Study on Inversion Algorithm of NMR T2 Spectrum

Nuclear magnetic resonance (NMR) logging technology is a kind of developed gradually in recent decades, which has great significance for the development technology. compared to traditional technology, not only ideal signal - hydrogen nuclei, and exploration original affected by influence rock skeleton, provides information that unrelated formation porosity, can better provide permeability, movable fluid, oil gas content, such as information, It method suitable low porosity permeability complex special lithology, overcomes defects methods .With NMR theory, been applied analysis well data. The data are spin-echo string observed spin echo string, its amplitude decays with time. T2 spectrum be inversely calculated from measured data, then further solved obtain key porosity. In this paper, simulated constructed forward modeling, inversion carried out under different SNR. On basis, specific relationship between optimal number singular values retained change SNR summarized, an improved value decomposition algorithm proposed. Inverse problem converted nonlinear fitting optimization problems nonnegative constraints, linear truncation on calculation precision increased significantly, algorithm, new signal-to-noise ratio solution smaller stability, through improvement constraint processing structure was more rapid iteration algorithm.