基于因果贝叶斯网络的接触网风险评估模型研究  

Research on risk assessment model of catenary based on causative Bayesian network

作　　者：陈雍君[1] 李晓健 王劲 王卫东[2,3,4] 邱实 CHEN Yongjun;LI Xiaojian;WANG Jin;WANG Weidong;QIU Shi(School of Urban Economics and Management,Beijing University of Civil Engineering and Architecture,Beijing 102616,China;School of Civil Engineering,Central South University,Changsha 410075,China;National Engineering Research Center for High-speed Railway Construction Technology,Changsha 410075,China;Center for Railway Infrastructure Smart Monitoring and Management,Central South University,Changsha 410075,China)

机构地区：[1]北京建筑大学城市经济与管理学院,北京102616 [2]中南大学土木工程学院,湖南长沙410075 [3]高速铁路建造技术国家工程研究中心,湖南长沙410075 [4]中南大学轨道交通基础设施智能监控研究中心,湖南长沙410075

出　　处：《铁道科学与工程学报》2023年第8期3061-3071,共11页Journal of Railway Science and Engineering

基　　金：国家自然科学基金资助项目(52178442,U1734208);京沪高速铁路有限公司重点项目(2020-16)。

摘　　要：针对接触网建设阶段风险管控研究相对薄弱的问题,在充分研析接触网施工过程风险因素的基础上,判断风险因素间因果关系,建立接触网建设阶段风险评估模型,从因果逻辑视角分析接触网建设阶段风险。首先,从建设阶段、参建部门及建设要素3个维度分析接触网建设过程风险因素的外生变量与内生变量;然后,利用结构方程式描述变量之间的因果关系,建立多层次接触网风险结构因果模型网络关系分析图,并引入专家评议确定外生变量的先验概率与内生变量的条件概率,构建基于贝叶斯网络的接触网建设阶段风险评估模型;最后,通过贝叶斯网络敏感性分析和逆向推理分析找到接触网建设风险的事故最大因果链以及重要风险节点,在施工过程薄弱环节制定相应的风险预防与控制措施,降低风险发生的概率及影响,保证接触网工程按期交付以及运维阶段的检修效率。研究结果表明:在接触网建设阶段,风险发生的概率为0.6,所构建的接触网建设过程风险评估模型中有敏感性较高的节点4个,事故致因链5条。在假设接触网风险发生的条件下,逆向推理的结果显示预装配阶段和安装阶段的后验概率较高,分别为0.74和0.67。通过严格执行提出的风险控制措施,有助于提升接触网建设效率,降低运维阶段的维护成本。Aiming at the problem that research on risk management and control of the catenary construction stage is relatively weak,on the basis of fully analyzing risk factors in the construction process of the catenary,a judgment was made about the causal relationship of risk factors,the risk assessment model of the catenary construction stage was established,and the risks of the catenary construction stage was analyzed from a causal logic perspective.First,the exogenous and endogenous variables of risk factors in the catenary construction process were analyzed from the three dimensions,namely the construction stage,construction departments and construction elements.The causal relationships between variables were described using structural equations,and a network relationship analysis diagram of a multi-level catenary risk structure causal model was established.Second,by introducing expert evaluation method to determine the prior probabilities of exogenous variables and the conditional probabilities of endogenous variables,a risk assessment model of the catenary construction process was established based on the Bayesian network.Through Bayesian network sensitivity analysis and reverse reasoning analysis,the largest accident causal chain and the important risk nodes of catenary construction risks were obtained.The corresponding risk prevention and control measures for construction process weakness were proposed,which can reduce the probability and impact of risk,make sure the catenary project is delivered on time,and ensure maintenance efficiency during the period of operation and maintenance phase.The results of the research show that:in the catenary construction process,the probability of engineering risk is 0.6.There are 4 higher sensitive nodes and 5 accident-cause chains in the risk assessment model of the catenary construction process.On the condition of assuming catenary risk occurs,the results of backward inference show that the posterior probabilities of the pre-assembly stage and the installation stage are higher

关 键 词：接触网 结构因果模型 贝叶斯网络 风险评估模型 

分 类 号：U225.5[交通运输工程—道路与铁道工程]