Study On Fault Feature Extraction Scheme Of Planetary Gearbox Bearing

Planetary gear box is widely used in practical work and production because of its compact structure,high transmission efficiency,smooth operation,large transmission ratio and high bearing capacity.There is usually not only one planetary bearing in the planetary gear box.once one of them breaks down,it is easy to cause uneven planetary gearbox to move unevenly.If it cannot be detected in time and continues to operate under its fault state,it will easily cause more serious consequences.Therefore,it is of great academic significance and engineering application value to carry out the research in the field of online monitoring and fault diagnosis.Compared with single bearing and fixed shaft gear box bearing,it is more difficult to extract the fault feature of planetary bearing in planetary gearbox.Firstly,the fault dynamic signal of planetary bearing is weak and easy to be annihilated by other strong non-bearing signal.Furthermore,the planetary gearbox has multiple meshing and rotating parts inside,and the coupling of multiple vibration sources results in mutual modulation of vibration signals and complex frequency components.Thirdly,there are usually many identical planetary gears and planetary bearings inside the planetary gear box,whose dynamic signals are similar and not easy to be separated.Finally,because of its planetary motion,the path of the dynamic signal from the planetary bearing to the sensor is time-varying.In the relevant research at the present stage,the fault diagnosis and feature extraction technology system for single bearing fault and bearing in fixed shaft gear box has been relatively mature,but due to the above reasons,these methods are not ideal when applied directly to the fault feature extraction of planetary bearing.In order to explore a more suitable fault feature extraction method of planetary bearing and improve the fault diagnosis efficiency of planetary bearing,this paper studies the fault feature extraction method of planetary bearing.In this paper,from the two directions of signal acquisition and signal processing.Different methods for fault feature extraction of planetary bearing are proposed,and apply separately to vibration signal and acoustic emission signal.In terms of vibration signal,aiming at the problem of multi-vibration source coupling of planetary gear box,based on the periodic signal and random signal of gear vibration signal and bearing vibration signal respectively,the Self Adaptive Noise Cancellation(SANC)technology was proposed.For the weak fault signal of planetary wheel bearing,the Time-delay Feedback Stochastic Resonance(TFSR)is used to enhance the relevant signal.For the problem of time-varying transmission path of planetary bearing vibration signal,the Multi-point Optimized Minimum Entropy Deconvolution Adjustment(MOMEDA)method is used to reduce noise and eliminate the influence of transmission path.Finally,combined with Fast Spectral Kurtosis(FSK)algorithm,envelope analysis is carried out to extract fault features of planetary bearing.In terms of acoustic emission signal,based on its advantages such as high sensitivity,high sampling frequency and clear components of high-frequency signals,Fast Spectral Kurtosis algorithm was improved in adaptability,and the fault feature information of planetary wheel bearing was successfully extracted in the high-frequency part of acoustic emission signal.Through experimental research and comparative analysis,the validity of the method and technical route proposed in this paper is verified,and the feature extraction results of each method are analyzed and evaluated,and the advantages and disadvantages and application scope of each method are summarized.