首页|基于贝叶斯网络的地铁施工风险管理研究

基于贝叶斯网络的地铁施工风险管理研究

Research on risk management of subway construction based on Bayesian network

扫码查看
为解决现有地铁施工风险管理方法在处理不确定性方面存在的不足,基于贝叶斯网络(BN)理论提出一种施工风险管理分析方法.首先,基于专家群决策等方法确定影响地铁风险管理的主要因素,并构建涉及多方参与主体的风险拓扑结构模型.然后利用贝叶斯正向因果推理预测不同情况下地铁施工风险发生概率,并结合反向诊断推理分析风险成因机理.最后,基于互信息(MI)指数法完成敏感性分析,找出敏感性致险因素.结合国内地铁项目历史数据,将该方法运用到地铁隧道工程安全风险管理实践中.结果表明,地铁施工风险概率为2.24%;未及时处理施工中存在的安全隐患是对其影响最大的风险因素.
In order to solve the deficiency problem existing subway construction risk management methods have in dealing with uncertainty,a BN-based risk management method was worked out.Firstly,main factors influencing subway construction risk management were identified and a BN topology model was built for the management.Then,the probabilities of subway construction risk under different conditions were predicted using forward causal reasoning and formation mechanism of subway construction risk was analyzed with reverse diagnosis reasoning.Finally,the sensitive risk factors were found with sensitivity analysis based on MI index.This method was employed in subway construction risk management practice.The results show that probability of subway construction project safety risk is 2.24% and failure to deal with hidden danger existing in construction timely is the most important factor influencing the probability.

Bayesian network(BN)subway constructionrisk managementuncertain reasoningmutual information(MI)

吴贤国、丁保军、张立茂、陈跃庆、薛莉敏、宋若昕

展开 >

华中科技大学土木工程与力学学院,湖北武汉430074

贝叶斯网络(BN) 地铁施工 风险管理 不确定推理 互信息(MI)

国家自然科学基金武汉市城建计划项目

51378235201208

2014

中国安全科学学报
中国职业安全健康协会

中国安全科学学报

CSTPCDCSCD北大核心
影响因子:1.548
ISSN:1003-3033
年,卷(期):2014.24(1)
  • 63
  • 8