Fault Diagnosis Of Rolling Bearings For Urban Rail Trains Based On Trackside Acoustic Signals

Urban rail transit plays an increasingly important role in the process of urbanization,and the safety performance of urban rail vehicles has received more and more attention.The traveling system is the most important part of urban rail vehicles,and the rolling bearings which are more prone to failure in the traveling system,are vital importance to the safe operation of trains.Therefore,it is very important for the condition monitoring of train rolling bearings.This paper summarized the former research basis,a fault diagnosis method for train rolling bearings based on wayside acoustic signals was presented.The main contents are as follows:Firstly,the rolling bearing structure of urban rail vehicles was introduced.The common failure modes of bearing were summarized,and the characteristics of fault signals of different parts were summarized.The mechanism,characteristics and propagation rules of bearing sound signals were analyzed.And the time-frequency characteristics and the energy characteristics in the signal analysis of rolling bearing were introduced;Secondly,the corresponding acoustic signal acquisition system was designed according to the rolling bearing experimental platform.According to the objective,the acoustic signal acquisition experiment of the array sensors were designed,the experiment under different parameters was designed,and the corresponding bearing acoustic signal was collected which provide data support for further study;Thirdly,aiming at signal incompleteness from the single sensor signal during the motion of the bearing,the acoustic signal fusion algorithm for the sensor array based on the instantaneous weight coefficient was proposed,and the validity of the fusion model was verified.In order to highlight the effectiveness of the fusion algorithm,the Adaptive Morphological Filtering(AMF)was applied to the collected acoustic signal before the fusion.Finally,an algorithm for rolling bearing fault diagnosis based on the Kurtosis to Shannon entropy ratio(KER)and Ensemble Empirical Mode Decomposition(EEMD)was proposed.This paper introduced the basic principle the wavelet transform,the wavelet packet decomposition,the KER and EEMD.In this paper,the proposed methods are effectively combined and the algorithm of this paper is presented.The bearing fault diagnosis were carried using the simulation signal,experimental signal and field measurement signal as the original signals.The experimental results show that the method can effectively extract the fault features contained in the acoustic signal,and prove the effectiveness and feasibility of the method.