Inversion method of nuclear magnetic resonance two-dimensional spectrum
A technology of nuclear magnetic resonance and two-dimensional spectrum, which is applied in the directions of magnetic resonance measurement, measurement using nuclear magnetic resonance image system, measurement of magnetic variables, etc., can solve the problem of slow evolution, achieve good robustness and improve resolution Effect
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Publication Date
- 2013-05-22
- Estimated Expiration
- Not applicable · inactive patent
Smart Images
Figure 1 Figure 2 Figure 3
Abstract
Description
technical field
[0001] The invention relates to signal processing in the field of nuclear magnetic resonance, in particular to a method for nuclear magnetic resonance two-dimensional spectrum inversion. Background technique
[0002] NMR technology has been widely used in the field of energy exploration. NMR logging can quickly and non-destructively provide accurate information for each step in the routine logging workflow such as reservoir fluid identification, petrophysical property evaluation and productivity evaluation. Compared with traditional methods such as electrical logging and acoustic logging, NMR logging is the only method that can identify fluid types without making the fluid flow. Solutions based on one-dimensional logging experiments have obvious shortcomings in efficiency and accuracy, and two-dimensional NMR logging technology has emerged as the times require. Using the D-T2 (diffusion-transverse relaxation) two-dimensional spectrum can quickly and intuiti...
Examples
Embodiment Construction
[0029] figure 1 It is the flow chart of the method for nuclear magnetic resonance two-dimensional spectrum inversion of the present invention, comprising steps: the first step, noise extraction and estimation: use wavelet transform to extract the noise in the acquisition data CPMG echo train and estimate its standard deviation; Step 1, data compression: generate an inversion kernel, and use the rank of the kernel matrix to perform truncated singular value decomposition and reconstruction to complete data compression; step 3, data fitting: regularize the fitting problem of the compressed data, And use Newton's method combined with non-exact one-dimensional search to iteratively solve the regularization factor and inversion spectrum to obtain the inversion spectrum.
[0030] 1. Noise extraction and estimation:
[0031] Such as figure 2 It is the algorithm flowchart of the noise extraction and estimation part. The wavelet transform is introduced in the noise extraction, and t...