基于贝叶斯网络的综合管廊运维灾害风险分析

Disaster risk analysis of utility tunnel operation and maintenance based on the Bayesian network

针对综合管廊运维风险安全管理和现行评估方法的研究不足,分析综合管廊运维过程中潜在的灾害风险因素,构建了基于贝叶斯网络的灾害风险评估模型。采用专家调查权重法计算父节点先验风险概率等级分布,根据贝叶斯网络链式传递原则计算模型中其余各节点风险概率等级分布,评估综合管廊运维期间潜在的灾害风险,得出整体风险概率等级及灾害风险事件发生的关键路径和关键风险因素。结果表明,该方法可以有效地评估综合管廊运维灾害风险等级,能够满足综合管廊运维风险管理及防灾减灾工作中的实际需求。

This paper is aimed at developing a disaster risk assessment method based on the Bayesian network theory,which can be used for analyzing and evaluating the disaster risk levels in the operation and maintenance of the utility tunnels.This is because the disaster events of such utility tunnels are usually uncertain and unpredictable,and,therefore,are not easy to be given accurate description with its probability being reasonable and dependable.What is more,Bayesian network should be regarded as a kind of uncertain knowledge denotation and the most effective theoretical reasoning model.To achieve the purpose of the risk assessment,it is thus necessary to introduce first of all the potential disaster risks with the operation and maintenance of the utility tunnel.According to the potential location of the disasters,such risks can be divided into 2 categories,the structure damage and the interior disasters of the utility tunnel.According to the above situation,it would be convenient to establish and propose a disaster risk assessment model based on the Bayesian network theory for the risk qualitative analysis.To calculate and find the probability level distribution of the risk factors of such tunnels,that is,the priori risk probability of the parent nodes,it is necessary to employ the expert survey weight method.According to the chain transfer rule of Bayesian network,the risk probability level distribution of the rest of the sub-nodes has been worked out.And,then,the status quo has been gained of the operation and maintenance of the utility tunnel in Beijing,and then the risk factors scoring data can be collected by using a questionnaire to survey the designers,contractors,the operation and maintenance workers of the utility tunnel and research members in the institutions concerned,so as to make the proposed method fit for calculation of the said Bayesian model.What is more,since the probability level distribution of the overall operation and maintenance disaster risk are high,it would be necessary to identify and determine the key path-ways and the key risk factors of the disaster risk event.Hence,finally,the risk assessment results have been proven to be in line with the overall status of the utility tunnels in Beijing.Thus,the method presented in this paper proves to be effective in supporting the disaster risk assessment and decision making approach during the stage of operation and maintenance of the utility tunnel.