ATT-Based Multi-Scale One-Dimensional CNN Laser Cleaning Equipment Motor Fault Diagnosis Technology Research

Since the motor is used as one of the essential mechanical devices in the laser cleaning equipment, its failure will seriously affect the normal operation of the equipment, resulting in certain economic losses, and it is difficult to repair after the accident. This paper proposes a multi-scale one-dimensional convolutional neural network fault diagnosis method based on the attention mechanism (ATT). The attention mechanism is introduced into the multi-scale feature fusion and screening module, which automatically assigns different feature weights to the parallel connected. features, thus distinguishing the importance of different features, strengthening the important features, weakening the redundant features, thus optimizing the learning mechanism of the CNN, and ultimately sending the new features into the classifier. Finally, it is validated against the multiscale one-dimensional convolutional neural network method without setting the attention mechanism as well as the single-scale one-dimensional convolutional neural network, which proves the superiority of the method.