Research On Fault Diagnosis And Performance Degradation Evaluation Technology Of Vibration Signal Of Transmission Chain

As an important equipment in rotating machinery,transmission chain shoulders the important task of transfering energy.It will cause performance performance degradation and many kinds of problems when the transmission chain is working for a long time,because it will be easy to aging and wear for the reason of complex structure which involves many mechanical parts.Serious failures can be avoided if the transmission chain can be diagnosed in time;then we can find the failure in advance and carry it out if a reasonable evaluation can be made on the status of its operation on this basis.This paper takes the gears and bearings of the transmission chain as the research object,from the perspective of vibration signals,studies the fault characteristics of gears and bearing outer rings.Aiming at fault diagnosis and performance degradation evaluation,studying the fault diagnosis of each component of transmission chain deeply based on the study of band-pass filtering of transmission chain vibration signal and researching the performance degradation evaluation of transmission chain further.The main research contents in this paper including:(1)Morlet wavelet filter based on improved whale algorithm.The principle and characteristics of Morlet wavelet filter are analyzed,and the noise components are eliminated by using it.In the aspect of filter design and implementation,make the correlation kurtosis(CK)as as the fitness function and the improved whale algorithm(IWOA)is used as the optimization algorithm.At the same time,the periodic update strategy is proposed to solve the periodic dependence when calculating the CK.Through simulation and experiment,it is proved that the filter can effectively obtain the periodic impact components in the vibration signal while filtering the noise.(2)Research on fault diagnosis of transmission chain.Studying three envelope demodulation methods which include Hilbert transform,Teager energy operator(TEO),and differential symmetric analytical energy operator(SD-AEO)from the beginning of enveloping demodulation method and pointing out these methodes difference which are used in fault diagnosis.Considering the fact that the rolling element fault features are easily submerged in the fault diagnosis of isochronous sampling signal,the order spectrum features under equal angle sampling are studied.The experimental results show the fault features of gears and bearings can be clearly observed from the order spectrum.Finally,the traditional autocorrelation function analysis and cepstrum method are studied,and the superiority of the order spectrum method is proved by comparing the experimental results.(3)Research on performance degradation assessment.Uesing local linear embedding(LLE)algorithm to reduce the extracted features which are studied in the time frequency and angle domain that many feature extraction methods are used.On this basis,the performance degradation evaluation model based on SVM is studied,and the performance degradation curve obtained can distinguish different failure degrees but cannot describe the state when the failure occurs.Based on this,the IWOA-SVDD evaluation model is proposed and the model is improved considering the irreversible and continuous characteristics of actual performance degradation.Finally,through the experimental platform,it is verified that the model can describe the entire process from normal to severe failure of the bearing and the degradation curve is better than before.