[1]李元军,李萧桐,罗旭,等.基于SVM的地质雷达检测道床病害智能识别[J].高速铁路技术,2025,16(06):83-89.[doi:10.12098/j.issn.1674-8247.2025.06.012]
　LI Yuanjun,LI Xiaotong,LUO Xu,et al.Intelligent Recognition of Ballasted Track Bed Diseases Using Ground Penetrating Radar Based on Support Vector Machine[J].HIGH SPEED RAILWAY TECHNOLOGY,2025,16(06):83-89.[doi:10.12098/j.issn.1674-8247.2025.06.012]
点击复制

基于SVM的地质雷达检测道床病害智能识别()

《高速铁路技术》[ISSN:1674-8247/CN:51-1730/U]

卷:
16卷
期数:
2025年06期
页码:
83-89
栏目:
建造运维
出版日期:
2025-12-30

文章信息/Info

Title:
Intelligent Recognition of Ballasted Track Bed Diseases Using Ground Penetrating Radar Based on Support Vector Machine
文章编号:
1674-8247(2025)06-0083-07
作者:
李元军1; 李萧桐2; 罗旭2; 李光耀1; 井国庆2
(1.大秦铁路股份有限公司, 山西 大同 030023; 2. 北京交通大学, 北京 100044)
Author(s):
LI Yuanjun1;  LI Xiaotong2;  LUO Xu2;  LI Guangyao1;  JING Guoqing2
(1. Datong-Qinhuangdao Railway Co., Ltd., Datong 030023, China; 2. Beijing Jiaotong University, Beijing 100044, China)
关键词:
铁路道床;  地质雷达;  无损检测;  病害识别;  支持向量机
Keywords:
railway track bed;  ground penetrating radar;  non-destructive testing;  disease recognition;  support vector machine(SVM)
分类号:
U216.3; P631.3
DOI:
10.12098/j.issn.1674-8247.2025.06.012
文献标志码:
A
摘要:
针对碎石道床病害检测中传统方法效率低、偏差大等问题,提出一种基于支持向量机(Support Vector Machine,SVM)的地质雷达道床病害智能识别方法。通过提取道床雷达图像中反映病害特征的灰度共生矩阵参数——能量、熵、对比度和相关性,构建病害特征数据库,并采用不同核函数的支持向量机模型,对翻浆冒泥、含水异常和下沉3种典型道床病害进行分类识别。研究表明,该方法能有效实现道床病害的智能识别与分类,其中基于线性核函数的支持向量机模型识别效果最优,其平均识别率、召回率与F-值分别达92.5%、98.4%和91.3%。该方法可为铁路道床病害的智能诊断与早期预警提供技术支持。
Abstract:
To address the issues of low efficiency and significant deviations in traditional methods for detecting diseases in ballasted track beds, this paper proposed an intelligent recognition method for track bed diseases utilizing ground penetrating radar(GPR)based on support vector machine(SVM). By extracting gray-level co-occurrence matrix(GLCM)parameters—energy, entropy, contrast, and correlation—that reflect disease characteristics from GPR images of track beds, a feature database was constructed. SVM models with different kernel functions were then applied to classify three typical track bed diseases: mud pumping, abnormal moisture content, and settlement. Experimental results demonstrate the proposed method can effectively identify and classify track bed diseases, with the linear-kernel SVM model achieving optimal performance, with average recognition accuracy, recall, and F-score reaching 92.5%, 98.4%, and 91.3%, respectively. This method offers a robust technical foundation for intelligent recognition and early warning of track bed diseases.

参考文献/References:

[1] 井国庆. 铁路有砟道床[M]. 北京: 中国铁道出版社, 2012.
JING Guoqing. Railway Ballasted Track Bed[M]. Beijing: China Railway Publishing House, 2012.
[2] ZHANG Zhihai, XIAO Hong, MA Chaozhi, et al. Ballast Failures Restored and Detecting Methodology for Lateral Resistance of Ballast Bed during Dynamic Stabilization[J]. Engineering Failure Analysis, 2025, 176: 109665.
[3] CIAMPOLI L B, CALVI A, OLIVA E. “Test-site Operations for the Health Monitoring of Railway Ballast Using Ground-penetrating Radar” [J]. Transportation Research Procedia, 2020, 45: 763-770.
[4] ARTAGAN S S, BORECKY V. Advances in the Nondestructive Condition Assessment of Railway Ballast: a Focus on GPR[J]. NDT & E International, 2020, 115: 102290.
[5] 伊小娟, 罗威, 李伟, 等. 基于Mask R-CNN的地质雷达岩溶预报图像识别研究[J]. 高速铁路技术, 2024, 15(2): 50-55.
YI Xiaojuan, LUO Wei, LI Wei, et al. Study on Karst Forecast Image Recognition with Geological Radar Based on Mask R-CNN[J]. High Speed Railway Technology, 2024, 15(2): 50-55.
[6] GUO Yunlong, LIU Guixian, JING Guoqing, et al. Ballast Fouling Inspection and Quantification with Ground Penetrating Radar(GPR)[J]. International Journal of Rail Transportation, 2023, 11(2): 151-168.
[7] 邹华胜. 路基病害识别算法与应用研究[J]. 地球物理学进展, 2009, 24(6): 2302-2307.
ZOU Huasheng. A Study on Roadbed Disease Recognition Algorithm and Application[J]. Progress in Geophysics, 2009, 24(6): 2302-2307.
[8] 井国庆, 卜俊杰, 彭湛. 基于探地雷达单道信号的碎石道床病害智能识别[J]. 铁道学报, 2025, 47(6): 170-178.
JING Guoqing, BU Junjie, PENG Zhan. Intelligent Recognition of Ballast Track Bed Defects Based on A-Scan GPR Data[J]. Journal of the China Railway Society, 2025, 47(6): 170-178.
[9] 王先义, 赵欢, 黄华东. 地质雷达探测原理及在工程中的应用[J]. 安徽建筑, 2013, 20(3): 108-110.
WANG Xianyi, ZHAO Huan, HUANG Huadong. Mechanism of Geological Radar Exploration and Its Application in Engineering[J]. Anhui Architecture, 2013, 20(3): 108-110.
[10] SONG Weimin, HUANG Baoshan, SHU Xiang, et al. Interaction between Railroad Ballast and Sleeper: a DEM-FEM Approach[J]. International Journal of Geomechanics, 2019, 19(5): 04019030.
[11] 刘欢, 王石磊, 杨维一, 等. 探地雷达在有砟铁路路基检测中的研究综述[J]. 铁道学报, 2025, 47(11): 166-175.
LIU Huan, WANG Shilei, YANG Weiyi, et al. A Review of Ground Penetrating Radar Research in Ballasted Railway Subgrade Detection[J]. Journal of the China Railway Society, 2025, 47(11): 166-175.
[12] 刘伟, 周超, 金俊俊. 不同地质雷达设备在铁路工程隐伏岩溶探测中的应用与对比[J]. 高速铁路技术, 2024, 15(4): 82-86.
LIU Wei, ZHOU Chao, JIN Junjun. Application and Comparison of Different Ground Penetrating Radar Systems in the Detection of Concealed Karst in Railway Engineering[J]. High Speed Railway Technology, 2024, 15(4): 82-86.
[13] 王石磊, 赵志刚, 刘宏江, 等. 铁路碎石道床雷达检测技术研究与应用进展[J]. 铁道学报, 2023, 45(2): 116-126.
WANG Shilei, ZHAO Zhigang, LIU Hongjiang, et al. Research and Application Progress of Radar Detection Technology for Railway Ballast Bed[J]. Journal of the China Railway Society, 2023, 45(2): 116-126.
[14] 刘浩. 基于振动原理及改进SVM的管道漏失检测分析[D]. 武汉: 华中科技大学, 2020.
LIU Hao. Analysis of Pipeline Leakage Detection Based on Vibration Principle and Improved SVM[D]. Wuhan: Huazhong University of Science and Technology, 2020.
[15] 廖立坚, 杨新安, 丁春林. 铁路路基雷达探测图像的自动解释技术研究[J]. 土木工程学报, 2009, 42(6): 102-107.
LIAO Lijian, YANG Xinan, DING Chunlin. Technique for Automatic Interpretation of GPR Images of Railroad Subgrades[J]. China Civil Engineering Journal, 2009, 42(6): 102-107.
[16] 杨晓美, 仰圣刚. 基于灰度共生矩阵的沥青路面病害雷达图像纹理特征提取研究[J]. 交通节能与环保, 2023, 19(4): 187-191.
YANG Xiaomei, YANG Shenggang. Research on Texture Feature Extraction of Asphalt Pavement Disease Radar Image Based on Grayscale Symbiosis Matrix[J]. Transport Energy Conservation & Environmental Protection, 2023, 19(4): 187-191.
[17] 王石磊, 彭湛, 齐法琳, 等. 碎石道床状态雷达测试适宜参数研究[J]. 铁路技术创新, 2021(6): 50-55.
WANG Shilei, PENG Zhan, QI Falin, et al. Study on Appropriate Parameters for Radar Test of Ballast Bed State[J]. Railway Technical Innovation, 2021(6): 50-55.
[18] FENG Deshan, LIU Yuxin, WANG Xun, et al. Efficient Source-independent Full Waveform Dual-parameter Inversion of Common-offset GPR Data[J]. Journal of Applied Geophysics, 2023, 208: 104891.

备注/Memo

备注/Memo:
收稿日期:2025-11-30
作者简介:李元军(1975-),男,高级工程师。
基金项目:中国铁路太原局集团有限公司科技开发计划课题(2024G08)
更新日期/Last Update: 2025-12-30