基于时频增强的滚动轴承少样本故障诊断方法  

A Method of Fault Diagnosis with Few Samples for Rolling Bearing Based on Time-Frequency Enhancement

作　　者：胡向东[1] 梁川 杨希 HU Xiang-dong;LIANG Chuan;YANG Xi(School of Automation/School of Industrial Internet,Chongqing University of Posts and Telecommunications,Chongqing 400065,China)

机构地区：[1]重庆邮电大学自动化学院/工业互联网学院,重庆400065

出　　处：《计量学报》2023年第1期12-20,共9页Acta Metrologica Sinica

基　　金：重庆市自然科学基金(cstc2021jcyj-msxmX0330)。

摘　　要：针对滚动轴承故障样本稀缺、振动特征提取困难导致故障诊断准确率低的难题,提出一种基于时频增强的滚动轴承少样本故障诊断方法。首先,对滚动轴承一维振动信号进行重叠采样,利用连续小波变换对采样信号段进行时频域特征映射,构造二维时频矩阵;其次,通过深度卷积生成对抗网络对真实时频样本进行训练后,将生成时频样本加入到训练集中;然后,采用时序卷积网络融合深层次的时频域特征;最后,构建Softmax分类器输出与故障类别对应的状态。仿真实验结果表明,在仅有10个训练样本的条件下,该方法在凯斯西储大学滚动轴承数据集中的诊断准确率均值达91.00%,相较未经时频增强的方法提高了7.56%,并利用实测数据验证了时频增强方法能够显著提升少样本情形下的故障诊断准确率。Aiming at the problem of low fault diagnosis accuracy due to the scarcity of fault samples of rolling bearings and the difficulty in extracting vibration features,a fault diagnosis method of rolling bearings with few samples based on time-frequency enhancement is proposed.Firstly,one-dimensional vibration signals of rolling bearings are overlapped and sampled,and a two-dimensional time-frequency matrix is constructed by using continuous wavelet transform to map the sampled signals in time-frequency domain.Secondly,the real time-frequency samples are trained by deep convolutional generative adversarial network,and the generated time-frequency samples are added to the training set.Then,the time-series convolutional network is used to fuse the deep time-frequency domain features.Finally,Softmax classifier is constructed to output states corresponding to fault categories.Simulation experiment results show that under the condition of only 10 training samples,the method in the diagnosis of rolling bearing at case western reserve university data set average accuracy of 91.00%,compared with an enhanced time-frequency method is increased by 7.56%,and the time-frequency method is validated by the measured data can significantly improve the fault diagnosis accuracy of less sample circumstances.

关 键 词：计量学 滚动轴承 故障诊断 时频增强 少样本 

分 类 号：TB936[一般工业技术—计量学]