弱胶结软岩隧道突水灾变演化特征与预测方法研究

Weakly cemented soft rock tunnels in water-rich regions are prone to water and mud inrush, collapses and other engineering disasters, resulting in serious casualties and huge economic losses. Based on model tests, the evolution characteristics of multi-field information such as displacement field, stress field, seepage field, temperature field, sound field, and geophysics field will be analyzed. The mathematical model based on the multi-field information coupling effect of water inrush in weakly cemented soft rock tunnels will be established. In addition, the relationship between multi-field parameter mutations and water inrush formation and its temporal and spatial evolution patterns and characteristics will be revealed. On this basis, a Bayesian Network probabilistic prediction model for tunnel water inrush considering multi-field and multi-factor information will be established. The guidelines and methods for early warning release at different water inrush levels will be proposed and water inrush prediction and early warning system based on the main controlling factors and changes of multiple information parameters would be built. Based on the numerical simulation of multi-physics coupling of seepage evolution, the evolution mechanism of water inrush in weakly cemented soft rock masses will be revealed. A multi-factor coordinated control method and timing for water inrush disasters of tunnel will be proposed. An expert emergency decision-making model considering multi-field and multi-factor information will be established. The research results would provide a scientific basis and theoretical basis for the prediction and control of water inrush disasters in weakly cemented soft rock tunnels.

弱胶结软岩隧道遇水极易发生突水突泥、塌方等工程灾害，导致严重的人员伤亡和巨大的经济损失。本项目拟通过模型试验，分析位移场、应力场、渗流场、温度场、声场及地球物理场等多场信息的演化特征，建立弱胶结软岩隧道突水多场信息耦合效应的数学模型，揭示多场参数突变与突水形成之间的关系及其时空演化规律和特征；在此基础上，建立隧道突水多场多因素的贝叶斯网络概率预测模型，提出不同突水级别预警发布的准则和方法，形成基于突水主控因素和多场信息参数变化的隧道突水的临突预报和预警体系；通过开展渗流演化多物理场耦合数值模拟，揭示弱胶结软岩隧道突水灾变演化机理，提出隧道突水灾害多因素协同控制方法与处置时机，建立多场多因素突水概率专家应急决策模型。研究成果能为弱胶结软岩隧道突水灾害的预测和控制提供科学依据和理论基础。

弱胶结软岩隧道遇水极易发生突水突泥、塌方等工程灾害，导致严重的人员伤亡和巨大的经济损失。本项目确定了PSTO流固耦合相似材料合理配比，开展了弱胶结软岩隧道突水灾变演化过程模型试验，采集了位移场、渗流场、应力场等多场监测数据，采用归一化方法分析了所采集多场信息的演化特征，揭示了多场参数突变与突水形成之间的关系及其时空演化规律和特征；开展了渗流演化多物理场耦合数值模拟方法研究，揭示了弱胶结软岩隧道突水灾变演化机理，开发了多场多因素突水概率专家应急决策软件，为隧道突水突泥灾害的快速决策和综合治理提供理论基础和高效工具；建立了隧道突水典型案例数据库，分析了突水主控因素，确定了贝叶斯网络模型结构，利用 EM 算法和所建立的数据库训练并建立了隧道突水多场多因素的贝叶斯网络概率预测模型，模型准确率达到90%以上，实现了对现场隧道突水等地质灾害进行突水级别的概率预测和有效预警。研究成果能为弱胶结软岩隧道突水灾害的预测和控制提供科学依据和理论基础。