Equal Frequency Division Method of Seismic Data Spectrum Based on Multiple Wavelets

Abstract

The invention provides a seismic data frequency spectrum equalization frequency division method based on multi-wavelet. The method comprises the steps that a seismic wavelet sequence is generated; multi-wavelet seismic trace data are decomposed; a wavelet or wavelet range involved in frequency division calculation is selected; the frequency spectrum and the amplitude envelope of the selected wavelet involved in frequency division calculation are calculated; equalization calculation is carried out on the frequency spectrum of the selected wavelet involved in frequency division calculation; the frequency division of the selected wavelet which is involved in frequency division calculation and undergoes frequency spectrum equalization is calculated; and the frequency division of the selected wavelet which is involved in frequency division calculation and undergoes frequency spectrum equalization is superposed. According to the method provided by the invention, the lateral resolution of the frequency division result is improved; the boundary distribution and changes of the underground rock stratum and physical characteristics are accurately reflected; and a more reliable scientific basis is provided for the interpretation and prediction of oil and gas resources.

Description

technical field [0001] The invention relates to an equal frequency division method of seismic data spectrum based on multiple wavelets, which belongs to the technical field of seismic data processing and interpretation in the exploration of seismic resources (petroleum, coal, minerals, etc.). Background technique [0002] Seismic data frequency division is to extract the energy of a given (frequency division) frequency from a seismic trace or seismic signal. This energy varies with time. After calculating the frequency division results for a section or data volume according to a given frequency division frequency, it can be used for display, further interpretation and analysis against known drilling results, and then for the corresponding formation characteristics and possible oil and gas Based on the spatial distribution and changes of features, further speculation can be made to improve the accuracy of oil and gas prediction. [0003] Currently commonly used frequency di...

AI Technical Summary

Problems solved by technology

[0004] For the STFT-based frequency division method, once the time window is selected, the resolution in time and frequency is fixed, and this single time-frequency resolution is not conducive to distinguishing low-frequency seismic signals and high-frequency seismic signals that exist simultaneously.

[0005] For the frequency division method based on wavelet transform, the obtained wavelet coefficients belong to l 2 Norm optimization, that is, to find the reflection coefficient The minimum value of is non-sparse decomposition, and the resolution of the seismic wavelet in the seismic trace is poor, so the frequency division result obtained has the problem of poor resolution

[0006] For the frequency division method based on matching pursuit seismic trace decomposition, it is a non-global optimization algorithm. The result of the decomposition is not unique and can change with the order of matching wavelets. It is also a non-sparse decomposition with poor analytical ability.

[0007] The above frequency division calculation method can only calculate the frequency division of a given frequency division frequency of the original seismic data, and cannot effectively remove the influence of geological background and possible noise

In addition, when the main frequency of the wavelet obtained by the multi-wavelet decomposition of the seismic trace is high, the frequency division result based on the multi-wavelet has a relatively low horizontal or spatial resolution due to the wide frequency spectrum of the wavelet (as shown in Figure 3). Difference

Images