Abstract: Dynamics data of the train contains a wealth of information on track irregularities, wheel-rail relationships and so on. Aiming at the high noise component and nonlinear characteristics of the data, an improved Hilbert-Huang analysis method for extracting the transient characteristics of non-stationary signals is proposed. The method draws on the idea of fast band-pass filtering, and obtains the intrinsic mode function through the adaptive band-pass filtering algorithm. It can detect the time frequency, amplitude and other information of transient components in the signal while reducing noise. By using the improved method to process the signal, modal aliasing can be avoided. The research result of the dynamic data of the sample section of the railway has different degrees of short-wave and medium-long-wave irregularities. It shows that the improved Hilbert-Huang transform method can provide a new way to ensure the safe operation of railways.

Key words:  Track irregularity, Hilbert-Huang Transform, time-frequency analysis 

摘要： 轨道车辆系统动力学数据包含轨道不平顺、轮轨关系等多方面的丰富信息。针对该类数据中噪声成分高、非线性特征明显，提出了提取非平稳信号中瞬态特征的信号的改进Hilbert-Huang分析方法。该方法借鉴了快速带通滤波思想，通过自适应带通滤波算法求得固有模态函数，能在降噪的同时检测出信号中瞬态成分的时间频率、幅值等信息。利用该改进方法对模拟信号进行处理，可以避免模态混叠现象。对某客运专线铁路实测轨道车辆动力学数据进行研究的结果表明：该铁路样本段轨道不平顺存在着不同程度的短波和中长波不平顺。说明改进的Hilbert-Huang变换方法可以为保障铁路安全运营提供一种新的途径。

关键词: 轨道不平顺, Hilbert-Huang Transform, 时频分析