Study On Fault Diagnosis Method Of EMU Anti-yaw Damper

Anti-yaw damper is an important part of vehicle suspension system.Its performance will directly affect the running stability and ride quality of the vehicle.However,it is difficult to directly judge its state from the appearance,which will increase the risk of the vehicle during operation.With the advancement of EMU optimization of repair,in order to match the extended maintenance period of the EMU,the disassembly and maintenance period of the anti-yaw damper also needs to be extended accordingly.Research on fault diagnosis methods has important research significance and application value.This paper combines modern signal processing methods and machine learning methods,construct fault diagnosis method of EMU anti-yaw damper.The fault diagnosis of the anti-yaw damper is realized by using the vehicle frame and the car body vibration acceleration signal data.This paper done the studies is as follows:(1)Analysis on the dynamics performance test data of 250 km/h Fuxing EMU on a250km/h line in China.Then study on the effect of anti-yaw damper failure on vehicle dynamics.The results show that the failure of the anti-yaw damper has a more obvious impact on the indicators related to the lateral vibration of the vehicle.(2)A vehicle dynamics simulation model was established by using SIMPACK multibody dynamics software.The validity of the established model is verified by the nonlinear critical velocity method and the car body vibration acceleration comparison method.(3)Analyze the time domain and frequency domain characteristic parameters of the car body lateral vibration acceleration signal of the vehicle dynamics simulation model under different vehicle operating conditions.Decomposing vibration signals based on EEMD method,using IMF with high correlation coefficient with original signal as fault feature extraction signal.Take the feature vector with more obvious differences,build a feature extraction method for fault diagnosis of EMU anti-yaw damper.(4)Build a fault diagnosis of EMU anti-yaw damper based on EEMD-SVM method through divided and normalized data.Select sample length,feature vector and penalty parameters.Identify and verify the model with the measured lateral vibration acceleration data of the car body and frame.The results show that the accuracy of the model recognition results is 90.53%.In this paper,the vehicle dynamics simulation model is used to vehicle vibration signal data of different vehicle operating conditions,different speed levels,and different line conditions.The fault diagnosis method of anti-yaw damper based on EEMD-SVM proposed in this paper has a good recognition rate in practical application cases,and provides a theoretical basis for enriching vehicle safety monitoring and early warning functions.30 figures,19 tables,and 36 references.