【摘要】 为解决隧道施工中围岩富水性信息获取困难且不能及时对突涌水风险进行预警的问题，提出一种评价掌子面前方围岩富水性的数字孪生与风险决策框架，该框架通过点云模型实现对围岩富水性的虚拟映射，并提出一种模糊云概率模型，用于精细化评估每米围岩的突涌水风险。框架包括3个步骤：（1）通过瞬变电磁法（TEM）获取围岩电阻率数据，并进行三维空间配准，构建富水性数据库；（2）基于掌子面的几何尺寸构建前方点云模型，应用KNN算法赋予点云富水性属性，构建围岩富水性数字孪生体；（3）基于模糊云概率模型量化围岩富水性数字孪生体的不确定性，智能评估掌子面前方突涌水风险。将该方法应用于晏家寨隧道工程，结果表明：所提框架可快速提取掌子面前方围岩富水性信息，并构建数字孪生体对围岩富水性进行虚拟映射；通过对比开挖揭露的围岩突涌水实际情况，验证了提出的模糊云概率模型在掌子面前方围岩突涌水风险评估中的准确性。更多还原

【Abstract】 To address the challenges of acquiring water-richness information in surrounding rock during tunnel construction and the lack of timely early warning for water-inrush risks, this study proposes a digital twin and risk decision-making framework for evaluating the water-richness of surrouding rock ahead of tunnel face. The framework achieves virtual mapping of rock water-richness through point cloud modeling and introduces a fuzzy cloud probability model for refined risk assessment of water-inrush per meter. The framework comprises three key steps:(1) Acquisition of rock resistivity data via transient electromagnetic method(TEM) with 3D spatial registration to construct a water-richness database;(2) Generation of forward point cloud models based on tunnel face geometry, where Knearest neighbors(KNN) algorithm assigns water-richness attributes to establish a digital twin of water-richness of surrounding rock;(3) Uncertainty quantification of the digital twin using the fuzzy cloud probability model for intelligent water-inrush risk assessment. Application in Yanjiazhai Tunnel demonstrates that the framework enables rapid extraction of water-richness information and virtual mapping through digital twins. Comparative analysis with actual water-inrush conditions during excavation verifies the accuracy of the proposed fuzzy cloud probability model in risk evaluation of water inrush ahead of tunnel face.更多还原