Research On Early Weak Fault Diagnosis Method Of Rolling Bearing And Its System Implementation

Rolling bearings are important components in rotating machinery and greatly affect the operation of mechanical equipment.Rolling bearings often work in high temperature and high pressure environments and are subjected to high bearing capacity,so they are also one of the most vulnerable parts of mechanical equipment.The failure of mechanical equipment caused by rolling bearing damage often leads to serious property damage and safety accidents.Early detection of faults can reduce losses as much as possible.In the early failure stage,the fault impact is weak,so extracting the weak fault feature from the noise is the main research content of early fault diagnosis.For weak fault diagnosis of rolling bearings,this paper mainly studies the bearing vibration signals.The main research contents include:signal preprocessing,fault feature extraction and pattern recognition diagnosis.In terms of signal preprocessing,the application of EMD,LMD and VMD in rolling fault diagnosis is analyzed theoretically and experimentally..The advantages and disadvantages of the three methods are compared and the feasibility of the three methods in early fault diagnosis.In view of the shortcomings of the LMD method in early fault diagnosis,the MCKD method is combined with the LMD method,and then the time-frequency analysis is used to judge the fault type,and the accuracy of the LMD method in early fault diagnosis is improved.The feature extraction inverse surface,using the bispectrum analysis to suppress the Gaussian noise,extract the high-order eigenvectors of the vibration signal,and verify the feasibility of bispectrum analysis in fault diagnosis.For the non-Gaussian noise component can not be processed by bispectrum analysis,and the fault impact in the early fault stage is susceptible to other noise components,an early fault diagnosis method combining MOMEDA and bispectrum analysis is proposed.Finally,according to the symmetry of bispectrum analysis,further The data is compressed to extract the fault feature vector.Finally,use the SVM to complete the fault type diagnosis.Finally,the theory combined with practice,designed and developed a simple Windows desktop application for rolling bearing fault diagnosis,the system combined with the main methods of this research,making the theoretical method more practical.