摘要
富水破碎不良地质区在隧道施工中容易诱发涌水灾害,为准确分析隧道围岩的富水破碎风险,且满足自动化、定量化风险分析需求,基于开挖数据构建模糊贝叶斯网络风险评估模型,通过隶属函数量化地质参数的不确定性,并结合贝叶斯概率推理,融合隧道地震预报法与瞬变电磁法的探测数据,得到围岩富水破碎风险概率;进一步利用三维体素模型将风险概率映射至三维坐标,可视化表达风险的空间分布特征.选取典型长大深埋隧道进行实验分析,结果表明:评估模型对地下水情况与岩体完整性分类的准确率分别为80.91%和82.81%,且不受数据完备性限制,能够在单一或多源数据条件下完成定量分析;所建三维体素模型为风险防控提供了有效参考,其中,相较于单一数据,多源数据融合分析结果与现场揭露的富水区、破碎带位置吻合度更高.
Unfavorable water-rich and fractured geological zones easily bring about water inrush disasters during tunnel construction.To accurately analyze water-rich and fractured risks in tunnel surrounding rock and address the need for automated and quantitative risk analysis,a fuzzy Bayesian network model for risk assessment was constructed by using tunnel excavation data.Geological parameter uncertainty was quantified via membership functions,and Bayesian probabilistic inference was employed to integrate data from tunnel seismic prediction and transient electromagnetic methods,yielding the probability of water-rich and fractured risks.A three-dimensional voxel model was used to map the risk probability to spatial coordinates,visualizing the spatial distribution of risks.A typical deep-buried long tunnel was selected for analysis.The results demonstrate that the assessment model achieves classification accuracies of 80.91%for groundwater conditions and 82.81%for rock mass integrity.Not affected by incomplete data,the model can conduct quantitative analysis under both single-source and multi-source data conditions.The constructed three-dimensional voxel model provides an effective reference for risk prevention and control.Analysis results of multi-source data fusion show higher spatial consistency with field-exposed water-rich and fractured zones than those of single-source data.
作者
朱庆
郑威鹏
吴浩宇
丁雨淋
郭永欣
王强
刘利
张骏骁
ZHU Qing;ZHENG Weipeng;WU Haoyu;DING Yulin;GUO Yongxin;WANG Qiang;LIU Li;ZHANG Junxiao(Faculty of Geosciences and Environmental Engineering,Southwest Jiaotong University,Chengdu 611756,China;Faculty of Geomatics,Lanzhou Jiaotong University,Lanzhou 730070,China;China Railway 18th Bureau Group Corporation Limited,Tianjin 300222,China;Qilu Aerospace Information Research Institute,Jinan 250100,China)
出处
《西南交通大学学报》
北大核心
2025年第5期1071-1079,共9页
Journal of Southwest Jiaotong University
基金
国家重点研发计划(2022YFB3904103)
国家自然科学基金项目(42001336)
国家铁路集团重大课题(K2021G027)
成都市重点研发支撑计划(2022YF0800001GX)。
关键词
隧道
风险分析
三维地质建模
模糊贝叶斯网络
数据融合
tunnel
risk analysis
three-dimensional geological modeling
fuzzy Bayesian network
data fusion
