Design and development of a fault identification system for reducers based on frequency domain analysis of vibration signals

Abstract

In order to improve the efficiency, accuracy and automation level of reducer fault diagnosis, a fault identification system based on vibration signal frequency domain feature extraction, SVM and CNN technology is designed in this paper. Firstly, the original vibration signal of the reducer is denoised, and then the frequency domain analysis is performed to extract the fault features. Then, Support Vector Machine (SVM) and Convolutional Neural Network (CNN) were constructed respectively for fault classification and recognition, and their performances were compared. The experimental results show that the recognition accuracy of CNN model reaches more than 80%, which is better than that of the traditional SVM method (about 70%), demonstrating the advantages of CNN method in the field of reducer fault recognition. In the future, we can try to use a combination of methods to reduce the probability of overfitting and improve the accuracy.