改进掩码自编码器的滚动轴承半监督故障诊断

针对滚动轴承在不同转速条件下数据分布不同以及实际工程应用中标签样本不足导致故障诊断精度低的问题，将领域适配模块融入掩码自编码器(MAE)中，提出了改进掩码自编码器(IMAE)的滚动轴承半监督故障诊断方法。首先，对滚动轴承振动信号进行连续小波变换(CWT)得到反应信号时频特征的二维时频图，然后对时频图随机掩码，利用无标签样本进行掩码自编码器预训练，获得数据中复杂的内在特征，减少对有标签样本的依赖；其次将领域适配模块引入到预训练后的编码器中，使用少量有标签源域数据对IMAE进行微调，在希尔伯特空间中利用最小化最大均值差异减小因转速不同造成的源域与目标域间数据分布差异；最后在Softmax分类层下实现滚动轴承半监督故障诊断。通过滚动轴承数据集实验验证，所提方法检测精度均达到94%以上，证明了该方法的可行性与有效性。

Improved semi-supervised fault diagnosis of rollingbearings with mask autoencoder

To address the problems of different data distribution of rolling bearings under different speed conditions and low faultdiagnosis accuracy caused by insufficient label samples in practical engineering applications, the domain adaptation modules areintegrated into masked autoencoders (MAE). An improved masked autoencoders (IMAE) method for semi-supervised fault diagnosis ofrolling bearings is proposed. Firstly, the two-dimensional time-frequency graph of the response signal is obtained by applying continuouswavelet transform (CWT) to the vibration signal of the rolling bearing. Then, the mask of the time-frequency graph is randomly masked,and the mask autoencoder is pre-trained with unlabeled samples to obtain the complex intrinsic features of the data. The reliance onlabeled samples is reduced. Secondly, the domain adaptation module is introduced into the pre-trained encoder, and a small amount oflabeled source domain data are used to fine-tune the IMAE, and the maximum mean difference is minimized in Hilbert space to reducethe data distribution difference between the source domain and the target domain caused by different rotational speeds. Finally, the semi-supervised fault diagnosis of rolling bearing is realized under the Softmax classification layer. Through the experimental evaluation of therolling bearing data set, the detection accuracy of the proposed method is more than 94% , which proves the feasibility and effectivenessof the proposed method.