岩溶地区隧道泄水洞精细化探测技术  

Refined Detection Technology for Discharge Tunnel in Karst Area

作　　者：张锡江 聂俊丽[2] Zhang Xijiang;Nie Junli(College of Resources and Environmental Engineering,Guizhou University,Guiyang Guizhou 550025,China;Key Laboratory of Karst Georesources and Environment,Ministry of Education,Guizhou University,Guiyang Guizhou 550025,China)

机构地区：[1]贵州大学资源与环境工程学院,贵州贵阳550025 [2]贵州大学喀斯特地质资源与环境教育部重点实验室,贵州贵阳550025

出　　处：《工程地球物理学报》2024年第5期847-854,共8页Chinese Journal of Engineering Geophysics

基　　金：国家自然科学基金(编号:42264008)。

摘　　要：在岩溶地区进行隧道工程建设时,经常遇到因岩溶作用导致的塌陷、突泥、突水等地质灾害,给隧道施工带来不同程度的安全风险。为了预防地质灾害、降低其对隧道施工的影响,需要对隧道工程进行精准的超前地质预报。而岩溶区的泄水洞隧道具有断面尺寸小、干扰大、探测要求高、地质条件复杂等特点,使得超前地质预报的难度变大、准确度降低。采用地质调查法、地震波反射法、基于卷积神经网络的地质雷达法、瞬变电磁法和超前地质钻孔的综合超前地质预报技术,能对岩溶体进行精准探测。本文采用该综合超前地质预报技术在青后山泄水洞桩号XSDK59+017—XSDK58+887段进行了实际应用,预报了桩号XSDK58+928—XSDK58+918段的泥质填充型溶洞。结果表明,在岩溶区的泄水洞中,使用经过训练后的YOLOV8模型对雷达图像进行溶洞识别,置信率能达到80%以上,避免了人的主观性,提高了数据解释效率。并且采取这种综合超前地质预报方式能有效预报隧道前方的溶洞空间分布、充填物;为设计和施工提供了依据,避免了地质灾害对工程的不利影响,对于类似工程具有一定的借鉴意义。When constructing tunnels in karst areas,geological disasters such as collapse,mud and water inrush caused by karst often occur,which bring varying degrees of safety risks to tunnel construction.In order to prevent geological disasters and reduce their impact on tunnel construction,it is necessary to conduct accurate and advanced geological prediction of the tunnel.The drainage tunnel in karst areas has the characteristics of small cross-sectional size,large interference,high detection requirements,and complex geological conditions,which makes advance geological prediction more difficult and less accurate.Comprehensive advance geological prediction techniques such as geological survey,seismic wave reflection,convolutional neural network-based geological radar,transient electromagnetic method,and advance geological drilling can be used to accurately detect karst.This article uses the comprehensive advanced geological prediction technology to apply it in the XSDK59+017—XSDK58+887 section of the Qinghou Mountain drainage tunnel,predicting the mud filled karst caves in the XSDK58+928—XSDK58+918 section,improving the prediction accuracy and reducing construction risks.The results show that using the trained YOLOV8 model to identify karst caves in radar images in drainage tunnels in karst areas can achieve a confidence rate of over 80%,avoiding human subjectivity and improving data interpretation efficiency.Moreover,this comprehensive advanced geological prediction method can effectively predict the spatial distribution and filling materials of karst caves in front of the tunnel,which provides a basis for design and construction,avoids the adverse effects of geological disasters on the project,and has certain reference significance for similar projects.

关 键 词：超前地质预报 地质雷达 地震波反射法 瞬变电磁法 岩溶探测 

分 类 号：P631[天文地球—地质矿产勘探]