Fault Diagnosis of Communication Equipment Gear based on Deep Learning

Abstract

Traditional mechanical fault diagnosis methods often need to process the collected fault wave signal, and then combine with neural network for feature extraction and classification, which not only has complex process, time-consuming, but also has low recognition accuracy. In this paper, one-dimensional convolutional neural network (1d-cnn) is used to extract and classify the features of gear fault vibration data of a communication equipment, and a one-dimensional convolutional neural network model of gear fault is established to diagnose the bearing fault of communication equipment. From the test and analysis results, the accuracy of the neural network model for gear classification can reach 78.81%, which is 15% higher than that of the traditional feedforward neural network with 63.71%; The accuracy of this method is 16% higher than that of SVM. This method can directly take the waveform vibration signal as the input, and output the final classification result through a series of operations such as convolution and pooling, which simplifies the traditional cumbersome steps of signal processing and machine learning diagnosis, and provides a feasible method for communication equipment fault diagnosis.