Fault Diagnosis Method with Improved Space-Channel Attention Mechanism Based on BiGRU

Most of the existing fault diagnosis methods based on deep learning rely on a large number of data samples, and there are often fewer training samples available in real industrial production. Aiming at this problem, this paper proposes a method with Improved Space-Channel Attention Mechanism (ISCAM) and combines Laplace wavelet convolution and Bidirectional Gated Recurrent Unit (BiGRU) neural network to diagnose and classify three common types of rolling motor bearing faults. Laplace wavelet convolution can convert the original signal from the time domain to the frequency domain, and more effectively identify the bearing state. BiGRU can get the hidden information of different time sequence position. Experiments show that the neural network has a good classification performance for bearing faults.