基于特征指标与模式识别的轨道层间脱空研究  

Study on Ballastless Track Interlayer Voiding Based on Feature Indicators and Pattern Recognition

作　　者：李文秀 黄浩[1] 赵坪锐[1] LI Wenxiu;HUANG Hao;ZHAO Pingrui(MOE Key Laboratory of High-speed Railway Engineering,Southwest Jiaotong University,Chengdu 610031,China)

机构地区：[1]西南交通大学高速铁路线路工程教育部重点实验室,成都610031

出　　处：《高速铁路技术》2025年第2期22-28,共7页High Speed Railway Technology

基　　金：国家重点研发计划资助(2021YFF0502100);高速铁路无砟轨道设计与维护四川省青年科技创新研究团队项目(2022JDTD0015)。

摘　　要：为评价无砟轨道层间的脱空程度,采用数值模拟方法建立无砟轨道动力学模型,设置不同程度脱空损伤的工况,提取移动列车荷载作用下无砟轨道各部位的振动响应;利用数理统计理论及信号处理手段,从时域、频域、时频域、相关分析、振动传递特性5个维度,建立能综合表征损伤程度的高敏感性特征指标;基于以上理论、技术,运用主成分分析模式识别技术,实现对无砟轨道层间脱空损伤程度的评价,并通过实例验证了该方法的有效性和正确性。结果表明,实例中50%脱空程度得出的加权得分值介于40%、60%脱空程度之间,提出的无砟轨道层间脱空损伤识别方法准确度较高。To assess the extent of interlayer voiding in ballastless track systems,a dynamic model of the ballastless track was established using numerical simulation methods,incorporating scenarios with varying degrees of interlayer voiding damage.Vibration responses at different locations of the track under moving train loads were extracted.Employing mathematical statistics theory and signal processing techniques,high-sensitivity feature indicators that comprehensively represent the damage severity were developed across five dimensions:time domain,frequency domain,time-frequency domain,correlation analysis,and vibration transmission characteristics.Utilizing these theoretical foundations and techniques,principal component analysis-based pattern recognition was employed to evaluate the degree of interlayer voiding damage in ballastless tracks.The validity and accuracy of this method were substantiated through a case study.Results show that the weighted score derived from a 50%voiding scenario falls within the range of scores obtained for 40%and 60%,indicating the proposed method’s high accuracy for identifying interlayer voiding damage in ballastless tracks.

关 键 词：无砟轨道 层间脱空 损伤识别 主成分分析 特征提取 

分 类 号：U213.244[交通运输工程—道路与铁道工程]