首页 | 本学科首页   官方微博 | 高级检索  
     检索      

川藏铁路桥隧施工安全风险评价
引用本文:张锦,徐君翔.川藏铁路桥隧施工安全风险评价[J].安全与环境学报,2020(1):39-46.
作者姓名:张锦  徐君翔
作者单位:西南交通大学交通运输与物流学院;西南交通大学综合交通大数据应用技术国家工程实验室
摘    要:通过评估川藏铁路工程建设的施工风险等级,为高质量推进川藏铁路工程建设提供理论支撑。剖析了国内外学者关于风险评估研究的理论与方法,针对川藏铁路施工建设中的5座特大桥梁工程和9座超长隧道工程,分析了桥梁隧道建设工程的特征,构建了川藏铁路桥梁工程的17个安全风险评价指标体系和隧道工程的20个安全风险评价指标体系,通过建立基于模糊综合评价法的风险评估模型完成了对川藏铁路重点桥梁和隧道工程的风险评价,最后构造BP神经网络模型对风险评估结果进行验证,以川藏铁路部分重点桥梁工程评分数据和部分重点隧道工程评分数据为训练数据,以剩余评分结果为验证数据,预测桥梁和隧道工程的风险等级。结果表明:采用BP神经网络预测桥梁隧道工程安全风险等级的准确率高达98. 82%,BP神经网络对于该工程施工安全风险评价具有适用性;川藏铁路重点桥隧工程项目有50%处于较危险以上,只有20%的工程处于安全级别。

关 键 词:安全管理工程  川藏铁路  风险评价  模糊综合评价法  BP神经网络

Approach to the safety risk assessment of bridge and tunnel construction of Sichuan-Tibet Railway
ZHANG Jin,XU Jun-xiang.Approach to the safety risk assessment of bridge and tunnel construction of Sichuan-Tibet Railway[J].Journal of Safety and Environment,2020(1):39-46.
Authors:ZHANG Jin  XU Jun-xiang
Institution:(School of Transportation and Logistics,Southwest Jiaotong University,Chengdu 610031,Sichuan;National United Engineering Laboratory of Integrated and Intelligent Transportation,Southwest Jiaotong University,Chengdu 610031,Sichuan)
Abstract:This paper intends to make an analysis and evaluation of the characteristic features of the bridge and tunnel construction projects to provide a theoretical support for high-quality construction of Sichuan-Tibet Railway project. To provide a theoretical support for the high-quality construction of the said railway project,the given paper has been trying to make a scientific and reasonable evaluation of the construction risk level of the project by establishing a risk assessment model by doing the risk assessment of key bridges and tunnels of the said railway. And,finally,the BP neural network model has been adopted to validate the results of the risk assessment. The said BP network model under discussion can be said consisting of 3 layers with the maximum number of training times being 10 000,and of 15 hidden layer neurons,with their activation function being tansig function,the output layer activation function being logig function,the training function being the trainlm function,as well as the learning function being learngdm function. As to the training length and quality concerned,the scoring data of some key bridges and key tunnels of the railway,the remaining scoring results should be taken as the validation data for predicting and forecasting the risk levels of bridge and tunnel projects. The results of our evaluation prove that the confidential accuracy of our prediction and forecast of the safety risk grades of the bridge and tunnel projects by the BP neural network turns out as high as up to 92. 86%. Since 50%of the key bridge and tunnel projects of the said railway are potential with risks,of which only 20% can be put into the safety grade category. Therefore,it is strongly recommended that the construction companies be demanded to pay close attention to the safety supervision over the construction personnel,and strict monitoring control of the use of quality-enough equipment and materials. Besides,emergency plans and safety protection measures should be taken to prevent from various likely accidents by the relevant departments and units,so as to get rid of continuous operation of cold and anoxia,minimize the time length of the frequent geological inspection activities. Further analysis and inspection of the hidden dangers of construction risks are needed timely to reduce and even to eliminate such risk.
Keywords:safety control  Sichuan-Tibet Railway  risk evaluation  fuzzy comprehensive evaluation method  BP neural network
本文献已被 CNKI 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号