Emu Bogie Health Assessment And Fault Diagnosis

High speed and heavy load is the established development strategy of China's railway transportation.The existing,under construction and future planned high-speed railway and EMU trains have become an important part of the "high speed" strategy.As the key part of EMU train,high-speed bogie is not only the safe part of EMU running smoothly,but also the dynamic load component bearing the complex alternating load of train.The research on the health management and fault diagnosis of EMU bogies in service has always been the focus in the field of rolling stock.This paper constructs a health management model to measure the service state of EMU bogies,and solves the problem of state information fusion of multiple data sources and multi-sensor sources.In the MATLAB software environment,based on the evaluation index,fuzzy analytic hierarchy process model and BP neural network model to measure the health status of EMU bogies,the field data are simulated to verify the health evaluation model of EMU bogies.The technical method is to train BP neural network with the sample data obtained by fuzzy analytic hierarchy process,and then use the test sample of the simulated field data to the fuzzy BP neural network,to make the available,maintenance,fault,overhaul and other hierarchical treatment countermeasures for the health status of EMU bogie.For the maintenance,fault and maintenance rating of EMU bogies,the basic work needs to carry out the prediction and maintenance of the deterioration process of the health status.In this paper,the LabVIEW software platform and matlab mixed programming function are combined to design a fault diagnosis system for the health status rating of the key parts bearing,gear and wheel set of EMU bogie,and simulate and analyze the vibration characteristic signals of the key parts bearing,gear and wheel set when they are in fault status.For the bearing vibration signal,the wavelet packet analysis is used to remove the noise in the original signal,the first five IMF components of the signal are obtained by EMD decomposition,the energy value is calculated as the fault feature vector of the input BP neural network,and the BP neural network is used to identify the bearing fault;for the gear vibration signal,the filter is used to eliminate the environmental noise,and the frequency spectrum analysis and power spectrum division after signal demodulation are used Analysis and FFT spectrum analysis and other signal frequency-domain analysis methods identify the frequency-domain waveform of the fault signal and diagnose the fault of the gear;for the vibration signal of the wheel set,wavelet analysis,discrete cosine transform and envelope spectrum analysis are used to identify the signal frequency component contained in the signal,and the characteristic frequency value of the fault signal of the wheel is compared and analyzed to realize the diagnosis process of the wheel set fault.In this paper,wavelet analysis,EMD decomposition,BP neural network,frequency domain analysis and envelope spectrum analysis are used to realize the fault diagnosis of bearings,gears and wheel sets of key parts of EMU bogie.