Planetary Gearbox Fault Diagnosis Under Time-varying Speed Conditions

Planetary gearboxes have appealing advantages of high transmission ratio,large bearing capacity,and compact structure over traditional fixed-shaft gearboxes.They are widely applied in wind turbines,heavy-duty vehicles,vessels and many other industrial equipment.However,planetary gearboxes often work in dusty and corrosive working environment,and are prone to faults like gear tooth profile damage,gear root crack and fatigue break.As importaint part of the power train,planetary gearbox fault will cause vibration intensification,efficiency loss,unplanned outage,and even structure destruction.Thus,planetary gearbox fault diagnosis is of significance.However,two main difficulties still hinder the development of planetary gearbox fault diagnosis techniques.(1)The fault charactersitcs of planetary gearboxes have not been clearly clarified.Current explanations on fault signatures mostly base on transverse vibration signals and assumption of stationarity.(2)Most available nonstationary signal analysis techniques still lack the ability to clearly reveal complex time-varying fault characteristics of planetary gearboxes.Powerful fault feature extraction methods are still in urgent demand to precisely and intuitively present planetary gearbox faults.In this study,three aspects of works have been done to solve the aforementioned two main difficulties in planetary gearbox fault diagnosis under variable speed condition.The main innovation points and contributions include:(1)In theoretical modeling aspect,the analytical model of planetary gearbox vibration in both transverse and torsional directions,and stator current of induction motor,are built under time-varying speed conditions.Extensitons from transverse vibration to torsional vibration and stator current,from constant to time-vaying speed conditions,and from gear meshing frequency sidebands to resonance frequency sidebands,are realized.(2)In time-frequency analysis aspects,improvements are made to frontier methods in order to accurately present planetary gearbox faults.Reassigned scalogram is successfully applied in extracting fault frequency harmonics and fault impluses in planetary gearbox vibration signals,and new methods including iterative generalized reassignment and iterative generalized synchrsqueezing transform are proposed to analyze complicatedly modulated nonstationary signals.(3)To facilitate fault feature extraction,traditional envelope analysis and instantaneous frequency analysis are extended to time-varying amplitude and frequency demodulation analysis,and new order analysis method based on iterative generalized demodulation is proposed.Besides,resonance frequency sideband analysis is proposed for utilizing torsional vibration signal,and time-varying space vector analysis is proposed for utilizing motor stator current signal,in order to provide new ideas and solutions to planetary gearbox fault diagnosis.Lab experiments are conducted under time-varying speed conditions to verify the accurateness of the proposd models for different gear fault cases.The verified models are further used to test the effectiveness of the proposed nonstationary fault feature extraction methods.All these models and methods are finally used in real world planetary gearbox fault diagnosis applications,and different gear faults are successfully diagnosed.In summary,a series of analytical models and signal analysis methods are proposed in this study to solve the problem of planetary gearbox fault diagnosis urtder time-varying speed conditions.Analytical model of transverse vibration,torsional vibration,as well as stator current signals are established and studied for revealing fault signatures,and a series of methods including synchrosqueezing,time-frequency reassignment,iterative generalized demodulation,and space vector analysis are exploited and improved for dealing with complex multi-component signals of planetary gearbox.These proposals will provide new ideas and solutions for real world fault diagnosis of planetary gearboxes under time-varying speed conditions.