Research On A Fault Feature Adaptive-Optimization Extraction Method For The Rolling Bearing Based On Maximum Correlation Kurtosis Deconvolution

As a key component of rotating machinery,rolling bearings are widely used in rotating machinery.It is one of the most vulnerable parts,and its running state directly affects the health of rotating machinery.Therefore,it is of great significance to monitor the running state of rolling bearings and make regular fault diagnosis.Taking the rolling bearing as the research object,this paper proposes a fault feature adaptive-optimization extraction method for the rolling bearing based on maximum correlation kurtosis deconvolution(MCKD).The main research contents of this paper are as follows:(1)The influence of each parameter in MCKD on the algorithm result is analyzed,and an optimal selection method of the MCKD parameters based on a comprehensive index of fault feature is proposed to optimize the selection of the deconvolution period T for the MCKD algorithm parameters.In addition,the selection criteria of the shift number M and the filter length L are also described,and the optimal selection of all three parameters of the MCKD algorithm is completed.(2)The adaptive optimal selection method of the MCKD parameters is proposed on the basis of the cuckoo search algorithm,which realizes the adaptive optimal selection of the deconvolution period T.At the same time,the adaptive optimal selection results of the particle swarm optimization algorithm are compared and analyzed to illustrate the superiority of the search results of the cuckoo search algorithm.(3)The selection criterion for the IMF components of the EEMD algorithm processing results is given,and a fault feature extraction method for the rolling bearing combining EEMD and adaptively parameter-optimized MCKD is proposed.By using a bearing fault simulation signal,EEMD and adaptively parameter-optimized MCKD are carried out respectively to extract the rolling bearing fault feature,and the effectiveness and superiority of the method are verified.(4)By using the inner ring fault experimental data and outer ring fault experimental data of the rolling bearing measured by the rolling bearing test bench,the inner ring fault feature and outer ring fault feature are extracted according to the method of EEMD and adaptively parameter-optimized MCKD,which further demonstrates the effectiveness and reliability of the method.