Abstract
The article considered methods for diagnosing bearing sliding units and methods for their implementation. Based on the information received, an experiment was carried out to diagnose a sliding bearing unit using a non-contact method of thermal control, followed by processing the da-ta obtained using ResNet convolutional neural networks. In the process of conducting a continuous experiment, four different states of the sliding bearing unit were identified based on the images of thermograms. On the basis of the obtained data, a convolutional neural network was trained with the subsequent solution of the problem of classifying defects according to the image data of thermograms.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
