Based On EMD And Wavelet Transform Nuclear Magnetic Logging Echo Signal Denoising Research

Nuclear magnetic resonance logging technology can measure rich strata information,thus we can obtain accurate evaluation of strata formation.But the echo signals generated bynuclear magnetic resonance (NMR) logging is very weak, with a lot of noise, so echo signal islow signal noise ratio (SNR).Therefore, researching the echo signal denoising to obtain moreuseful information is a very important step in nuclear magnetic resonance (NMR) logging.Inthis paper, the main work is as follows:Firstly, the paper made a detailed introduction of wavelet de-noising theory in detail, and useMATLAB to denoise the simulation signal by wavelet threshold algorithm. Then the paperstudies the denoising effect which is about the selection of threshold function, the selection ofwavelet basis function and wavelet soft threshold and hard threshold selection in the waveletthreshold denoising. Besides these, the paper analyzes the propagation characteristicsof modulus maxima of signal and noise. the maximum decomposition levels by signalsimulation are analyzed, and discusses the advantages and disadvantages of the waveletthreshold denoising and wavelet modulus maxima denoising. Then the paper use the waveledenoising method to study the denoising effect of nuclear magnetic resonance(NMR) loggingecho signal from the field.Secondly, this paper studies the basic theory of empirical mode decomposition and thedenoising process which based on empirical mode decomposition. In view of endpoint effectquestion of the empirical mode decomposition, this paper puts forward a method based onlocal extremum continuation and endpoint judgment. the simulation experiment shows thasthe method can effectively restrain the endpoint effect., and its application in echo signaldenoising theory has obatainde the good effect. Beside these, this paper use the empiricalmode decomposition which based on improving to denoise the nuclear magneticresonance(NMR) logging echo signal from the field. The denoising effect is good.Finally, the paper study the production factors of noise of the echo signal in nuclearmagnetic resonance (NMR) logging.According to the theory of wavelet denoising thresholdand the selection problem of wavelet basis function, empirical mode decomposition methoddenoising is rough, it has the problem such as loss of useful information, combined with theadvantages of both, it is proposed denoising method based on empirical mode decompositionand the combination of wavelet transform, and use this method to denoise the nuclearmagnetic resonance(NMR) logging echo signal from the field. The denoising results showsthat the fusion method has the better denoising effects, the SNR improve obviously, the usefulinformation are further improved.