Research On Fault Diagnosis Method Of Rolling Element Bearing Of Numerical Control Machine Based On Weighted Log-envelope Harmonic-to-Noise Ratio

As the core and supporting part of spindle rotation,rolling element bearing is usually working under the conditions of alternating load,high temperature and heavy load which makes it one of the most vulnerable parts of the rotating machinery.Therefore,condition monitoring and fault diagnosis of the rolling element bearings are the keys to ensure the normal operation of the equipment.Envelope analysis has become one of the most widely used and mature methods in fault diagnosis of the rolling element bearings,but there are still some problems in the selection of demodulation frequency band,which is the critical step of the whole procedure.The selection of demodulation frequency band is easy to be interfered by irrelevant noise and choose the wrong frequency band,resulting in poor diagnosis effect.Therefore,based on the rolling element bearing' vibration signals as the research object,the optimal demodulation frequency band selection of rolling element bearing fault diagnosis technology under complicated interferences as the main research target,this paper proposes a novel method which can select the optimal demodulation frequency band more effectively in the complex cases,and designs a rolling element bearing fault diagnosis software using this method that is combined with Huazhong No.9 intelligent CNC system.The main research contents of this paper are as follows:Aiming at the commonly used classical optimal demodulation frequency band selection methods(Fast Kurtogram,Harmonic-to-Noise Ratio,Ratio of Cyclic Content and Logcycligram)that are based on impulsiveness and second order cyclostationarity of the bearing element fault signal,starting from the background of each method,the principle in selecting the optimal filter band,calculation process,advantages and disadvantages of these methods are introduced systematically.To cover the shortage of the definition of the narrow band in Ratio of Cyclic Content,two types of refined narrowbands are proposed.Combined with the new narrowbands and the key idea of Harmonic-to-noise Ratio,a new index called Cyclic Logarithmic Harmonic-to-Noise Ratio(CLHNR)is proposed on the basis of Log-cycligram.By establishing two thresholds,the new index can not only focus on the fault characteristic frequency but also reduce the interference of independent noise.Considering the existence of multiple faults,an optimal demodulation frequency band selection method based on the improved Log-cycligram is proposed,which is called Weighted Log-envelope Harmonic-to-Noise Ratio(WLHNR).In order to reduce the influence of false detection,the parameters used in this index are discussed in details.By simulating the outer race fault signal,the influences of different types and levels of noise on the demodulation frequency band selection using three methods(Fast Kurtogram,Logcycligram and Weighted Log-envelope Harmonic-to-Noise Ratio)are illustrated.Analyzing the results,the advantages of the new index in selecting frequency band under different noise conditions are verified.By comparing the performances of the three methods with the experimental data of different fault types,it is proved that the method proposed in this paper can correctly diagnose the bearing fault when in practical use.According to the proposed optimal demodulation frequency band selection method,a rolling element bearing fault diagnosis software is designed based on Qt and it can be used in the Huazhong No.9 intelligent CNC system.On the basis of the actual requirements,the overall scheme is determined,and each part of the modules and windows is designed and implemented.