Research On Local Mean Decomposition And Its Application In Rolling Bearing Fault Diagnosis

Rotating machinery is an important composition part of modern industrial production equipment,and it is also an important categories of various machinery equipments.In all kinds of rotating machineries,rolling bearing plays an important role.To guarantee the normal operation of machinery equipment,condition monitoring and fault diagnosis faced to rolling bearing of rotating machinery has become an important research subject.Local Mean Decomposition(LMD)is a new kind of adaptive signal analysis method that appeared in recent years.Due to it's advantages in nonstationary signal processing,this method has been widely used in rolling bearing fault diagnosis field.Focusing on LMD method and it's application in rolling bearing fault diagnosis,main contents of this paperare as follows:First,aiming at the mode mixing problem existing in the decomposition process of LMD,a method of Masking Local Mean Decomposition(MLMD)based on masking signal and correlation coefficient is studied.The feasibility of this method is verified by simulated signal and actual vibration signal of rolling bearing.Then,aiming at the limitations existing in the coarse graining process of Multi-Scale Permutation Entropy(MPE),Composite Multi-Scale Permutation Entropy(CMPE)is studied.Combining the CMPE with LMD and synthetical associatied analysis,a rolling bearing fault diagnosis method based on LMD-CMPE and synthetical associatied analysis is studied.In this method,the rolling bearing vibration signal is decomposed into several Product Functions(PF)adaptively,and the PF components that contain the main fault information are selected to calculate the CMPE as fault feature vectors.Introducing proximity factor into grey similar incidence,this method takes both similarity and proximity into consideration to identify the CMPE feature vectors,and realizes the recognition of different fault types.Finally,from the perspective of instantaneous characteristic analysis,based on the characteristics of LMD,the instantaneous energy spectrum based on LMD is put forward,and then,introducing the concept of information entropy into the instantaneous energy spectrum,a new fault feature parameter,instantaneous energy spectrum entropy is proposed.Combined with mathematical morphology denoising,the proposed method is applied to the feature extraction of bearing signals.Instantaneous energy spectrum based on LMD can accurately reflect the fault feature frequency,and the instantaneous energy spectrum entropy has stable differentiation degree on the signals of different running states,as a kind of fault feature parameter,it can effectively distinguish different types of faults.