A Study On Low Speed Bearing Fault Diagnosis Based On Acoustic Emission Technique

Bearing is an indispensable part of mechanical equipment;its stable operation has a great impact on the reliability of the machinery and the safety of the operator.The fault diagnosis of the rolling element bearing plays a very important role in the industrial production.Current technologies used in condition monitoring and fault diagnosis of rolling element bearings include vibration monitoring technology,lubricating oil composition analysis technology,Acoustic Emission(AE)technology,and ultrasonic detection technology.Among them,the vibration monitoring technology is the most mature and most widely used technology in the field of rolling element bearing fault diagnosis.Vibration monitoring technology in the low-speed rotating machinery such as wind turbines,large cranes and material transmission machine and other low speed machine,usually have an unsatisfactory effect owing to their large moment of inertia,strong background noise,weak vibration energy.In order to overcome this problem,the application of high frequency acoustic emission technology in low speed rotating machinery monitoring has attracted more and more attention of the relevant researchers and engineering personnel.In general,the processing of acoustic emission signals is mainly focused on signal de-noising and time-frequency analysis.Common signal de-noising methods are wavelet de-noising,EMD de-noising,singular value decomposition and high frequency filtering.The time-frequency analysis of the signal mainly depends on Fourier transform and envelope analysis.Now,in order to improve the accuracy of the fault diagnosis of bearings,acoustic emission signal analysis has gradually developed to the direction of intelligent.The commonly intelligent diagnostic techniques include support vector machine,artificial neural network,fuzzy c-means clustering,and genetic algorithm and so on.In this paper,based on the previous experience in the diagnosis of acoustic emission signals of the rolling element bearings,the intelligent diagnosis of low-speed bearing fault is realized from three aspects: signal de-noising,time-frequency domain feature analysis and artificial intelligence identification.The main tasks accomplished by this subject are as follows:(1)Studying the acoustic emission technology and apply it to low-speed bearing fault diagnosis.Analyze the characteristics of acoustic emission signals generated bydifferent bearing faults.The bearing fault characteristics are obtained by the wavelet transform,wavelet packet decomposition,empirical mode decomposition and ensemble empirical mode decomposition.Then,we briefly discussed the advantages and disadvantages of these methods.(2)Carrying out rolling element bearing fault simulation experiment.The low-speed rolling bearing experiment was carried out on the SQI mechanical fault test platform.The vibration and acoustic emission signals of the rolling element bearing,which is under normal working condition,outer race fault,inner race fault and rolling element fault respectively,were recorded.And then,by the analysis of the signals that we recorded with the bearings working under the conditions mentioned above,we can get the signal characteristics.(3)A low-speed bearing fault diagnosis method based on Empirical Mode Decomposition(EMD)-Clear Iterative Interval-Thresholding(CIIT)and Kernel-based Fuzzy C-Means(KFCM)algorithms is proposed,and this method is applied to low-speed bearing fault diagnosis.