Research On Risk Assessment Of High-Speed Railway Tunnel Based On Bayesian Network Analysis

With the rapid development of high-speed rail construction in China,risk assessment and control of high-speed rail tunnel construction have received more and more attention.Relying on its advantages in uncertainty analysis and deduction,Bayesian network analysis has been increasingly used in the field of tunnel risk assessment.A high-speed rail tunnel risk assessment model is constructed through statistical analysis of accident cases,and the Bayesian network analysis method based on the accident tree is used,and the fuzzy hierarchical comprehensive evaluation method to realize the quantitative and qualitative assessment of the risk of tunnel construction is applied.At the same time,Risk assessment and risk cause diagnosis for a high-speed railway tunnel under complicated geological conditions are carried out.And corresponding control measures are proposed according to the risk acceptability.The main contents and conclusions of the thesis are as follows:(1)Through the analysis of the construction process of high-speed railway tunnel and the summary of the content of risk assessment and management,the core contents of the research is determined in the stage of tunnel construction with the highest risk.At the same time,based on the logical advantages of fault tree analysis and the dynamic analysis and data processing ability of Bayesian network,combined with the characteristics of high-speed railway tunnel engineering,the Bayesian network analysis method based on Fault Tree is selected as the main method of tunnel risk assessment.(2)According to the statistics of risk accidents in high-speed railway tunnels,the accidents of collapse,water and/or mud inrush,explosion(fire)happened frequently in the construction process are taken as the core risks for the analysis,and the risk index system is established through the identification of risk factors.The risk assessment model is constructed combining the probability of risk accident with the loss of risk.(3)A risk assessment model is applied to determine the risk level and acceptability of the three types of risk accidents in a constructing high-speed railway tunnel.Firstly,the prior probability of risk factors is determined by combining construction data and expert investigation,then the conditional probability of Bayesian network is modified by triangular fuzzy number and Noisy-Max/Min model according to the actual situation,the level of risk probability is determined,and the degree of risk loss is determined by combining AHP and fuzzy comprehensive evaluation method.After comprehensively considering the probability and loss level of the three types of accidents,the risk level and acceptable degree can be determined.(4)Use the posterior probability reasoning and sensitivity analysis functions of Bayesian network software Netica to identify the "most dangerous factors" that determine the occurrence of risk accidents and propose corresponding control measures.In the thesis,there are 23 pictures,46 tables and 82 references.