Research On Fault Diagnosis Method Of Gearbox Bearing Based On Improved Enhanced Envelope Spectrum

At present,China’s high-speed railway is in the stage of vigorous development,the train operation speed continues to improve,so it has put forward higher requirements for its operation safety.Bearing is one of the most important parts in the train gearbox,which is under the excitation from wheel,rail and car body,and operates in a complex environment.Once the fault occurs,it will directly threaten the safety of train operation,so it will be the key monitoring object.Therefore,the bearing will be the key monitoring object.The vibration sources of gearbox bearing vibration signal are various,and the signal is weak,which is generally submerged in the strong background noise.The emphasis of this thesis is to study how to enhance the fault characteristics of bearing vibration signals and to realize bearing fault diagnosis.Spectral correlation(SC)is one of the most important tools for the analysis of cyclostationary signals.It can demodulate signals in the carrier frequency domain and the cyclic frequency domain.In addition,it has a certain effect of noise reduction in signal processing.However,due to the large amount of computation,it is hindered in practical application.Fast Spectral Correlation(Fast-SC)is an effective method for SC estimation.Compared with other SC estimation methods,Fast-SC can change the carrier frequency resolution and cyclic frequency resolution at will,and can estimate SC efficiently and accurately.Compared with traditional signal processing methods,more bearing fault information can be obtained by Fast-SC spectrum.Enhanced Envelope Spectrum(EES)is a new spectral quantity proposed on the basis of Fast-SC,which is the integral of the Fast-SC spectral graph.It can enhance the fault characteristics of signals to a certain extent.Although Fast-SC and EES have some advantages in bearing fault diagnosis,their application in bearing fault diagnosis under strong background noise is hindered.In this thesis,the advantages and disadvantages of Fast-SC and EES in handling bearing fault vibration signals are analyzed by establishing simulation signals.Based on the advantages of Fast-SC and EES,the method of locating the high frequency resonance band caused by bearing failure is proposed to improve the EES.The Harmonic Product Spectrum(HPS)can enhance the bearing fault characteristics in the signal envelope Spectrum,and is robust to outliers and measurement noise.The Harmonic Significance Index(HSI)calculated by HPS can effectively reflect the real bearing health.Whale Optimization Algorithm(WOA)is a new global optimization algorithm,which has the advantages of simple operation,few parameters and excellent performance.In this thesis,on the basis of Fast-SC analysis,HIS and WOA are used as fitness functions to find the bearing fault sensitive frequency band in the carrier frequency domain,so as to further obtain the improved enhanced envelope spectrum(IEES)of the signal.The IEES algorithm is applied to the processing and analysis of simulation signal of gearbox and bearing vibration signal of bearing failure in high speed train gear box bench test,and compared with EES and traditional fault diagnosis algorithm.The results show that the IEES can effectively enhance the bearing fault characteristics under the background noise,realize the bearing fault diagnosis,and avoid misdiagnosis and missed diagnosis of bearing fault.