基于多源特征融合的无砟轨道砂浆层脱空病害检测方法  

Method of ballastless track mortar layer void disease detection based on multi-source feature fusion

作　　者：张广远 王哲 王保宪 李义强 赵维刚 ZHANG Guangyuan;WANG Zhe;WANG Baoxian;Li Yiqiang;ZHAO Weigang(Institute of Large-scale Structure Health Monitoring and Control,Shijiazhuang Tiedao University,Shijiazhuang 050043,China;School of Electrical and Electronic Engineering,Shijiazhuang Tiedao University,Shijiazhuang 050043,China;Heibei Key Laboratory for Large-scale Structure Health Monitoring and Control,Shijiazhuang 050043,China)

机构地区：[1]石家庄铁道大学大型结构健康诊断与控制研究所,河北石家庄050043 [2]石家庄铁道大学电气与电子工程学院,河北石家庄050043 [3]河北省大型结构健康诊断与控制重点实验室,河北石家庄050043

出　　处：《现代电子技术》2020年第22期62-66,共5页Modern Electronics Technique

基　　金：国家重点研发计划(2016YFC0802207);国家自然科学基金(51978423);河北省高等学校科学技术研究项目青年基金(QN2016221)。

摘　　要：高速铁路轨下多层结构承载着高速列车通行,因此其健康状态直接影响到列车运营的安全性。利用弹性波、探地雷达等传感器提取的轨道病害特征可有效地实现轨道结构病害检测,但单一特征不能全面地对无砟轨道病害进行描述,导致部分病害不能被检测,从而影响无砟轨道的病害检测精度。基于此,提出一种基于多源特征融合的无砟轨道砂浆层脱空病害检测方法。该方法将弹性波与探地雷达两种特征进行量化分析,并将量化结果利用特征堆栈的方式实现特征融合,最后利用支持向量机完成病害融合特征的分类识别。在无砟轨道实体结构上采集大量的脱空病害数据并测试该文方法,实验结果验证了该文算法对脱空病害检测的有效性。Multi-layer structure under the track of high-speed railway bears up the high-speed train,so its health directly affects the safety of train operation.The track defect features extracted by elastic wave and ground penetrating radar sensors can effectively detect the track structure diseases,but the single feature cannot describe the ballastless track damage comprehensively,which results in the some diseases cannot be detected,and thus influences the accuracy of the diseases detection of the ballastless track.On this basis,a method of ballastless track mortar layer void disease detection based on multisource feature fusion is proposed.In this method,the two features of the elastic wave and ground penetrating radar are analyzed quantitatively,the feature fusion of the quantification results is carried out by means of the feature stack,and the classification and identification of the disease fusion feature are completed by means of the support vector machine.A large number of void disease data are collected from the solid structure of ballastless track and the method in this paper is tested.The experimental results have verified the effectiveness of the proposed algorithm for void disease detection.

关 键 词：无砟轨道 病害监测 特征融合 量化分析 特征堆栈 分类识别 

分 类 号：TN911.23-34[电子电信—通信与信息系统]