Convolutional Neural Network Bearing Fault Diagnosis Method Based on Time Frequency Domain Information Fusion

Currently, in bearing fault diagnosis, only the timedomain characteristics of the original signal are usually considered, while the frequency-domain characteristics of the signal are ignored, resulting in low data utilization and insufficient extraction of fault information. In response to this issue, this paper proposes a convolutional neural network fault diagnosis method that simultaneously considers the time-frequency domain characteristics of the signal. This method uses principal component analysis (PCA) to reduce the dimensionality of the original time-domain signal, and extracts the frequency domain features of the vibration signal using the WFEE algorithm. The extracted time-frequency domain features are simultaneously input into the convolutional neural network for learning and classification of fault types. The validation was conducted on the publicly available rolling bearing dataset from Case Western Reserve University, and the results showed that the proposed method has higher prediction accuracy than networks that only consider time-domain or frequency-domain features.