Abstract：Research purposes: The fracture zone is one of the most common geological anomalies in tunnel construction, which is prone to gushing water and mud, landslides and other geological disasters. To help solve the problems of relying on experience and low accuracy of traditional geological prediction interpretation, an intelligent identification algorithm of tunnel fault fracture zone based on Convolutional Neural Network (CNN) was proposed to assist professionals in identifying fault fracture zone quickly and effectively. Research conclusions:（1）Based on the geological background and standardized forecast picture data set, the YOLOv5 deep learning framework was adopted, the BoTNet module was introduced, and the self-attention mechanism was combined to form an intelligent identification algorithm for fault fracture zone (New_YOLOv5), which can realize the intelligent identification of its location, scale and other elements.（2）Compared with the traditional YOLOv5 algorithm, the optimized algorithm has a higher accuracy in identifying undesirable geologic bodies, with the growth rates of mAP as well as mAPmax values of 13.68% and 9.96%, respectively, where the highest mAP value can reach 84.79%.（3）To a certain extent, the research results can promote the further improvement of the interpretation level of the results of the over-advance prediction of tunnel fault fracture zones, and realize the technological progress from "experience-based, varying quality" to "intelligent identification, fast and effective", which can provide favorable technical support for the intelligent construction of tunnels, and has a good prospect of engineering application.
卢松,汪旭,李苍松,等.应用HSP法的TBM隧道施工地质预报技术研究[J].现代隧道技术,2020(3):30-35.Lu Song, Wang Xu, Li Cangsong, etc. Study on Geological Prediction Technology of HSP Method for TBM Tunnel[J]. Modern Tunnelling Technology, 2020(3):30-35.
[2]
雷冬,杜文康,朱国靖,等.基于机器视觉方法的高铁桥梁监测技术研究[J].铁道工程学报,2023(3):45-49.Lei Dong, Du Wenkang, Zhu Guojing,etc. Research on the Monitoring Technology of High-speed Railway Bridge Based on Machine Vision Method[J]. Journal of Railway Engineering Society,2023(3):45-49.
[3]
刘杰,许建国,高春丽,等.基于机器视觉的接触网吊弦缺陷检测研究[J].铁道工程学报,2022(5):91-97.Liu Jie, Xu Jianguo, Gao Chunli,etc. Machine Vision-based Research on the Inspection of Dropper Defects of Overhead Contact Line[J]. Journal of Railway Engineering Society,2022(5):91-97.
[4]
陈培帅,袁青,张子平,等.基于卷积神经网络的隧道富水破碎带地质超前预报图像解译方法[J].应用基础与工程科学学报,2022(1):196-207.Chen Peishuai, Yuan Qing, Zhang Ziping, etc. Image Interpretation Method for Geological Advance Prediction in Water-rich Fracture Zone of Tunnel Based on Convolutional Neural Network[J]. Journal of Basic Sciences and Engineering,2022(1):196-207.
[5]
闫星宇,李宗杰,顾汉明,等.基于深度卷积神经网络的地震数据溶洞识别[J].石油地球物理勘探,2022(1):1-11.Yan Xingyu, Li Zongjie, Gu Hanming, etc. Identification of Karst Caves in Seismic Data Based on Deep Convolutional Neural Network[J]. Oil Geophysical Prospecting, 2022(1):1-11.
[6]
罗虎,Miller Mark,张睿,等.基于计算机视觉技术和深度学习的隧道掌子面岩体裂隙自动识别方法研究[J].现代隧道技术,2023(1):56-65.Luo Hu, Miller Mark, Zhang Rui,etc. Research on the Automatic Identification Method for Rock Mass Fracture in Tunnel Face Based on Computer Vision Technology and Deep Learning[J]. Modern Tunnelling Technology,2023(1):56-65.
[7]
程久龙,王慧杰,徐忠忠,等.基于全卷积神经网络的钻孔瞬变电磁法岩层富水性预测研究[J].煤田地质与勘探,2023(1):289-297.Cheng Jiulong, Wang Huijie, Xu Zhongzhong,etc. Research on Aquifer Water Abundance Evaluation by Borehole Transient Electromagnetic Method Based on FCNN[J]. Coal Geology & Exploration,2023(1):289-297.
[8]
何易龙,文晓涛,王锦涛,等.基于3D U-Net++L3卷积神经网络的断层识别[J].地球物理学进展,2022(2):607-616.He Yilong, Wen Xiaotao, Wang Jintao, etc. Fault Recognition Based on 3D U-Net++L3 Convolutional Neural Network[J]. Progress in Geophysics, 2022(2):607-616.
2024, Vol. 41 
版权所有 © 2014 《铁道工程学报》编辑部