Bayes-DCGRU with bayesian optimization for rolling bearing fault diagnosis

摘要

Deep learning has been diffusely used in bearing fault diagnosis. In view of the increasing complexity of the model and the exponential growth of the hyperparameters, the adjustment becomes increasingly difficult. In this paper, a Bayesian optimization-Deep convolution gate recurring unit (Bayes-DCGRU) based on Bayesian optimization is proposed. It adopts Bayesian optimization algorithm to automatically adjust the hyperparameters of the model, and convolutional neural network (CNN) adaptively extracts the spatial characteristics of bearing signals. Combined with the Gate recurring unit (GRU) to learn the time series characteristics of signals, and then achieve high precision bearing fault identification. This method overcomes the shortcomings of the traditional hyperparameter adjustment method based on experience and feeling. It provides a solution for the hyperparameter adjustment of bearing fault diagnosis model under multi-hyperparameter coupling. The experimental results show that the model obtained by this method converges quickly and the fault identification accuracy is higher.