基于自适应连续小波模极大值算法的车轮擦伤定量评估

Evaluation of railway wheel-flat via adaptive continuous wavelet transform modulus maximum algorithm

为实现高速动车组车轮踏面擦伤长度定量评估,提出了一种自适应连续小波模极大值算法,应用到基于轴箱振动加速度信号的车轮擦伤状态识别与损伤评估。利用连续小波变换系数模极大值对自适应滤波后的轴箱振动加速度信号突变点进行定位,从而建立擦伤车轮脱离钢轨腾空运行时间与擦伤长度及其影响参数的关系模型。仿真实例结果表明,该方法可在车辆高速运行时实现车轮擦伤程度的定量识别。与其它文献识别结果对比分析表明,该方法原理简单、计算结果精度高。

In order to evaluate quantitatively wheel-flat length of high-speed EMUs,an adaptive continuous wavelet modulus maximum algorithm was proposed,applying to detect wheel-flat based on axle box vibration acceleration.The continuous wavelet modulus maximum was used to detect abrupt change points of adaptive filtered axle box acceleration signals,and the relationship between wheel-rail detachment time and wheel-flat length and its influence parameters was established.The simulation results indicate that this approach is effective to identify and quantitatively assess the state of railway wheel-flat at high speed operation.Compared with traditional methods,this algorithm has brief principle and accurate results.