基于空间谱的钢轨波磨识别技术研究  

Research on Rail Corrugation Identification Technology Based on Spatial Spectrum

作　　者：孙效杰 赵宇伦 易广记 罗进 陈迪来 SUN Xiaojie;ZHAO Yulun;YI Guangji;LUO Jin;CHEN Dilai(School of Railway Transportation,Shanghai Institute of Technology,Shanghai 201418,China)

机构地区：[1]上海应用技术大学轨道交通学院,上海201418

出　　处：《铁道标准设计》2024年第12期31-36,共6页Railway Standard Design

基　　金：上海市科技创新行动计划基金项目(21210750300);上海市科委启明星计划项目(22YF1447600)。

摘　　要：伴随轨道车辆运行里程的增加和运行速度的提升,轮轨间的异常磨耗也逐渐凸显,为及时了解钢轨表面波浪形磨耗的存在现状,对波磨的波长及波深的识别研究成为一类重要的研究方向。目前列车变速下的波磨波长识别方法有所欠缺,基于加速度二次积分的波磨波深识别方法由于零点漂移问题也存在计算结果不够准确的问题。使用空间谱代替时间谱采集和处理列车轴箱加速度数据,对波磨波长计算公式进行改进,避免速度变化对波长识别带来的影响;将采集和计算得到的数据整合成数据集并进行自然邻点插值,得到完整的波长-波深-幅值数据库,对波磨波深进行量化识别;同时提出完整的钢轨波磨故障识别流程,为实际工程应用提供参考。研究结果表明,基于空间谱的傅里叶变换能够精确获取波磨分布的异常频率,并通过改进后的波长计算公式得到波磨波长,在不同速度段变化下仿真得到的识别误差低于5%,实现了变速条件下的准确识别;使用自然邻点插值方法处理数据契合波磨相关数据的分布特性,插值结果遵循列车轴距滤波规律,基于该方法的波磨波深识别数据误差小于7%。上述方法实现了不同复杂运行工况下的波磨波长与波深的准确识别,可用于指导线路上钢轨异常磨耗的打磨工作。With the increase in the mileage and speed of railway vehicles,abnormal wear between the wheel and rail has become increasingly prominent.To understand the current status of wave-shaped wear on rail surfaces in a timely manner,identifying the wavelength and depth of rail corrugation has become an important research focus.Current methods for identifying corrugation wavelengths under variable train speed are limited,and the wave depth identification method based on quadratic integration of acceleration also faces inaccuracies due to zero-drift issues.This study proposed the use of spatial spectrum instead of time spectrum to collect and process axle box acceleration data,modifying the corrugation wavelength calculation formula to eliminate the impact of speed changes on wavelength recognition.The collected and calculated data were integrated into a dataset and processed with natural neighbor interpolation to obtain a complete wavelength-depth-amplitude database,allowing for quantitative identification of corrugation wave depth.A comprehensive rail corrugation fault identification procedure was also proposed,providing a reference for practical engineering applications.The results showed that Fourier transform based on spatial spectrum could accurately identify the abnormal frequencies of corrugation distribution,and the modified wavelength calculation formula could effectively determine the corrugation wavelength,achieving an identification error of less than 5%under variable speed conditions in simulations.Using natural neighbor interpolation aligned well with the distribution characteristics of rail corrugation-related data,and the interpolation result followed the train axle distance filtering rule,yielding a wave depth identification error below 7%with this method.The proposed method achieves accurate identification of rail corrugation wavelength and depth under various complex operating conditions,providing guidance for grinding operations to address abnormal rail corrugation along the track.

关 键 词：钢轨 波磨识别 空间谱 自然邻点插值 轴箱振动 变速行驶 

分 类 号：U270.33[机械工程—车辆工程] U213.2[交通运输工程—载运工具运用工程]