Reweighted Kurtogram Method And Its Application In Bearing Fault Diagnosis

As an indispensable part of modern industrialization,rotating machinery is widely used in production equipment and daily life.As an important part of connecting and supporting rotation,rolling bearings play a pivotal role in the normal operation of equipment.Especially,with the continuous improvement of the speed of modern rotating machinery,the health status of bearings directly affects the working efficiency and reliability of machinery.Therefore,detecting the operation state of mechanical equipment,finding faults and timely maintenance are of great significance to ensure the normal operation of equipment and prevent economic losses and accidents.There are many problems in rolling bearing health monitoring and fault diagnosis,such as strong vibration signal interference and weak fault characteristic information,while resonance demodulation technology can extract weak fault information from noise interference through band-pass filtering and envelope demodulation.As the most representative method in resonance demodulation technology,the fast Kurtogram(FK)can adaptively select the best band-pass filter parameters,which makes up for the deficiency of manual selection of filter parameters in traditional resonance demodulation technology.Therefore,it has been widely used in the field of rolling bearing fault diagnosis.However,the FK method still has some limitations,mainly reflected in the kurtosis index and frequency band division method.In view of this,this paper has carried out the following in-depth research on the two limitations of the FK method:Firstly,aiming at the problem that the kurtosis index used in the FK method is vulnerable to strong impact interference and becomes invalid,this paper defines a robust index—reweighted kurtosis(RK),and based on this,a reweighted Kurtogram method based on the“Binary-Ternary” wavelet packet transform is proposed.The robustness of the proposed method under strong impact interference is verified by simulation signals and actual signals,At the same time,the comparison with the analysis results of the FK and other methods shows that the proposed method has a better diagnostic effect.Secondly,aiming at the problems of spectral energy leakage and lack of shift-invariance in the frequency band division method used by FK,this paper proposes a new frequency band division method called sub-bands rearranged and ensemble dual-tree complex wavelet transform(SRE-DTCWPT).This method can not only solve the frequency band disorder but also maintain the nearly shift-invariance and almost no presence of frequency aliasing.In addition,the SRE-DTCWPT can also provide more frequency band division forms,which can more effectively capture the fault characteristics in the transition area and prevent frequency band leakage.In view of the above advantages of SRE-DTCWPT,combined with RK,a reweighted Kurtogram method based on SRE-DTCWPT is proposed.At the same time,according to the fault signals in different situations,an envelope analysis framework is proposed to improve the applicability of the method in practical fault diagnosis.The analysis results of simulation signals verify the robustness of the proposed method under various disturbances,and the analysis results of several different fault cases in the bearing data set of Case Western Reserve University further verify that the proposed method can still effectively extract fault features and realize bearing fault diagnosis under complex disturbances.Finally,based on the above research,to provide an application platform for the proposed diagnosis method,a rolling bearing fault diagnosis system integrating data acquisition and storage,fault analysis,and automatic diagnosis is built with the help of the Lab VIEW software development platform,and the proposed method is integrated into the diagnosis system through the Lab VIEW and MATLAB mixed programming.The feasibility and effectiveness of the function of the fault diagnosis system have been verified on the bearing fault simulation test bench of the research group.