Fault diagnosis model training method and device, electronic equipment and storage medium

Abstract

The invention provides a fault diagnosis model training method and device, electronic equipment and a storage medium, and the method comprises the steps: respectively inputting a sample target domainsignal of first equipment and a sample source domain signal of second equipment into corresponding convolution layers, and obtaining a first target domain feature and a first source domain feature; respectively inputting the first target domain feature and the first source domain feature into respective corresponding domain adaptation layers to obtain a second target domain feature and a second source domain feature, and determining domain adaptation loss based on the second target domain feature and the second source domain feature; determining a first probability that the first device belongs to a preset fault type based on the second target domain feature and the second source domain feature, and determining a first classification loss based on the first probability and a target fault type of the first device; and if the domain adaptation loss and the first classification loss of the fault diagnosis model meet a first training condition, determining the current fault diagnosis modelas a target fault diagnosis model.

Description

technical field [0001] The present application relates to the field of computer information technology, in particular, to a training method, device, electronic equipment and storage medium of a fault diagnosis model. Background technique [0002] Rolling bearings are one of the most numerous and most critical components in rotating machinery, and their health status directly affects the overall stability, reliability and safety of rotating machinery. Therefore, it is very important to carry out state detection and fault diagnosis of rolling bearings. [0003] At this stage, a supervised learning model matching each type of rolling bearing is usually used to detect the failure of each type of rolling bearing. When training the supervised learning model, it is necessary to obtain a large amount of fault data for each type of rolling bearing. . [0004] However, for precision rolling bearings, it is difficult to obtain a large amount of fault data, and the supervised learning ...

AI Technical Summary

Problems solved by technology

[0004] However, for precision rolling bearings, it is difficult to obtain a large amount of fault data, and the supervised learning model is trained based on a small amount of fault data. The accuracy of the trained supervised learning model is low, which leads to low efficiency of rolling bearing fault diagnosis

Images