Bearing Fault Classification Using Multi-Scale Dilated Convolution and Temporal Convolutional Networks

Rolling bearings play a critical role in fault diagnosis to ensure the stable operation of equipment. While existing fault diagnosis methods, particularly those based on deep learning, have achieved success in feature extraction, they continue to face challenges related to model complexity and training efficiency. This paper proposes a bearing fault diagnosis method that integrates multi-scale dilated convolutional residual blocks, the convolutional block attention module (CBAM), and the temporal convolutional network (TCN). By expanding the receptive field with dilated convolutions and enhancing feature selectivity with CBAM, the method captures temporal dependencies using TCN, leading to improved fault classification. Experiments on the Case Western Reserve University (CWRU) and Paderborn University (PU) datasets show that the proposed method achieves high accuracy in multi-class fault diagnosis.