Data Analysis Of Acoustic Emission Signal Of Axle Fatigue Crack Based On Deep Belief Network

With the rapid development of rail transit industry,the speed and load of rail vehicles are gradually increasing.At the same time,security has always been the top priority of the industry.The axle is the main part of the wheel sets,and it is also an important part of the running part of the vehicle.It is an indispensable part in the operation of the railway vehicle.Axles are faced with changing external environment and complex working conditions,which inevitably lead to cracks.The development of axle cracks will cause axle fracture to a certain extent,and bring extremely serious consequences.Therefore,the fault diagnosis of axle has practical significance.This paper studies how to extract features and classify AE data of axles,and proposes an analysis method based on Deep belief network(DBN).Firstly,the basic principle and network structure of DBN are introduced.On this basis,the Restricted Boltzmann machine(RBM),which is a part of DBN,is studied.The influence of the number of iterations of weight updating on feature extraction and network computing time is analyzed.In the process of extracting features of AE signals from axles by DBN,principal component analysis(PCA)is used to reduce the dimensions of the extracted multi-dimensional features.The first three principal components are retained so that the features can be represented in three-dimensional space.In the process of classification and recognition of axle AE signals by DBN,the change of signal recognition accuracy caused by different hidden layers is mainly analyzed.Then,by comparing with the classification accuracy of Extreme Learning Machine(ELM),a machine learning network model,the advantages of DBN in crack signal recognition are verified.For different axle cracks,DBN is used to carry out experiments in other axle acoustic emission signal datasets.Finally,aiming at the complete AE signal of axle crack,according to the change of signal kurtosis,the crack is divided into four different stages,and the signals of different stages are put into the DBN network for classification and recognition.Experiments on different data sets show that the deep belief network can recognize crack signals from various AE signals of axles and has good classification ability for different stages of crack signals.