Reasearch And Application Of Fault Diagnosis On The Running Gear Resonance Demodulation In Locomotive

With the rapid development of rail transportation, the safe operation of the locomotive has been paid more and more attention. If the mechanical fault cannot be found and eliminated in time during the operation of the locomotive, it will cause serious damage once the accident happen. So, as the key part of high speed locomotive, the safety of the locomotive running gear becomes particularly important, and it is necessary to monitor the locomotive running gear in real-time during the operation of the locomotive. Because of these, this research mainly focuses on the real-time online monitoring and the fault diagnosis method of the locomotive running gear. Based on the mechanism of the common faults of the locomotive running gear, the research realized the processing structure as FPGA+ARM by using the diagnosis technique based on vibration signal analysis, and completed the vibration signal acquisition, processing and analysis of the locomotive running gear.The thesis is mainly about the real-time online monitoring of locomotive running.After analyzing the direction of common faults of locomotive mechanism and contrasting different analysis methods of the common vibration signals, such as the Hilbert-Huang Transform and wavelet transform of vibration signal, the research selected the resonance demodulation method, which is easy to realize in hardware and convenient for quantitative analysis, to extract the fault signals of locomotiverunning gear, and processed the resonance demodulation signals by FFT transform.In order to improve the accuracy of the data of the FFT transform, the window function would be simulation compared before the FFT transform. The research completed the processes, such as fault signal acquisition, data processing and storage, and signal transmission, on the Xilinx ISE platform, and simulation analysis of various modules was carried out by using Modelsim. Finally, the research integrated the monitoring system and tested its resource utilization ratio and running speed. The results of the test show that, compared with the current design, the design in this research shortened the single period time, which is required to complete the acquisition, transmission, processing of the fault signal and ensures the design to meet the real-time requirements. Another thing is in the FPGA, the output of FFT is consistent with the results of MATLAB simulation analysis, which also verified the reliability of the data processing in the research.