Research On Fault Diagnosis Methods Of Rolling Bearing Under Variable Rotational Speed Conditions

As a key component of rotating machinery,the healthy state of rolling bearing is one of the important factors that determine the safe operation of machinery.In lots of engineering practice,the work speed of bearing usually fluctuates in a certain range of speed,and the frequency modulation,amplitude modulation,phase modulation and other effects under the condition of variable speed make the traditional fault diagnosis strategies based on spectrum analysis useless.Therefore,how to effectively extract the time-varying fault information is one of the key problems of bearing fault diagnosis under the condition of variable speed.In this background,the characteristics of different diagnosis methods and the characteristics of the bearing fault signal of variable speed conditions were analysed.Two fault diagnosis methods were put forward respectively aiming at different working conditions.Specific research contents are as follow:(1)For the working environment where it is convenient to collect the tacho signals,the real-time speed of the tested axis was calculated according to the tacho signals and the instantaneous frequency curve of the tested axis was drawn;The fault characteristic coefficients of inner and outer ring of the bearing were calculated based on the geometric dimensions of the bearing,and the time-varying fault characteristic frequency curves of inner and outer rings were drawn.The time-varying amplitude demodulation spectrum based on sychroextracting transform was calculated and matched with the fault characteristic frequency curves of the inner and outer rings of the bearing so as to identify the fault types.(2)For the working environment where it is not convenient to collect the tacho signals,the instantaneous frequency estimation algorithm based on Veiterbi algorithm was applied to the bearing vibration signal.Based on maximum estimation of the STFT of the vibration signal in time-frequency grid,the instantaneous frequencies of tested bearing were got and the instantaneous frequency curve was drawn;The validity of the instantaneous frequency estimation algorithm was verified by experiments;The time-varying amplitude demodulation spectrum based on sychroextracting transform was calculated and matched with the fault characteristic frequency curves of the inner and outer rings of the bearing so as to identify the fault types.(3)The vibration response of the mass spring damper model of single degree of freedom was used to simulate the amplitude demodulated vibration signal caused by bearings with fault.The length of the signal and the number of the impact can be set by the impact number caused by the fault.The above methods were validated by the simulatied signal,and for comparision different methods and working conditions with different SNRs were validated with the simulated signal.(4)The working condition where it is convenient to collect the tacho signals was simulated by the vibration and fault simulation test bench of rotating machinery,with which the fault diagnosis method based on the tacho signal and synchroextracting transform were verified;The working condition where it is not convenient to collect the tacho signals was simulated by the integrated bench of high-speed rail bearings,with which the fault diagnosis method based on instantaneous frequency estimation and synchroextracting transform was verified.The diagnosis effects of diagnosis method proposed and another diagnosis method were compared.The two fault diagnosis methods of bearing and analysis results can provide a new idea for the study of health monitoring and precision fault diagnosis of bearing under variable speed conditions,and have certain theoretical and application value.