基于FFT⁃1D⁃CNN的细纱机罗拉轴承故障诊断  

Fault diagnosis of roller bearing in spinning frame based on FFT-1D-CNN

作　　者：陈宇航 李正平 肖雷 CHEN Yuhang;LI Zhengping;XIAO Lei(Donghua University,Shanghai,201620,China)

机构地区：[1]东华大学,上海201620

出　　处：《棉纺织技术》2023年第1期16-21,共6页Cotton Textile Technology

基　　金：国家自然科学基金面上项目(52075094);国家自然科学基金青年基金(51705321)。

摘　　要：提出一种基于快速傅里叶变换(FFT)及一维卷积神经网络(1D⁃CNN)的变转速细纱机罗拉轴承故障诊断方法。将故障罗拉轴承安装在细纱机小样机上监测振动加速度信号,经FFT转化成频域信号,利用1D⁃CNN快速从频域信号中学习故障特征进行诊断。结果表明:该方法的诊断准确率达99.917%,且耗时能满足企业端到端的诊断需求。与其他主流方法对比,该研究方法在准确率上具有优越性,能有效提高变转速下对细纱机罗拉轴承故障诊断的效果。A fault diagnosis method was proposed for roller bearing of variable speed spinning frame based on fast Fourier transform(FFT)and One-dimensional Convolutional Neural Network(1D-CNN).Fault bearing was installed on the spinning frame to monitor its vibration acceleration signals.Then,collected vibration signals were transformed into frequency domain signals by FFT.1D-CNN was used to quickly learn fault characteristics from frequency domain signals for diagnosis.Result showed that diagnostic accuracy of the method was 99.917%and time consuming could meet the end-to-end diagnostic needs of enterprises.Compared with other mainstream methods,the proposed method had advantage of higher accuracy.It is considered that the proposed method can effectively improve fault diagnosis effect of roller bearings of variable speed spinning machines.

关 键 词：罗拉轴承 故障诊断 快速傅里叶变换 一维卷积神经网络 变转速 

分 类 号：TS101.9[轻工技术与工程—纺织工程]