Font Size: a A A

Application Of EMD To Diagnosis Of Machine Fault

Posted on:2004-01-02Degree:MasterType:Thesis
Country:ChinaCandidate:X L ZhangFull Text:PDF
GTID:2168360092496767Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
Machine fault diagnosis is a science of identifying the running state of device by studying the varying information of the device running state. A new signal processingmethod-EMD(empirical mode decomposition) method has been developed byHuang in 1998. The essence of the method is to identify the intrinsic oscillatory modes by their characteristic time scales in the data empirically and then decompose the data accordingly. EMD method is local adaptive.lt not only can be used to analyse linear data,but also can be used to analyse nonlinear and non-stationary data. Fourier transform method use high order frequency to analyse and imitate nonlinear and non-stationary data. EMD method based on local characteristic time scale of the data is more fit to analyse nonlinear and non-stationary data than Fourier transform method.In this paper,the use of EMD method in machine fault diagnosis is discussed,and two new fault diagnosis methods are introduced.1. Application of EMD and wavelet transform to diagnosis machine faultIn this paper a signal singularity detection method is introduced,with which the signal is decomposed into IMFs(intrinsic mode function) by the EMD(empirical mode decomposition) and the singularity of the IMFs is extracted by the Wavelet transform. Based on the singularity the machine fault can be diagnosed. The numerical and the expriment show that this method is effective.2. Application of EMD and correlative dimension to diagnosis machine fault Fractal dimension can be used to show the structure character of the signal,and iswidely used in denoting the nonlinear system. In this paper one kind of fractal dimension-correlative dimension is discussed thoroughly.And the parameters (include the delay time and the embedding dimension) that affect the reconstruction of phase space and the result of the correlative dimension are discussed thoroughly. In this paper the C-C method is selected to determine the delay time and the embedding dimension. In this paper,the average square root of the signal's first three IMFs is used to diagnosis the machine fault. Use this standard can improve the precision of the fault diagnosis,and reduce the false diagnosis in contrast to using the signal's correlative dimension.
Keywords/Search Tags:empirical mode decomposition (EMD), wavelet transform, phase space reconstruction, correlative dimension, fault diagnosis
PDF Full Text Request
Related items