Due to the fact that existing methods do not consider the nonlinear and non-stationary characteristics of the fault signals of paper machine press bearings,the accuracy of fault diagnosis is not high.Therefore,in order to effectively ensure the safe operation of paper machines,a transfer learning based fault diagnosis method for paper machine press bearings is proposed.This method first collects the vibration signal of the paper machine press bearing fault based on the signal acquisition system,and implements noise removal processing on the collected bearing fault vibration signal through local projection method;Based on the denoising results of the fault vibration signal,transfer learning method is used,combined with convolutional neural network and artificial neural network,to establish a diagnostic system for paper machine press bearing fault diagnosis;Finally,by establishing a system to extract fault signal features and complete fault classification,precise diagnosis of paper machine press bearing faults can be achieved.The experimental results show that using the above method for fault diagnosis of paper machine press bearings has good diagnostic effect and high accuracy.
关键词
迁移学习/造纸机/压榨轴承/故障诊断方法/故障信号去噪
Key words
transfer learning/paper machine/pressing bearings/fault diagnosis methods/fault signal denoising
维普
万方数据