考虑施工时序的隧道工作面坍塌风险分析方法与验证?

岩石隧道钻爆法开挖过程中存在多工序交叉作业，且频繁遭遇复杂的地质环境，导致工作面坍塌事故时有发生。为分析隧道施工中工作面坍塌风险，结合 T‐S（Takagi‐Sugeno）模糊事故树与贝叶斯网络，提出了一种考虑岩石隧道施工基本事件时序性的工作面坍塌风险分析方法。广泛调研了 46 起钻爆法岩石隧道工作面坍塌事故，筛选辨识出影响工作面稳定性的 30 项致险因子与孕险环境作为基本事件，基于统计数据量化计算各基本事件的发生概率模糊子集与各节点间条件概率，构建了隧道工作面坍塌 T‐S 模糊事故树并映射为贝叶斯网络，运用贝叶斯网络的双向推理功能，计算得到各基本事件的后验概率与重要度，筛选出部分对隧道工作面坍塌影响较大的关键基本事件。随后结合隧道施工过程中各基本事件发生的时序性，通过在贝叶斯网络中逐次更新对应事件的风险状态， 得到工作面坍塌风险发生概率的时序性变化。以油坊坪隧道坍塌事故为工程实例，讨论了隧道建设中考虑施工时序的工作面坍塌风险分析的可行性与挑战，给出了岩石隧道不同施工阶段的风险管控方案与重点，为岩石隧道施工安全风险分析与管控提供了新的理论基础与技术手段。

Method and Validation for Tunnel Face Collapse Risk Analysis Considering Construction Sequence

Rock tunnel excavation using the drill-and-blast method often involves multiple concurrent operations and encounters complex geological conditions, leading to the occurrence of tunnel face collapse accidents. To analyze the risk of tunnel face collapse during tunnel construction, a risk analysis method considering the temporality sequence of basic events during rock tunnel construction was proposed, utilizing T-S (Takagi-Sugeno) fuzzy fault trees and Bayesian network. Extensive research was conducted on 46 tunnel face collapse accidents in rock tunnels excavated using the drill-and-blast method. Thirty hazardous factors and risk-prone environments were identified as basic events. The occurrence probabilities and conditional probabilities between nodes were quantitatively calculated based on statistical data, generating fuzzy subsets for the occurrence probabilities of each basic event. A T-S fuzzy fault tree for tunnel face collapse was constructed and mapped to a Bayesian network. By employing the bidirectional inference capability of Bayesian networks, the posterior probabilities and importance of each basic event were calculated, and the key basic events that had a significant impact on tunnel face collapse were identified and selected. Furthermore, considering the temporal occurrence of basic events during the tunnel construction process, the risk status of corresponding events was sequentially updated in the Bayesian network, enabling the assessment of the temporal variations in the probability of tunnel face collapse risk. Using the collapse accident in the Youfangping Tunnel as an engineering example, the feasibility and challenges of considering construction sequencing in the risk analysis of tunnel collapses during tunnel construction were discussed, Risk management strategies and focal points for different construction phases of rock tunnels were provided. This research offers a novel theoretical framework and technical means for the analysis and management of safety risks in rock tunnel construction.