WVD Based On Dual EEMD And Using In Vibration Signal Analysis

Wigner-Ville distribution(WVD) is widely used in the filed of signal processingduing to all kinds of excellent mathematical characteristics. However, the cross-terms ofthe WVD often bring us lots of problems, because the WVD is not additive in thecalculation. Aiming at this problem, cross-terms erasing method based on dual EnsembleEmpirical Mode Decomposition(EEMD) is put forward in this paper. This method firstlydecomposes the original signal into several Intrinsic Mode Functions(IMFs) by EEMD.Then deal with the IMFs obtained with the EEMD again and now a group of new IMFscan be gained. After calculating each of the IMFs’ WVD we can gain the WVD of originalsignal by adding the results together. The method presented in this paper effectively avoidsthe cross-terms of WVD with less mode mixing while keeps the advantages of WVD.Computer simulation and experimental study indicates that this method could eliminatethe cross-terms effectively.In order to induce the steps and increase the efficiency when analyze the fault signals,this text designed an aided analysis tool for fault signals’ analysis. In this tool, MATLABScript works as a link, which makes LabVIEW and MATLAB work together. This toolcontains Wavelete Analysis, Wigner-Ville Distribute, Hilbert-Huang Transform and othermethods of signal analysis, any-thing else, this tool can also share the database with thesignal collector. Because of that, this aided tool provides an easier communication waybetween users and computer, and that induces the steps and increase the efficiency greatlywhen analyze fault signals off line. Of course, this tool also contains the DEEMD-WVDmethod (Wigner-Ville Distribute based on Dual EEMD, DEEME-WVD), we applied thismethod through the aided tool to collecting and analysis of fault signals in practicalengineering in Chengde steel mill. It turns out that the DEEMD-WVD method is preciseand feasible, and the aided analysis tool of fault singals is practical.