Research on Multi-stage Noise Reduction Method of NMR Logging Signal

Abstract

By inversion calculation of T2 spectrum of echo string, the permeability, saturation, fluid type and other important parameters of porous media can be obtained. However, since the inversion algorithm itself is susceptible to noise signal [1], improving the SNR level of echo signal has been one of the key problems in the application of low-field NMR technology. In order to obtain more accurate analysis results, the corresponding noise reduction method needs to be developed. In the past, the research on noise reduction of NMR signal mainly focused on echo string. This paper constructs a complete NMR signal model and discusses the de-noising of echo and echo string. Aiming at noise reduction of echo, FIR filter and Orthogonal Modulation Filter(OMF) are discussed. Aiming at the de-noising of echo string, this paper proposes an improved CEEMD threshold de-noising, and makes a comparative analysis with the heuristic threshold de-noising based on sym4 wavelet and EMD threshold de-noising proposed by predecessors. It is found that the method proposed in this paper can invert the more accurate T2 spectrum while improving the SNR.