基于高精度地质雷达大断面隧道含水构造连续智能感知技术研究  

Research on Continuous Intelligent Perception Technology for Water-Bearing Structure of Large-Section Tunnel Based on High-Precision Geological Radar

作　　者：高军 马岩 Gao Jun;Ma Yan(Inner Mongolia Autonomous Region Institute of Transportation Science Development,Hohhot 010020,Inner Mongolia Autonomous Region,China;School of Civil Engineering,Tsinghua University,Beijing 100083,China;Hohhot Intelligent Technology Application School,Hohhot 010020,Inner Mongolia Autonomous Region,China)

机构地区：[1]内蒙古自治区交通运输科学发展研究院,内蒙古呼和浩特010020 [2]清华大学土木工程系,北京100083 [3]呼和浩特市智能技术应用学校,内蒙古呼和浩特010020

出　　处：《科学与信息化》2023年第13期95-97,共3页Technology and Information

摘　　要：大断面隧道含水地质构造具有地质条件复杂、地质灾害多发、孕灾特性隐蔽等特点,传统的检测方法常采用高精度地质雷达进行超前地质检测预报,存在杂波影响大、人工依赖强、检测精度低等问题。本文从高精度地质雷达检测原理出发,分析了GPR检测在大断面隧道含水构造态势感知中面临的难题,为GPR检测雷达特征提取、分析和灾害预测分别构建了数学模型,并提出了一种基于ResNet卷积神经网络的连续智能感知技术实现。The water-bearing geological structure of large-section tunnel has the characteristics of complex geological conditions,frequent geological disasters and hidden disaster concealment.Traditional detection method often uses highprecision geological radar for advanced geological detection and forecasting,which has problems such as large clutter influence,strong artificial dependence and low detection accuracy.Starting from the principle of high-precision geological radar detection,this paper analyzes the problems of GPR detection in the trend perception of water-bearing structure in largesection tunnels,constructs mathematical models for GPR detection radar feature extraction,analysis and disaster prediction,and proposes a continuous intelligent perception technology based on ResNet convolutional neural network.

关 键 词：高精度地质雷达 大断面隧道 含水构造 智能感知技术 

分 类 号：TN9[电子电信—信息与通信工程]