Application of Neural Network Based on Transformer Model in Rolling Bearing Fault Diagnosis

Fault diagnosis of rolling bearings is one of the important tasks in the industry. Rolling bearing fault diagnosis using neural networks has become a common method in recent years. In this study, a rolling bearing fault diagnosis method based on the Transformer model is constructed, which mainly uses an encoder to encode the position of the input signals and identify the multi-head self-attention mechanism to diagnose the faults of rolling bearings. The model was trained using a publicly available bearing dataset from Case Western Reserve University. The trained model was then used to process the actual data collected by the fault diagnosis laboratory of Inner Mongolia University of Science and Technology. After a series of experiments, the model was verified to perform well in the rolling bearing fault diagnosis task, effectively identifying and extracting key features and accurately identifying the fault state of the bearing, which can provide a reliable solution for fault detection and prevention in industrial production. This study lays the foundation for further research and application in the field of rolling bearing fault diagnosis, which is of great practical significance.