基于全矢谱和径向基概率神经网络的旋转机械故障诊断方法研究  

Fault diagnosis on rotary machinery based on vector spectrum and radial basis probabilistic neural networks

作　　者：杨春燕[1] 丁静[1] 

机构地区：[1]郑州轻工业学院机电学院,郑州450002

出　　处：《现代制造工程》2010年第1期141-144,共4页Modern Manufacturing Engineering

基　　金：河南省教育厅自然科学研究资助计划项目(2009A460013)

摘　　要：结合全矢谱和径向基概率神经网络的优点,提出一种故障诊断的新方法,该方法是以提取全矢幅值谱的特征输入到径向基概率神经网络分类器进行故障识别。试验结果表明,该方法与传统单通道相比故障正确识别率很高,把它应用于旋转机械故障诊断是有效的。Combining full vector spectrum and RBPNN, a new fault diagnosis approach is proposed, this approach is that the full vector spectrum is used as eigenvectors, RBPNN as a classifier. The experiment result shows that the method has high correct recognition rate comparing with the traditional single-channel, and the proposed approach is very effective applying to the fault diagnosis of rotating machinery.

关 键 词：全矢谱 径向基概率神经网络 故障诊断 旋转机械 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术]