关键词: 核磁测井解谱, 迭代Tikhonov正则化, L—曲线, 正则化参数, 收敛速度

Abstract:

Echo-stripping T₂ relaxation spectra from nuclear magnetic resonance log data is one of the most principal and important jod during NMR logging processing and interpretation,it directly affects the accuracy of geological estimations.Echo-stripping T₂ relaxation spectra is a linear inverse problem.In view of the singularity of the coefficient matrix,regularization methodology can obtain steady solution.In this paper,Tikhonov iteration regularization methodology is applied in order to inverse T₂ spectra,and a posteriority choices of the regularization parameter and proper criteria of the ill-posed problem are fast convergence rate and can process the lower SNR logging data,case study shows that the method is efficient and feasible.Convergence rate with the method is much ,the solution obtained by this method has faster convergence rate and higher accuracy than SIRT,and this method can achieve much better and more stable results in the processing of the low SNR data.

Key words: NMR logging Echo-stripping, Tikhonov iteration regularization, L-curve, regularization parameter, convergence rate