Fault diagnosis of rolling bearings based on time-frequency plots and convolutional neural networks

Fault diagnosis technology overly relies on manual feature extraction. This restricts the use of technology in bearing fault diagnosis. In order to solve this problem, a fault diagnosis method of rolling bearing based on convolutional neural network (CNN) and time-frequency graph is put out in this work. Firstly, the time-frequency image in frequency domain is generated by continuous wavelet transform(CWT) of the time domain vibration signal collected under multiple working conditions, and a labeled training dataset and validation set are constructed. Secondly, the proposed CNN network model is used to learn spatial features of two-dimensional signals. Finally, use Softmax classifier for fault recognition. Through the validation of the model method using the rolling bearing vibration signal dataset publicly available at Case Western Reserve University Bearing Experimental Data Center, compared to other methods, this method can reflect the fault information of rolling bearing more accurately.