首页|采用轮轨振动加速度信号对车轮扁疤的识别研究

采用轮轨振动加速度信号对车轮扁疤的识别研究

扫码查看
实现车轮扁疤损伤的早期识别,并跟踪其发展过程,是研发轨道交通故障智能识别系统的重要内容.目前相关研究着重于分析车轮扁疤激励对车辆和轨道系统相关部件的影响,以及如何在确保行车安全的条件下确定扁疤的安全限值,缺乏通过钢轨振动响应识别车轮扁疤机理的研究.因此基于车辆-轨道耦合动力学理论,通过动力学仿真计算得到车轮扁疤激励下,车辆系统和轨道系统的动力学响应.采用小波包分解算法,提取不同部件的垂向振动加速度的能量值作为评判指标,从能量特征的角度研究车轮扁疤冲击响应的评价方法.考虑了轮轨之间平顺、随机不平顺、以及叠加粗糙度 3 种激励状态,对车体、构架、轴箱、轮对以及钢轨的加速度时域信号特征和能量特征进行了对比分析.研究发现钢轨小波包能量值与扁疤深度之间具有线性递增关系,车轮扁疤冲击仅影响同侧钢轨的小波包能量值,因此可以利用钢轨振动响应的小波包能量值进行车轮扁疤检测.
Study on Recognition of Wheel Flat Scars Using Acceleration Signal of Wheel-rail Vibration
Early detecting of wheel flat and tracking development process are important contents for the rail transit fault intelligent identification system.Current work emphasizes the impact of wheel flat excitation on vehicle and track system components,and how to determine the safety limit of the scar to ensure driving safety,lack of research on the recognition mechanism of wheel flat.Therefore,in terms of the vehicle-track coupled dynamics theory,the dynamic response of vehicle system and track system under wheel flat excitation is obtained by using the dynamic simulation.Using wavelet packet decomposition,the energy values of vertical vibration acceleration of different components are extracted as the evaluation indexes,and the evaluation method of wheel flat impact response is studied from the perspective of energy characteristics.The acceleration time domain the signal eigenvalue and energy eigenvalue of the vehicle body,frame,axle box,wheelset and rail are compared and analyzed by considering three excitation states of wheel-rail irregularity,random irregularity and superimposed roughness.It is found that there is a linear increasing relationship between the wavelet packet energy of the rail and the flat depth,the wheel flat only affects the wavelet packet energy of identical side rail.Therefore,the wavelet packet energy of the rail vibration response can be used to detect the wheel flat.

intelligent recognitionwavelet packet decompositionvehicle-track coupling dynamicsflat detection

杨丽蓉、和振兴、王开云、刘旭麒、曹子勇

展开 >

兰州交通大学机电工程学院,兰州 730070

西南交通大学牵引动力国家重点实验室,成都 610031

智能识别 小波包分解 车辆-轨道耦合动力学 扁疤检测

国家自然科学基金牵引动力国家重点实验室开放课题甘肃省科技计划项目

52162047TPL190220JR5RA393

2024

机械科学与技术
西北工业大学

机械科学与技术

CSTPCD北大核心
影响因子:0.565
ISSN:1003-8728
年,卷(期):2024.43(2)
  • 16