低信噪比核磁共振回波数据降噪方法OA

Nuclear magnetic resonance(NMR)logging can effectively evaluate petrophysical parameters of reservoirs,providing an important basis for oil and gas resource exploration.It exhibits distinct advantages in the identification and quantitative evaluation of fluids.However,unconventional reservoirs are characterized by low porosity,resulting in weak fluid signals detected by NMR logging,low signal-to-noise ratio of echo data,and high uncertainty in the inverted T2 spectra,which in turn affects the reliability of NMR logging formation evaluation.To address these issues,this paper proposes an SGMD-Hurst method combining symplectic geometry mode decomposition(SGMD)and the Hurst index for noise reduction of low signal-to-noise ratio NMR echo data.In this method,the echo data are first decomposed into multiple symplectic geometry components(SGC)using SGMD.Secondly,the Hurst index is calculated for each SGC,and effective SGCs with long-range correlation are screened out based on the Hurst index.Finally,the selected effective SGCs are reconstructed to obtain denoised echo data that retain the characteristics of the original signal while effectively suppressing noise.On this basis,the denoising performance and adaptability of the SGMD-Hurst method are analyzed using numerical simulations and lowsignal-to-noise ratio NMR logging data respectively.The results show that:①Compared with the T2 spectra inverted from the original echo data,as well as data denoised by empirical mode decomposition and the traditional SGMD method,the T2 spectra inverted from the SGMD-Hurst denoised echo data have clearer peak shapes and can more accurately distinguish the positions between the irreducible water peak and the movable fluid peak.②In numerical simulations,the porosity inverted and calculated from the echo data denoised by the SGMD-Hurst is closer to the true value of the model,with a lower root-mean-square error.③In the processing of NMR logging data,the SGMD-Hurst method can still restore the main distribution interval of the T2 spectrum under low-SNR conditions,and the calculated porosity is more consistent with core analysis data than that obtained by traditional denoising methods.It is concluded that the SGMD-Hurst method effectively improves the quality of low signal-to-noise ratio NMR echo data and could provides reliable data preprocessing technical support for the fine evaluation of low-porosity and low-permeability oil and gas reservoirs.