基于BP神经网络的木材着火时间预测
详细信息    查看全文 | 推荐本文 |
  • 英文篇名:Prediction for ignition time of wood based on BP neural network
  • 作者:翟春婕 ; 唐松泽
  • 英文作者:ZHAI Chun-jie;TANG Song-ze;Department of Forest Fire Protection, Nanjing Forest Police College;State Key Laboratory of Fire Science, University of Science and Technology of China;Department of Criminal Science and Technology, Nanjing Forest Police College;
  • 关键词:木材 ; 着火时间 ; BP神经网络
  • 英文关键词:wood;;ignition time;;BP neutral network
  • 中文刊名:XFKJ
  • 英文刊名:Fire Science and Technology
  • 机构:南京森林警察学院森林消防学院;中国科学技术大学火灾科学国家重点实验室;南京森林警察学院刑事科学技术学院;
  • 出版日期:2019-04-15
  • 出版单位:消防科学与技术
  • 年:2019
  • 期:v.38;No.286
  • 基金:江苏省教育厅高校哲学社会科学研究基金项目(2018SJA0592);; 中央高校基本科研业务费专项资金项目(LGYB201807);; 国家自然科学基金项目(61702269)
  • 语种:中文;
  • 页:XFKJ201904039
  • 页数:5
  • CN:04
  • ISSN:12-1311/TU
  • 分类号:107-111
摘要
提出了一种基于BP神经网络获得木材辐射热流作用下着火时间与物性参数相互关系的方法。建立了木材热解数值模型,模型中考虑了水蒸发、活性物质热解、热解气体流动等物理化学过程并使用实验数据进行了验证;基于数值模型提供的数据搭建并训练了BP神经网络;利用随机生成的数值数据验证了BP神经网络模型的可靠性。该方法结合了数值模拟和BP神经网络算法预测的优点,克服了实验样本数据量有限及数值模型求解较慢的问题。
        BP neural network is proposed to predict the relationship between pyrolysis parameters and ignition time of wood under radiant heat fluxes. The wood pyrolysis model was built. The model considers water evaporation, pyrolysis action and gas flow, and was verified by experimental data. Based on the data from the model, BP network was built and trained. By data generated randomly,the reliability of the BP network was verified. The method combined the advantage of mathematical simulation and BP network,and overcome the problems of limited experimental data and the big time requirement of solving the mathematical model.
引文
[1] KUNG H. A mathematical model of wood pyrolysis[J]. Combustion and Flame,1972,18(2):185-195.
    [2] WICHMAN I,ATREYA A. A simplified model for the pyrolysis of charring materials[J].Combustion and Flame,1987,68(3):231-247.
    [3] WOJTOWICZ M A,BASSILAKIS R,SMITH W. Modeling the evolution of volatile species during tobacco pyrolysis[J]. Journal of Analytical and Applied Pyrolysis,2003,66(1):235-261.
    [4] BAL N. Forty years of material flammability:An appraisal of its role,its experimental determination and its modelling[J]. Fire Safety Journal,2018,96:46-58.
    [5]叶茂昌,王金广.典型胶合板的着火时间研究[J].消防科学与技术,2017,36(1):31-33.
    [6]徐亮,丁严艳.锥形量热仪中材料点燃的数值模拟研究[J].消防科学与技术,2010,29(7):569-571.
    [7]杨守生,段楠.木材热解特性研究[J].消防科学与技术,2001,20(3):9-11.
    [8]沈永腾.氧浓度对瓦楞纸着火特性影响[J].消防科学与技术,2013,32(6):585-588.
    [9] BILBAO R,MASTRAL J,LANA J,et al. A model for the prediction of the thermal degradation and ignition of wood under constant and variable heat flux[J]. Journal of Analytical and Applied Pyrolysis,2002,62(1):63-82.
    [10] BILLAUD Y,KAISS A,CONSALVI J,et al. Monte Carlo estimation of Thermal Radiation from wildland fires[J]. International Journal of Thermal Science,2011,50(1):2-11.
    [11] DIDOMIZIO M,MULHERIN P,WECKMAN E. Ignition of wood under time-varying radiant exposures[J]. Fire Safety Journal,2016,82:131-144.
    [12] GUO Z F,YANG L Z. Experimental study on relation between ignition time of wood and heat flux increasing rate exposed to variable heat flux[J]. Journal of Thermal Science and Technology,2005,4(3):224-227.
    [13] RESZKA P,BOROWIEC P,STEINHAUS T,et al. A methodology for the estimation of ignition delay times in forest fire modelling[J]. Combustion and Flame,2012,159(12):3652-3657.
    [14] LAMORLETTE A. Analytical modeling of solid material ignition under a radiant heat flux coming from a spreading fire front[J]. Journal of Thermal Science&Engineering Application,2014,6(4):044501.
    [15] ZHAI C J,GONG J H,ZHOU X D,et al. Pyrolysis and spontaneous ignition of wood under time-dependent heat flux[J]. Journal of Analytical and Applied Pyrolysis. 2017,125:100-108.
    [16] ZHAI C J,PENG F,ZHOU X D,et al. Pyrolysis and ignition delay time of poly(methyl methacrylate)exposed to ramped heat flux[J]. Journal of Fire Sciences,2018,36(3):147-163.
    [17]丛爽.面向MATLAB工具箱的神经网络理论与应用[M].合肥:中国科学技术大学出版社,1998.

© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号

地址:北京市海淀区学院路29号 邮编:100083

电话:办公室:(+86 10)66554848;文献借阅、咨询服务、科技查新:66554700