Abstract
In order to improve the diagnosis rate of bearing signals, the bearing fault diagnosis model ODCNN based on one-dimensional convolutional neural network is proposed to adapt to one-dimensional time-domain signals on the basis of AlexNet. The model is composed of a multilevel alternating convolutional layer and a pooling layer, which can complete the adaptive extraction of the original input signal features, and combine the fully connected layer to realize fault classification and recognition. Support vector machine, convolutional neural network classic AlexNet architecture, LeNet-5, particle swarm optimization algorithm support vector machine and BP neural network are introduced for comparison and verification by using the Shanghai Tiandi bearing test data set for fault diagnosis test. The results indicate that the algorithm proposed in this paper has high recognition accuracy. Finally, Principal component analysis is used to verify the model's feature learning and classification capabilities of vibration signals.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have