Vehicular Track Dynamic Research Of Fault Diagnosis Method

It is necessary to strengthen the management and monitoring of the trackirregularity in order to increase the speed of train and achieve the target of high density,high speed of the passenger transport and heavy freight, ensure safety and comfort ofthe travel. Traditional detect method which used special track inspection car or light railmonitoring car, is not suitable to use high speed and high density of the rail linefrequently, so we need to study other auxiliary detecting means.The auxiliary means commonly used at home and abroad is to install dynamicmonitoring device in the high speed train, through detecting the operating trainacceleration of vibration in horizontal and vertical directions, to synthetically evaluateline smooth state. The processing method of acceleration measured in this device isnot reliable and accurate, which has high rate of false alarm and distress railway repairwork. Therefore it is urgent to find an easier, more accurate fault diagnosis method.This paper is based on the vehicle track dynamics, using SIMPACK to set up themodel of vehicle track and verify the correctness of the model. On the basis of thesamples of American’s class6track spectrum into time domain signal, stacking sinesignal, as the excitation function of track, studying the train vibration on the track whichunder the action of fault irregularity. At last, in order to distinguish the fault oftrack irregularity, using wavelet decomposition to the simulation of the vibrationacceleration signal from the perspective of fault diagnosis and extracting the irregularitycharacteristic of each band as the eigenvalue of fuzzy neural network.