Fault Diagnosis Of Servo Drive System Of Cnc Machine Based On Deep Learning

The drive system of computer numerical control(CNC) machine is characterized by nonlinearity, uncertainty and so on. In this paper, a fault diagnosis algorithm based on dual stack sparse autoencoder model is proposed by utilizing the powerful feature extraction and data compression function of deep learning network. The method is to use a pile of sparse autoencoder network to learn the high-level features of the data, and use the softmax classifier to classify the data, and solve the diagnosis of the overloading and lubrication of the CNC machine. Through the simulation experiment of the servo drive system of CNC machine, it is shown that the learning network has a high success rate of failure detection.