摘要
利用小波分析和BP神经网络相结合的方法对旋转机械的故障进行识别。首先运用小波分析对故障信号进行降噪处理,然后运用小波包对信号进行分解和重构,提取各频带能量值,将该能量值作为BP神经网络输入端的特征向量,训练网络进行故障模式识别。实验表明,该方法在旋转机械故障诊断中切实可行。
The wavelet analysis and BP neural network are applied to the fault diagnosis of the rotating machinery.First,use the wavelet analysis to reduce the noise effectively for fault signals processing,then adopts the wavelet packets to decompose and reconstruct the signal,extract the energy of the different frequency bands,take the energy as the characteristic vector of the BP neural input layer and train the network.The various faults can be distinguished by the trained network.Test shows that this method is efficiently in the fault diagnosis of rotating machinery.
出处
《电子机械工程》
2011年第1期56-59,共4页
Electro-Mechanical Engineering
关键词
旋转机械
故障诊断
小波分析
BP神经网络
rotating machinery
fault diagnosis
wavelet analysis
BP neural network
