摘要
高速铁路曲线地段轮轨之间的动力相互作用复杂,智能识别算法在曲线地段的适应性是实现对车轮扁疤全线不间断识别跟踪的前提。考虑高速铁路车轮扁疤信号的随机性特征,提出了一种基于变分模态分解(variational modal decomposition, VMD)的高速铁路车轮扁疤识别方法。考虑曲线线路不同地段以及不同轨侧的影响,对动力学计算得到的轮轨力随机响应进行变分模态分解并将信号重构,通过包络谱特征识别车轮扁疤冲击频率和对应扁疤的长度。研究表明:在曲线地段正常车轮与扁疤车轮对应的包络谱之间存在显著差异,包络谱中10 mm以上扁疤冲击倍频特征明显;倍频峰值频率特征与列车速度对应扁疤冲击频率一致;前4阶包络谱均值与扁疤长度之间呈线性关系,可以通过包络谱均值直接识别扁疤的长度。对曲线外轨侧识别的车轮扁疤长度进行修正后,可以实现车轮扁疤全线不间断识别跟踪。
Dynamic interaction between wheel and rail in curve section of high-speed railway is complex. The adaptability of intelligent recognition algorithm in curve section is the premise to realize continuously recognizing and tracking wheel flat along whole line. Here, considering the randomness of high-speed railway wheel flat signal, a method of high-speed railway wheel flat recognition based on variational mode decomposition(VMD) was proposed. Considering effects of different sections of curve line and different rail sides, VMD was performed for wheel/rail force random response obtained with dynamic calculation and the signal was reconstructed. Wheel flat impact frequency and corresponding flat length were identified with envelope spectrum features. The results showed that there is a significant difference between normal wheel envelope spectrum and flat wheel envelope spectrum in curve section, and impact frequency doubling features of flat with length >10 mm in envelope spectrum are obvious;frequency doubling peak value frequency features are consistent to those of flat impact frequency corresponding to train speed;mean values of the first 4 order envelope spectra are linearly related to flat length, so flat length can be directly identified with mean value of envelope spectrum;after the identified wheel flat length on curve outer rail side is modified, continuously recognizing and tracking wheel flat can be realized along the whole line.
作者
刘旭麒
和振兴
杨丽蓉
LIU Xuqi;HE Zhenxing;YANG Lirong(School of Mechanical Engineering,Lanzhou Jiaotong University,Lanzhou 730070,China)
出处
《振动与冲击》
EI
CSCD
北大核心
2023年第5期223-232,共10页
Journal of Vibration and Shock
基金
国家自然科学基金项目(52162047
52062028)
牵引动力国家重点实验室开放课题:高速铁路减振轨道关键动力学参数研究(TPL1902)
甘肃省科技计划项目(20JR5RA393)。
