基于贝叶斯网络的岩溶隧道涌水风险评估

The Assessment of Water Inrush Risk in Karst Tunnel Based on Bayesian Networks

在岩溶富水隧道施工过程中,隧道施工涌水灾害的原因十分复杂,而岩溶涌水风险的评估方法都有一定的局限性和地域性。为了较为准确地预测岩溶隧道的涌水灾害,以云桂铁路对门山隧道为工程背景,基于贝叶斯网络的不确定性的推理,构建岩溶隧道涌水风险的贝叶斯网络模型,并运用Netica软件的案例学习功能对统计数据进行分析,并与地质雷达及红外探水相结合。结果表明,理论评估结果与现场实际相符合。因此运用贝叶斯网络的岩溶隧道涌水风险评估,不仅能对岩溶隧道涌水风险进行科学的风险评估及预警,而且对制定岩溶隧道涌水防治对策和处治措施具有重要的理论意义和工程借鉴价值。

During the construction of tunnel with rich karst water,the causes of water in -rush disasters in tunnel construction are very complex and the assessment methods of karst water inrush have limitations and regions to a certain extent. To predict precisely the disaster of water inrush in karst tunnel,this paper on the background of Duimenshan tunnel of Yun-Gui rai lway,based on the uncertain inference of Bayesian network,builds the Bayes-ian network model of water inrush risk in karst tunnel and uses the case study of Netica software to analyze the statistical d a tawith the combination of geological radar and infrared water detector. The result shows that the theory assessmentmatches the actual condition. Therefore,using theBayes iannetwork toevaluate the risk of water inrush in karst tunnel can not only conduct a scientific risk assessment and warning,but also have important the-oretical significance and reference value of engineering for making prevention and treating measures of water inrush in the karst tunnel.