Research On The Technology Of Gear Transmission System Compound Fault Diagnosis

Gear transmission system is a main power transmission device in modern machinery,and often works in the harsh internal and external environments.The gears and bearings,as the core parts of the gearbox,are prone to malfunction.If the fault is not discovered in time,a small incipient fault in gear or bearing may cause compound faults in gear-bearing system.Compound faults are difficult to be completely identified.Therefore,promoting research into the compound fault diagnosis method and explore the effective compound fault diagnosis method,is important for ensuring the stable operation of equipment,reduce the loss and improving theoretical value and application value.This paper takes compound fault diagnosis of gear-bearing transmission system as the research object,researches on the diagnosis method suitable for compound faults of gear-bearing transmission system.The main research work is as follows:The dynamic model of normal state and compound fault state of the multi-stage gear transmission system is established by using the lumped-mass method.The time-varying meshing stiffness is used to describe the normal state of the gear,the tooth root crack and the single-tooth broken fault state.By solving the dynamic differential equations,the simulation vibration signals of the normal state and the compound fault state are obtained.By comparing the simulation signal of normal state with the experimental signal,the rationality of the dynamic model is verified,and by comparing the fault signal with the normal state signal,the fault characteristics of the compound fault state of tooth root cracksingle tooth broken are obtained.A bearing compound fault diagnosis method named EWT-MCKD,which combines empirical wavelet transform(EWT)and maximum correlated kurtosis deconvolution(MCKD)is proposed.EWT is used as a pre-processing method for fault signals to filter out the noise interference in the original fault signal.MCKD is used as the post-processing method to separate and identify the compound fault.The proposed method is used to analyze the simulation signal and experimental signal of bearing compound fault,and realize the identification of single fault in compound fault signal.In order to overcome the insufficiency of MCKD in parameter selection and improve the diagnosis quality,a method of adaptive parameter selection named QGA-MCKD is proposed.QGA is used to automatically search the calculation parameters(period of interest and filter size)of MCKD by considering the interaction effect between two key parameters.Verification is performed on simulation and experimental signals.Results show that QGA can get the best combination of parameters and QGA-MCKD has a good effect in improving the accuracy of planetary gear-bearing compound fault diagnosis.Try to introduce Wilcoxon rank sum test into the field of fault diagnosis,and make a preliminary exploration of the application of the method in fault diagnosis.A pattern recognition diagnosis method based on Wilcoxon rank sum test is proposed as a intelligent diagnosis method for gearbox driving system.The proposed method Wilcoxon rank-sum tests and maximum amplitude selection is used as feature extraction method for vibration signals of different states and to establish the training samples and test samples.The K nearest neighbor method is used as a classifier to classify and identify the fault types.This method is applied to the analysis of experimental signals,and the diagnosis of 5 different types of gear-bearing faults is realized.