Bearing Fault Diagnosis Under Small Sample Data Based on SE-VAE and M1DCNN
Aiming at the problem of low diagnostic accuracy caused by the small number of fault samples in bearing fault diagnosis,a new bearing fault diagnosis method based on attention mechanism variation autoencoder(SE-VAE)and multi-scale one-dimensional convolutional neural network(M1DCNN)was proposed.Firstly,the training set of bearing data set is input into SE-VAE for training,generated samples with similar distribution to the training samples are obtained and added to the training set to increase the number of samples in the training set.Then,the extended training set is input into M1DCNN for training,and finally the trained model is applied to the test set to output the fault diagnosis results.Experimental results show that the proposed method can achieve better fault diagnosis accuracy on small sample bearing fault data sets with different loads.
国家科技期刊平台
NSTL
万方数据
