Risk Assessment Of Tunnel Construction Ventilation Based On Bayesian Network

With the rapid development of China’s infrastructure,the scale of tunnel construction has been gradually expanding.At the same time,the number of complex tunnels characterized by “long,large,deep,difficult and high” has been increasing.In large complex tunnel construction projects,tunnel related problems become more and more serious,with the increase of driving length.The ventilation environment of tunnel construction is a key issue,mainly manifested in the problems of insufficient ventilation supply,inappropriate temperature or humidity,and excessive concentration of harmful substances,which may lead to safety problems in tunnel construction.It can be seen that the research on risk assessment of tunnel construction ventilation is particularly important.In this thesis,the risk assessment of tunnel construction ventilation is studied based on Bayesian network,that aims to explore the key risk factors of tunnel construction ventilation,and that points out the key direction for tunnel construction ventilation design and site management.This thesis carries out the research based on the national key R & D plan,the theory and method of urban underground space development and construction(2019yfc0605104).The main research results of the thesis are as follows:(1)Taking Qinling Tiantaishan tunnel project as the research object,this thesis has tested the related indexes of ventilation system and tunnel environment.The test results show that there are many bad ventilation conditions in the construction of the project.Furthermore,based on the test results and the ventilation system establishment operation process,this thesis has identified the construction ventilation risk factors in the Qinling Tiantaishan Tunnel.And through the logical relationship of risk factors,a risk coupling model has been established.(2)This thesis introduces Monte Carlo method and expert investigation method to analyze the prior probability and conditional probability of ventilation risk factors in the construction of Qinling Tiantai Mountain tunnel.For the risk factors that have known the theoretical relationship with their parent nodes,the conditional probability is calculated by Monte Carlo method through tens of thousands of simulations.For the other risk factors,expert investigation method transforms expert knowledge experience into prior probability or conditional probability based on probability quantification standard.(3)Based on the results of risk identification and probability analysis,this thesis uses Bayesian network analysis software to realize the research on ventilation risk assessment of Qinling Tiantaishan tunnel construction.The research is mainly divided into risk event reasoning,risk factor diagnosis and risk control degree.The research results show that the ventilation risk of tunnel construction is likely to occur and that the control factors are mainly the layout of ventilation system or construction conditions.Among them,the more obvious risk factors are as follows: the length of air duct,the number of duct worn,the replacement of damaged duct,the position of duct end and the number of internal combustion engines.Furthermore,the thesis puts forward targeted control measures for the risk factors with obvious control degree.