声发射-小波包-BP网络在振动筛故障诊断中的应用  

Application of acoustic emission, wavelet packet and BP network to fault diagnosis of vibrating screens

作　　者：燕碧娟[1] 李昕涛[1] 

机构地区：[1]太原科技大学机械工程学院,山西太原030024

出　　处：《矿山机械》2013年第12期92-95,共4页Mining & Processing Equipment

基　　金：山西省回国留学人员科研资助项目(2012-073);山西省科技攻关项目(20110321005-08);太原科技大学博士启动基金项目(20132002)

摘　　要：以振动筛下横梁作为损伤结构识别的研究对象,对其进行实验室加载试验。利用声发射检测技术采集与部件疲劳损伤相关的声发射波形信号,并应用"小波包-能量"法提取其信号特征量,将其作为神经网络的输入向量。在MATLAB 6.5的环境下进行了神经网络识别计算,结果表明,将小波包-神经网络用于声发射信号处理及零件早期疲劳损伤的诊断是可行的。The lower crossbeam of the vibrating screen was regarded as the research object, and the loading test for it in laboratory was conducted. The acoustic emission technology was applied to collect the acoustically emitted waveform signals related to fatigue damage of the part, and the wavelet packet energy method was used to abstract the signal eigenvectors which served as the input vector of the neural network. Identification of the neural network model was conducted in MATLAB 6.5. The results showed that the application of wavelet packet and neural network to processing of acoustically emitted signals and diagnosis of initial fatigue damage of the part was feasible.

关 键 词：振动筛 故障诊断 声发射 小波包分析 BP网络 

分 类 号：TD452[矿业工程—矿山机电]