机电工程技术2025，Vol.54Issue(19)：70-76,88,8.DOI:10.3969/j.issn.1009-9492.2025.19.013
基于S变换与改进AlexNet网络的滚动轴承故障智能诊断
Intelligent Diagnosis of Rolling Bearing Faults Based on S-transform and Improved AlexNet Network
摘要
Abstract
Aiming at the problem that insufficient extraction of the correlation characteristics of information time series in traditional fault identification methods leads to low accuracy of fault diagnosis and identification,the S-transform graph encoding method is introduced into the fault identification of rolling bearings,and an intelligent fault diagnosis method for rolling bearings based on S-transform and improved AlexNet network is proposed.The S-transform graph encoding technology is used to convert one-dimensional vibration signals into two-dimensional image feature data,so that the loss of weak feature information is effectively avoided.The size of the converted two-dimensional images is adjusted to an appropriate size,and finally,the adaptive feature extraction capability of the improved AlexNet network is utilized to realize the intelligent fault identification of rolling bearings.Verification is carried out based on the rolling bearing fault data collected by the fault simulation experimental platform.The results show that,compared with other graph encoding methods as well as the improved LeNet and SVM classifier methods,the proposed method achieves a rolling bearing fault identification accuracy of 99.57%,and it exhibits higher identification accuracy and better generalization performance.关键词
滚动轴承/S变换/故障智能识别/图编码方法/改进AlexNet网络/二维图像Key words
rolling bearing/S-transform/fault intelligent recognition/graph encoding method/improved AlexNet network/2D image分类
机械制造引用本文复制引用
雷兵,李响,陈红,唐佳桃..基于S变换与改进AlexNet网络的滚动轴承故障智能诊断[J].机电工程技术,2025,54(19):70-76,88,8.基金项目
江西省教育厅科学技术研究项目(GJJ2207604,GJJ2207601,GJJ2403503) （GJJ2207604,GJJ2207601,GJJ2403503）
江西开放大学校级科研项目(JKND2407) （JKND2407）
