沈阳航空航天大学学报2024，Vol.41Issue(3)：37-42,6.DOI:10.3969/j.issn.2095-1248.2024.03.006
基于swin transformer与无监督学习的滚动轴承故障诊断方法
Fault diagnosis method of rolling bearing based on swin transformer and unsupervised learning
摘要
Abstract
In order to carry out fault diagnosis with only health status data,an optimized swin trans-former deep neural network architecture was constructed to extract and reconstruct health data features,which proposed an unsupervised learning method for rolling bearing fault diagnosis.Compared with au-toencoder,depth encoder,convolutional autoencoder,and sparse autoencoder,the accuracy is 98.62%,76.46%,68.69%,77.69%,68.00%,respectively,which is more than 20%higher than the accuracy of comparison network.关键词
滚动轴承/故障诊断/深度学习/无监督学习/swin transformerKey words
rolling bearing/fault diagnosis/deep learning/unsupervised learning/swin transformer分类
航空航天引用本文复制引用
张鸾,闵思垚,张微..基于swin transformer与无监督学习的滚动轴承故障诊断方法[J].沈阳航空航天大学学报,2024,41(3):37-42,6.基金项目
国家自然科学基金(项目编号:11902202) （项目编号:11902202）
