Fault Diagnosis Based on improved Deep Belief Network

A fault diagnosis method based on improved Deep Belief Network is proposed. This paper proposed an network named Semi-DBN, which adapts an adaptive semi supervised method to train the RBM structure to improve the performance of DBN, replaces the Sigmoid activation function with Rectified Linear Units to improve the performance of the network, and combines the model with Transfer Learning to solve the problem of lack of data. It can get rid of the dependence of the traditional machine learning methods on the extraction of the sample characteristics, and effectively overcome the problem of gradient vanishing, local extremum and so on. The identification and generalization ability of the method is verified when we used it in the experiment of rolling bearing data.