Fault diagnosis method for deep adversarial migration network based on Wasserstein distance

Abstract

The invention discloses a fault diagnosis method for a deep adversarial migration network based on a Wasserstein distance. The distance of the feature distribution of the two fields in the feature space is measured through the wassertein distance, the feature distribution is adapted, the difference between two fields is reduced, field-independent features are learned to train an effective classifier, the classifier is responsible for mapping the field-independent features to a category space to complete a classification task, and the problem of unsupervised transfer learning of vibration datawithout labels in a target domain is solved.

Description

technical field [0001] The invention relates to the technical field for identifying the fault category of unlabeled vibration data in the field of fault diagnosis, in particular to a fault diagnosis method based on Wasserstein distance-based deep adversarial migration network. Background technique [0002] In complex industrial systems, the study of advanced mechanical fault diagnosis methods is an important content to ensure the safety of equipment and personnel. With its powerful modeling and representation capabilities, deep learning theory has become one of the most active frontiers of data-driven intelligent fault diagnosis. However, using deep learning to train fault classification models requires a large amount of labeled data, and the training data and test data satisfy independent and identical distribution. These two conditions are usually difficult to satisfy in practical applications. How to use auxiliary domain data to establish a reliable mathematical model a...

AI Technical Summary

Problems solved by technology

However, due to the shortcomings of these domain adaptive methods to measure the distribution distance algorithm, the final accuracy of the classifier is not ideal enough

Images