Fault Diagnosis of Rolling Bearing Based on Edge Calculation

In view of the problem that the traditional rolling bearing fault diagnosis method needs to extract the characteristics artificially, take a long time and have low diagnostic accuracy, this paper puts forward a bearing fault diagnosis method based on edge computing. Cloud data is employed to train the convolutional neural network in order to get the cloud model. The feature extraction layer of cloud model is migrated to edge end to get edge end fault diagnosis model. The experimental results show that when the number of fault samples is only 20, the fault diagnosis model at the edge end will have 96.4% accuracy, and the fault diagnosis time will be reduced by 5.96, which meets the high precision and real-time requirements of rolling bearing fault diagnosis.