基于区间的小波包振动信号特征提取方法研究  

Wavelet Packet Method Based on Sections of Picking-up Characteristic from Vibration Signals

作　　者：黄强[1] 刘永长[1] 叶晓明[1] 

机构地区：[1]华中科技大学,湖北武汉430074

出　　处：《车用发动机》2003年第4期49-51,共3页Vehicle Engine

摘　　要：基于小波分析与神经网络的振动信号故障诊断方法,提出采用基于区间小波包分解方法来提取振动信号的特征向量来取代传统的小波包分解方法,并以295柴油机进排气系统故障诊断为例验证采用该方法的有效性,结果表明,基于区间的小波包特征提取方法较传统方法能大大提高进排气系统的故障识别率。In the fault diagnosis based on Wavelet Analysis and Neural Networks, the wavelet packet method based on sections is used to replace the traditional one to pick-up characteristics of vibration signals.Then the efficiency of this method is validated on the inlet and exhaust systems fault diagnosis of 295 Diesel Engine.The result shows that wavelet packet method based on sections of picking-up characteristic can improve the veracity of fault diagnosis.

关 键 词：柴油机 小波包 神经网络 特征提取 

分 类 号：TK402[动力工程及工程热物理—动力机械及工程]