A Novel Signal Identifying Method Of Oil-Gas-Water Phase Fractions In Nmr Measurement

Wuming Zhu,Jiang Yang,Jiajun Lao
2015-01-01
Abstract:Free induction decay (FID) signal of oil-gas-water nuclear magnetic resonance (NMR) responses contain three phase fraction information. A novel signal identifying method including filter and fitting algorithm to determine the oil-gas-water fractions from FID has presented in this paper. The filter algorithm based on empirical mode decomposition (EMD) and Pearson correlation coefficient (PCC) is used to restrain noise of the FID signal. EMD can transfer a signal into several intrinsic mode functions (IMF) that contains different noises, PCC abrupt increasing change indicates demarcation point after which the IMF are signal leading. A reconstruction function for those signal leading IMFs can be regard as the FID signal that measurement noises have been restrained. The segmental fitting algorithm for coefficients of FID has smaller error than that of direct fitting algorithm. MATLAB simulations show the phase amplitude estimating errors are less than 10% that can satisfy with calculating oil-gas-water fractions.
What problem does this paper attempt to address?