基于小波包分析和BP神经网络的滚动轴承非接触声发射诊断方法  

Non-Contacting Acoustic Emission Inspection Technique for Rolling Bearings Based on Wavelet Packet Analysis and BP Neural Network

作　　者：戴光[1] 余永增[1] 张颖[1] 于江林[1] 

机构地区：[1]大庆石油学院

出　　处：《化工机械》2008年第5期274-277,294,共5页Chemical Engineering & Machinery

基　　金：黑龙江省自然科学基金资助项目(E2007-02);黑龙江省教育厅科学技术研究资助项目(11521004)

摘　　要：采用声发射技术对滚动轴承进行非接触诊断,以小波包分析方法提取故障信号的能量特征向量,作为BP神经网络的输入向量进行模式识别,区别完好轴承和各类故障轴承。A non-contacting inspection was carried out for rolling bearings using acoustic emission technique. The energy characteristics vector of the fault signals extracted with the wavelet packet analysis method was taken as the input vector of BP neural network to identify the models for distinguishing the good bearings from the damaged bearings.

关 键 词：声发射 小波包分解 特征提取 BP神经网络 

分 类 号：TQ050.7[化学工程]