A Study Of Mechanical Fault Diagnosis Technology Based On Deep Neural Network

Nowadays,mechanical equipment fault diagnosis technology has received more and more attention in modern production.If equipment failure has not been discovered and eliminated in time,the result will not only cause equipment damage,but also cause serious economic losses or even threaten personal safety.Therefore,the study of mechanical equipment fault diagnosis technology has very important significance.This paper firstly investigates the current situation of mechanical equipment fault diagnosis technology,and studies the fault mechanism,vibration signal feature extraction technology and intelligent fault diagnosis method in the field of fault diagnosis,and then make an empirical mode decomposition algorithm based on mechanical equipment vibration signal.This method decompose the vibration signals of mechanical equipment by empirical mode decomposition,and selects some intrinsic mode function which contains the main fault information and extract the energy feature,and inputs the energy feature into the Siamese network to realize the mechanical fault diagnosis.In order to make the Siamese network model better adapt to the fault diagnosis problem,this paper improves the structure and decision-making process of twin neural network.Finally,the model proposed in this paper is validated by using the open datasets of rolling bearings in the Bearing Data Center of Case Western Reserve University.The experimental results show that the proposed fault diagnosis model based on deep neural network can effectively identify rolling bearing faults with a small number of training samples.