11731/j.issn.1673-193x.2020.07.017]
[10]王羽尘,马健霄,刘宇航.公路隧道火灾环境人群逃生行为分析及多项Logit建模*[J].中国安全生产科学技术,2020,16(12):129.[doi:10.11731/j.issn.1673-193x.2020.12.021]
　WANG Yuchen,MA Jianxiao,LIU Yuhang.Analysis of crowd escape behavior in highway tunnel fire environment and multinomial Logit modeling[J].Journal of Safety Science and Technology,2020,16(9):129.[doi:10.11731/j.issn.1673-193x.2020.12.021]

备注/Memo

备注/Memo:
收稿日期： 2024-06-07
* 基金项目： “十四五”国家重点研发计划项目（2023YFC3010205）；北京市科技新星计划项目（20220484050，20230484417）；中国安全生产科学研究院基本科研业务费专项资金项目（2024JBKY02）
作者简介： 何志平，本科，高级工程师，主要研究方向为城际铁路安全。
通信作者： 洪瑶，博士，工程师，主要研究方向为火灾动力学。
更新日期/Last Update: 2025-09-30