An Intelligent Fault Diagnosis Method Based on Importance Weighted Domain Adversarial Adaptive

Abstract

An intelligent fault diagnosis method based on importance-weighted domain confrontation and self-adaptation, which collects vibration signals of rotating machinery under different working conditions, and uses moving time windows to segment the data sets under different working conditions; constructs a domain category recognition network , output the importance weight of the source domain samples in the confrontation training; extract the discriminative features in the data set; combine the feature extractor and the domain discriminator to construct the importance weighted domain confrontation adaptive network; use the training strategy of the confrontational network to train the network model to The model converges, using the trained category classifier to identify the bearing health status of the target domain dataset lacking fault labels. The present invention performs fault diagnosis on working conditions with insufficient data information by means of working conditions with rich data information, completes the migration of diagnostic knowledge, and constructs a deep learning network at the same time, overcomes the dependence on expert knowledge in traditional diagnostic methods, and in order to reduce The cost of future intelligent fault diagnosis system provides an effective tool.

Description

technical field [0001] The invention relates to a rolling bearing state evaluation method, in particular to an intelligent fault diagnosis method based on importance weighted domain confrontation self-adaptation. Background technique [0002] As an important rotating part in mechanical equipment, rolling bearings are widely used in aerospace, engineering machinery, marine equipment, water conservancy engineering and other fields. The health status and performance of rolling bearings directly affect the safety and reliability of mechanical equipment. However, rolling bearings are vulnerable parts in mechanical equipment, and failure of bearings may cause the shutdown of the entire mechanical system, resulting in unimaginable economic losses. Therefore, the research on bearing health status detection and fault diagnosis technology is of great significance. [0003] By using the sensor to collect the vibration signal of the bearing for analysis during the operation of the bea...

AI Technical Summary

Problems solved by technology

[0004] In view of the deficiencies in the prior art, the purpose of the present invention is to provide an intelligent fault diagnosis method based on importance weighted domain confrontation self-adaptation, which solves the deficiencies of the traditional intelligent diagnosis method based on transfer learning in industrial applications, overcomes the The target domain training set is a data set that does not contain complete fault types, resulting in negative transfer

Images