Study On Vibration Characteristics And Fault Diagnosis Method Of Locomotive Gear Transmission System With Root Crack

Gear box transmission system whose function is transferring the torque of motor to wheelset and drive train is an important part of locomotive.The working condition of gearbox transmission system is generally hard.Specifically,it is not only affected by transmission load and wheel-rail impact load,but also adapted to the complex relative motion between wheelset and frame.Under the condition of train operation,once the gear box system occur breakdown,the breakdown will deteriorate rapidly.If the breakdown is not found in time or without corresponding measures,it will lead to serious and extraordinary accidents such as bearing combustion,gear breakage,box damage,power transmission interruption and vehicle destruction.These accidents will seriously endanger the safety of driving.Therefore,the research on breakdown detection and diagnosis of gearbox transmission system is important to ensure the safe operation of train.The core of effective detection for gearbox system's faults is correctly understanding and mastering the fault behavior rule and the essential characteristics of the fault mode of the gearbox system under the actual operation conditions of trains.Based on the vertical dynamic model of vehicle-track coupling with gear transmission system,the research on the fault diagnosis method of gear box system under train operation condition is processing in this thesis.The main contents are as follows:(1)In order to understand the influence of root crack faults on the vibration characteristics of various parts of locomotive,the root crack faults are introduced into the model which is the vertical dynamic model of train track coupling with gearbox transmission system.Firstly,the vibration signals of locomotive components,such as wheelset,motor,frame and car body,are obtained by using numerical simulation method under the conditions that the gearbox gears contain root cracks or do not contain root cracks.Then,the vibration signals are analyzed and compared by using time domain and frequency domain analysis methods.At the same time,the vibration characteristics of locomotive components with track excitation are analyzed.(2)Additionally,the mechanism of modulation of vibration signals of locomotive components collected from gearbox gears with root cracks and the envelope demodulation method based on complex Morlet wavelet are analyzed in this thesis.By establishing the dynamic model of the fixed-axle gear system,the frequency characteristics of vibration signals for the fixed-axle gear system under normal meshing and impulse-type faults are analyzed.Aiming at the issue that the signals collected in the gear system with impulse-type faults have modulation phenomenon,the research on commonly used signal demodulation methods are processed.This thesis mainly introduces Hilbert envelope demodulation method,complex Morlet wavelet envelope demodulation method,complex Morlet wavelet cluster envelope demodulation method,complex Morlet wavelet cluster bandpass filter and its envelope demodulation method.(3)Furthermore,due to that the vibration signal of each part of the locomotive is weak and susceptible to noise interference when the gear box contains the root crack fault,the fault feature enhancement method of rotating machinery based on noise assistant analysis is researched in this thesis.The numerical solution method of discrete model of bistable system and the normalized scale transformation method of bistable system are introduced.Meanwhile,a weak aperiodic signal detection method based on time domain synchronous averaging method and stochastic resonance principle is proposed.(4)Moreover,the method of extracting impulse components from vibration signals of locomotive components caused by gear box with root cracks has been researched in this thesis.The method used in this thesis for extracting impulse components of vibration signals based on the theory of graph Fourier transform.Aiming at the problem that the impulse components extracted by graph Fourier transform are affected by noise,the fault diagnosis methods based on stochastic resonance and graph Fourier transform are proposed.By applying the above method to process the simulation signal and the actual signal,we can find that the fault diagnosis method based on stochastic resonance and graph Fourier transform theory can carry out the fault of locomotive gear root crack...