基于切片谱RBF神经网络的旋转机械故障诊断  

Diagonal Spectrum-RBF Neural Networks Fault Diagnosis for Rotating Machinery

作　　者：周鹏[1] 秦树人[1] 

机构地区：[1]重庆大学,重庆400044

出　　处：《中国机械工程》2008年第12期1488-1491,共4页China Mechanical Engineering

基　　金：国家自然科学基金资助项目(50605065)

摘　　要：结合切片谱和RBF神经网络提出一种切片谱RBF神经网络旋转机械故障诊断方法。该方法以信号的切片谱为故障特征量,以RBF神经网络为分类器,进行旋转机械故障诊断。轴承故障的诊断结果表明,该方法是可行和有效的。A method of fault diagnosis for rotating machinery, combing diagonal spectrum with RBF（radial basis function）neural networks, was presented. In our applications, diagonal spectrum, as input feature vector, which was calculated from vibration rotating machinery, was classified by RBF neural networks, and finally determined fault modes of rotating machinery. The experiment of rolling element bearing shows that the approach based on diagonal spectrum - RBF neural networks is practical and effective.

关 键 词：故障诊断 切片谱 RBF神经网络 轴承 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]