Aiming at the high dependence of super parameter selection on artificial experience in rolling bearing state identification based on convolution neural network(CNN)

a fault diagnosis model(CNN⁃MA)based on mayfly algorithm(MA)was proposed.Firstly

the model used the powerful optimization ability of MA

took the diagnostic accuracy of CNN as the optimization objective

and adaptively adjusted the super parameters in CNN.Secondly

the normalized original signal image set was used to preserve the characteristics of the signal as much as possible.Finally

in order to evaluate the effectiveness of the parameters in the optimization model

compared with the CNN model optimized by particle swarm optimization(PSO)algorithm.The results show that the proposed model has more stable performance

higher recognition accuracy and good anti⁃noise ability.It fully shows the feasibility and reliability of CNN⁃MA model in fault diagnosis of rolling bearings.