Ground NMR Inversion Method Based on Dynamic Profile Model

Abstract

The invention discloses a dynamic contour model based surface nuclear magnetic resonance inversion method. The dynamic contour model based surface nuclear magnetic resonance inversion method comprises setting the number of water content layers in a detection area and initializing the thickness and a water content value of every layer in the SNMR (Surface Nuclear Magnetic Resonance) layered inversion solution process; dividing a sub-water-content vertical distribution graph into MN micro elements to meet a matrix equation of a dynamic model; performing iteration solution on the matrix equation and performing dynamic adjustment on the thickness and the water content value of every water content layer in the iteration process to search an optimal solution which meets the matrix equation. The dynamic adjustment is constantly performed on a contour of the vertical distribution graph of the water content values in the detection area in the integral solution process and accordingly the dynamic contour model based surface nuclear magnetic resonance inversion method is called the dynamic contour model and an SGD (Stochastic Gradient Descent) method is designed to solve the model. According to the dynamic contour model based surface nuclear magnetic resonance inversion method, the convergence speed is high, the inversion result is high in accuracy and stable, the performance is excellent in comparison with a regularization method, and the high-accuracy solution on an SNMR inversion problem can be achieved.

Description

Technical field [0001] The invention relates to the field of Surface Nuclear Magnetic Resonance (SNMR), in particular to a ground nuclear magnetic resonance inversion method based on a dynamic profile model. Background technique [0002] Surface Nuclear Magnetic Resonance (SNMR) technology is currently the only geophysical method in the world that directly finds water. This technology has been used in groundwater detection, archaeology, groundwater pollution detection and other fields. [0003] In recent years, with the gradual in-depth research of experts and scholars, SNMR technology has been further improved. The inversion calculation of water content is a key link in the research process of this technology, and the accuracy of the inversion result and the operation efficiency of the inversion process are the key indicators to measure the performance of the inversion algorithm. Among them, the one-dimensional forward and inversion theory is relatively mature, and a variety of e...

AI Technical Summary

Problems solved by technology

[0005] Ground NMR inversion can be abstracted as a problem of solving ill-conditioned matrix equations. In the existing literature, regularization methods are often used to solve it. However, there are two problems in the existing standard regularization methods and improved methods based on them. The deficiencies in the following aspects are: the selection of regularization parameters is difficult; the approximate solution of the inversion problem is forced to add smoothness constraints, which weakens the difference in water content between adjacent geological layers

Images