基于改进TDA与WPA的地铁齿轮箱振动信号消噪研究  

Research of Vibration Noise Elimination Based on Improved TDA and WPA for Subway Gearbox

作　　者：白永亮[1] 杨建伟[1] 

机构地区：[1]北京建筑大学机电与车辆工程学院,北京100044

出　　处：《北京建筑大学学报》2016年第2期51-55,共5页Journal of Beijing University of Civil Engineering and Architecture

基　　金：国家自然科学基金项目(51175028);轨道列车走行部轴承故障机理和辨识算法研究项目(00331615015);城市轨道交通车辆服役性能与保障北京市重点实验室项目(06080915001)

摘　　要：为提高齿轮箱振动信号的消噪水平,提出一种基于改进的时域平均(Time Domain Averaging,TDA)与小波包分析(Wavelet Packet Analysis,WPA)的信号消噪算法.首先引入零工况(主要运动部件静止但信号采集系统正常工作)下的参照信号取代经典TDA中的随机信号,然后将改进的TDA与经典WPA消噪理论中的DB4小波结合,从而构成组合消噪算法,最后将该算法应用于地铁齿轮箱故障轴承的振动信号处理中,取得了良好的信号消噪效果.In order to improve the level of vibration noise elimination in subway gearbox,a new method of signal noise elimination based on the improved TDA（ Time Domain Averaging） and WPA（ Wavelet Packet Analysis） was introduced. Firstly,the zero-condition signal（ main moving parts keep still while signal acquisition system works as usual） was used to replace the random signal in classic TDA,and then the improved TDA and the DB4 wavelet transformation in classic WPA were combined to constitute a combination denoising algorithm. Finally,the algorithm was applied to deal with the vibration signals collected from failed bearing in subway gearbox and a good performance was acquired.

关 键 词：齿轮箱 信号消噪 时域平均 小波包分析 

分 类 号：TH165.3[机械工程—机械制造及自动化] TH132.41