Research On Fault Diagnosis Method Of Motor Bearing Based On Tacholess Order Analysis

Rotating machines,such as turbines,pumps,engines,and electric machines,are widely used in household appliances,vehicles,and industrial applications.Rolling bearing is one of the most widely used and easily damaged mechanical parts in the motor equipment,especially in the start-stop stage where the load is relatively large and the rotation speed changes violently.Therefore,in order to solve the problem of bearing fault diagnosis under non-stationary condition,order analysis(OA)has been developed,which has advantages over other methods.Traditional methods often need additional keyphasor devices or speed sensors,which will take up extra space and cost,and increase the complexity of the system.It is necessary to study the convenient,efficient and low-cost variable speed motor bearing fault diagnosis method.The strategy of tacholess OA(TOA)is a developing trend.Therefore,from the point of view of convenience,high efficiency and low cost,this paper focuses on the study of frequency estimation methods of current and vibration signals,and takes the bearings of brushless direct current motor(BLDCM)and permanent magnet synchronous generator(PMSG)which are commonly used in industry as the research object.Finally,a series of convenient and highefficiency TOA methods are proposed for variable speed motor bearing fault diagnosis.First,from the point of accuracy and efficiency,an intelligent method based on the synchrosqueezing wavelet transform(SWT)and the TOA technology of current combined with vibration signal is proposed for the bearing fault diagnosis of the AC motor under the variablespeed condition.This method accurately extracts the mechanical rotation angle curve required by the OA through the SWT for the current signal,then the vibration signal is resampled to realize the fault diagnosis of the motor bearing.The final result is verified in the BLDCM and the PMSG with different fault bearings.Through the comparative experiment,the accuracy,efficiency and generalization of the proposed method are verified.Aiming at the problems of inconvenient installation and high cost existing in the traditional order analysis method,this paper proposes a TOA algorithm based on vibration signal to diagnose the bearing fault of wind turbine.This method only extracts the rotation angle necessary for OA through time-frequency analysis of vibration signal,which is convenient and efficient.On this basis,a bearing fault diagnosis scheme based on multi-sensor signal fusion is proposed to improve the accuracy and effectiveness,which still only needs to measure the threeaxis vibration signals by three-axis accelerate sensors,then the frequency information of the three-axis vibration signals is extracted,the frequency information is screened and fused by using the correlation coefficients.Finally,the original vibration signals are resampled by the fused rotating angle,and the final results of the OA are verified by the wind PMSG test bench,the results show that the inner and outer circle faults have good diagnosis effect,and the angle error extracted is also reduced.To sum up,the OA method based on current signal has higher accuracy and better applicability.And two kinds of OA methods based on vibration signal solve the problem of convenience and high cost.In addition,the common method has high efficiency,and the multisensor information fusion method has high accuracy.Three schemes complement each other,which can realize a variety of circumstances of motor bearing fault diagnosis quickly and accurately.