基于小波包ICA的齿轮箱信号分离系统*

作者：叶红仙,项文娟,刘仲宇,周晓峰,杨世锡 日期：2008-04-23/span> 浏览：3872 查看PDF文档

基于小波包ICA的齿轮箱信号分离系统*
叶红仙,项文娟,刘仲宇,周晓峰,杨世锡
(浙江大学 机械与能源工程学院,浙江 杭州 310027)

摘 要：由于齿轮箱振动信号混叠、信噪比低，给早期的故障诊断造成了一定的困难。结合小波包和独立分量分析(ICA)在信号处理中特征提取的方法，以LabVIEW和Matlab语言为开发工具，设计了齿轮箱故障源信号分离系统，并以真实信号进行了实验验证。实验结果表明，该系统可以很好地实现故障源信号的分离，有利于后续的故障诊断。
关键词：信号分离；小波包；独立分量分析；齿轮箱

Signal separation system of gearbox based on waveletpacket and ICA
YE Hongxian, XIANG Wenjuan, LIU Zhongyu, ZHOU Xiaofeng, YANG Shixi
（College of Mechanical and Energy Engineering, Zhejiang University, Hangzhou 310027, China)

Abstract: Owing to commingle of vibration sources signals of gearbox and low signaltonoise, some difficulty is made in early fault diagnosis. The system was designed by LabVIEW and Matlab for the separation of fault sources signals of gearbox. The algorithms of inndependent component analysis (ICA) and wavelet packet were used in the system. The result of experiment shows that independent signal is separated by the system which is good for the source identification and fault diagnosis.
Key words: signal separation; wavelet package; independent component analysis (ICA); gearbox