Fault Diagnosis of Rotor System Based on Attention Mechanism and Lenet5 Network
The high efficiency of vibration parameter processing is difficult to achieve when the empirical value method is used to deal with partitioned data,which will reduce the accuracy of the reconstructed data.In order to further improve the fault diagnosis ability of rotating machineries,a fault diagnosis method based on the attention mechanism and the LeNet5 network is proposed and successfully applied to a rotor system.The results show that compared with the traditional LeNet5 network,the proposed method overcomes the limitations of the inattentive mechanism,and the overall performance is better without the need to change the parameter settings.By adding the attention mechanism to the LeNet5 network,the accuracy of fault identification and the training speed of the model are significantly improved.Compared with other methods,the fault type detection results of this method demonstrate the highest accuracy rate,all of which are above 95％,and meet the practical requirements.In the diagnosis of rotating machinery faults,the convolutional neural network and attention mechanism are combined to test and study,and finally proved that the feasibility of this method is high,and the research can be extended to other mechanical transmission fields,which has certain application value.
NETL
NSTL
万方数据
