Decomposition method of attenuation signal based on q-value difference reflection

Abstract

The invention provides a decay signal decomposition method based on Q value difference reflection, comprising the following steps: selecting a first matching atom of a to-be-decomposed signal from an over-complete dictionary; decomposing the signal into a component on the first matching atom and a residual; iteratively decomposing the residual obtained through decomposition and judging the residual energy; and when the residual energy is less than a threshold, determining that decomposition is completed. Through the technical scheme, basis functions of a matching pursuit algorithm are constructed based on Q value difference reflection and a frequency domain seismic convolution model, and a decay seismic signal is decomposed using the set of basis functions, so that decay sub-signals of the seismic signal can be reconstructed to the maximum degree, a result of higher precision can be provided for subsequent rock physics analysis, decay parameter estimation or spectrum analysis, and the reliability and accuracy of reservoir oil-gas possibility prediction are improved.

Description

technical field [0001] The invention relates to the field of petroleum geophysical exploration, in particular to an attenuation signal decomposition method based on Q value difference reflection. Background technique [0002] In the traditional seismic exploration process, the attenuation signal contains a wealth of information, including thin layer thickness information, formation attribute information and gas-bearing information. Various time-frequency transformation methods are usually used to describe the non-stationary time signal of the decay signal, including short-time Fourier transform (DFT), maximum entropy method (MEM), continuous wavelet transform (CWT) and matching pursuit decomposition ( MPD) and other methods. The above methods have their own advantages and disadvantages, but none of them can fully utilize the characteristics of seismic convolution forward modeling to obtain more accurate time-frequency spectrum. [0003] Both the short-time Fourier transfor...

AI Technical Summary

Problems solved by technology

The above methods have their own advantages and disadvantages, but none of them can fully utilize the characteristics of seismic convolution forward modeling to obtain more accurate time-frequency spectrum

[0003] Both the short-time Fourier transform and the maximum entropy method explicitly use the window function, so that the resolution of the time-spectrum in the frequency domain and the time domain cannot be optimal at the same time

When the window function grows, the time-spectrum has a higher resolution in the frequency domain, but the resolution in the time domain is relatively reduced; and vice versa

The continuous wavelet transform uses a frequency-related window function, so that the high-frequency components and low-frequency components of the attenuation signal are better described, which is better than the short-time Fourier transform, but it still does not solve the problem when The spectral interference problem faced when there are multiple waves or thin layers in the seismic signal makes the continuous wavelet transform still inapplicable when dealing with such problems

[0004] The traditional matching pursuit method is a signal decomposition method based on the greedy algorithm. On the basis of not considering the seismic forward modeling process, the time-frequency spectrum generated by the signal decomposed by it is superior to the above-mentioned time-frequency analysis methods in terms of sparsity. , but still face the problem of spectral interference

[0005] For the above problems, there is no good solution in the prior art

Images