Research On Gearbox Fault Diagnosis Based On Synchrosqueezing Wavelet Transform And Order Tracking

Gearbox is a universal component of mechanical equipment to transmit motion and power.It is widely used in mechanical equipments of electric power,aviation,transportation,metal-cutting machine tools,mining metallurgy and so on.Fault characteristics of low-speed gear fault signal in multi-stage gearbox under condition of variable speed are very weak due to time-varying and non-stationarity,and fault diagnosis of gear fault is difficult.Therefore,it is very necessary to extract fault features of variable speed gearbox.In this paper,the multi-stage gearbox with variable speed is taken as the research object,and order tracking and synchrosqueezing wavelet transform(SWT)are taken as the main technical means,combined with continue wavelet transform(CWT)and improved empirical wavelet transform(IEWT)respectively.Weak fault feature extraction of variable speed gear is studied.The contents are as follows:(1)Basic theories of order tracking technology and SWT are introduced.This paper focuses on the principle and method steps of tacho-less order tracking technology based on instantaneous frequency estimation(IFE),and compares SWT and CWT instantaneous frequency estimation based on peak search.Simulation analysis shows that synchrosqueezing wavelet transform has certain advantages in time-frequency analysis and signal decomposition and reconstruction.(2)Aiming at the problem that vibration fault characteristics of variable speed gears are particularly weak,a tacho-less order tracking method combining continuous wavelet transform-elliptic time-varying filtering(CWT-ETVF)and SWT is proposed to extract low frequency fault characteristics of gear.Firstly,CWT-ETVF and SWT are combined to estimate the instantaneous frequency of vibration fault signal to obtain the reference axis phase.Then the original signal is resampled at equal angles to obtain the angular stationary signal,and its order spectrum analysis and SWT decomposition are performed.Finally,SWT reconstruction component is selected for order spectrum analysis and order envelope spectrum analysis to extract the time-varying fault characteristics of gear full broken teeth.(3)Aiming at the problems that difficult extraction of low-frequency weak fault features of local broken teeth in multistage gear transmission,mode aliasing and component impact feature weak in SWT decomposition,a fault diagnosis method of tacho-less order tracking combining SWT and IEWT is proposed.Firstly,SWT is used for instantaneous frequency estimation,angular stationary signals are obtained after uniform angular sampling.Then IEWT is used for frequency band division and signal decomposition.Finally,an appropriate component is selected for autocorrelation denoising,and order demodulation analysis is carried out to extract the low-frequency and weak fault characteristics of local broken teeth under variable speed gearbox effectively.