改进HHT方法在轨道不平顺信号分析中的应用  

Analysis of track irregularity signal by improved Hilbert-Huang transform

作　　者：李再帏[1] 练松良[1] 

机构地区：[1]同济大学道路与交通工程教育部重点实验室,上海201804

出　　处：《深圳大学学报（理工版）》2012年第3期270-275,共6页Journal of Shenzhen University(Science and Engineering)

基　　金：国家自然科学基金资助项目(50878158);铁路环境振动与噪声教育部工程研究中心开放研究资助项目(20091001)~~

摘　　要：针对希尔伯特-黄变换(Hilbert-Huang transform,HHT)中经验模态分解法(empirical modedecomposition,EMD)的端点效应问题,提出一种镜像延拓和灰色神经网络预测法相结合的改进方法,即先对信号进行灰色神经网络延拓,再利用镜像延拓法对数据进行延拓.分析仿真信号和实际轨道不平顺信号认为,该方法可以有效抑制EMD的端点效应.利用改进HHT方法对轨道轨距和水平不平顺信号进行研究,结果表明,该方法可有效去除轨道轨距和水平不平顺中低频趋势项,提高轨道不平顺信号的精度,有利于轨道质量的评价.To solve the problem of end effects of empirical mode decomposition （EMD） in Hilbert-Huang trans- form, a method combining gray neural network with mirror extension as to the ends of data is presented, in which the original time series was extended by means of gray neural network, and subsequently, by mirror extension method. The comparative results of the simulated signal and the actual detected track irregularity signal indicated that the improved algorithm restrained the end effect effectively. The method was applied to analyzing the track irregularity signal. Results show that by removing the trend of track gauge and cross level irregularity, this method can improve the accuracy of track irregularity, which is beneficial to track quality evaluation.

关 键 词：铁路轨道 信号处理 神经网络 希尔伯特-黄变换 经验模态分解法 端点效应 不平顺 

分 类 号：U216.3[交通运输工程—道路与铁道工程]