Frequency band enhancement method for weak signals in seismic data

Abstract

The present invention provides a method for enhancing weak signal frequency bands of seismic data, comprising: step 1, decomposing a series of frequency band component data by wavelet packet frequency band method; step 2, identifying strong shielding components; step 3, performing empirical mode decomposition in the determined strong reflection dominant frequency band component data, and calculating the instantaneous amplitude, instantaneous main frequency, and instantaneous phase of the strong reflection; step 4, fitting the optimal main lobe wavelet by adopting the main lobe wavelet approximation method controlled by three factors; step 5, subtracting the optimal main lobe wavelet obtained by approximation from the strong reflection dominant frequency band component data; step 6, synthesizing all the frequency band component data to obtain seismic data after weak signal enhancement. The method for enhancing weak signal frequency bands of seismic data can more specifically solve the problem that the energy of the underlying stratum reflection signal is weak and the signal ratio of the seismic data is low, and finally realizes the weak signal frequency band enhancement processing.

Description

Technical Field

[0001] The invention relates to the technical field of oil and gas geophysical data processing, and in particular to a method for enhancing weak signal frequency bands of seismic data. Background Art

[0002] With the exploration and development of conventional oil and gas reservoirs, the types of newly discovered oil and gas reservoirs are becoming more and more complex, and the reservoir burial depth is also increasing, which puts forward higher requirements for the enhancement of weak signals in seismic data. Since igneous rocks, gypsum layers, coal seams, etc. are often developed in the middle and deep layers, these strong reflection strata will produce strong shielding effects, resulting in blank reflections in the lower strata of strong reflection strata, or the reflection signal energy of the lower strata is weak, making it difficult to accurately characterize the underlying oil and gas reservoirs, resulting in large reservoir description errors.

[0003] At present, there are mainly two types of weak signal enhancement methods. The first type is the direct enhancement method of weak signals, which uses methods such as curvelet transform, empirical mode decomposition, compressed sensing theory, etc. to suppress noise in the transform domain, or perform targeted processing on the decomposed signal, and finally perform inverse transform or data synthesis processing, so as to achieve the purpose of enhancing weak energy signals while suppressing noise or improving resolution.

[0004] In the Chinese patent application with application number: 201410225122.0, a method for reconstructing weak seismic signals based on curvelet transform is involved, which belongs to the technical field of seismic data processing for petroleum seismic exploration, and includes the following steps: 1. Determine the noise model and its parameters; 2. Read the seismic data; 3. Perform curvelet decomposition on the read seismic data to obtain the curvelet coefficients of the seismic data; 4. Perform curvelet decomposition on the obtained noise model to obtain the curvelet coefficients of the noise model; 5. According to the set filtering threshold, filter the curvelet coefficients of the seismic data in the curvelet domain to filter out the noise part in the curvelet coefficients; 6. Perform inverse curvelet transform on the filtered curvelet coefficients to reconstruct the seismic signal after noise removal.

[0005] In the Chinese patent application with application number: 201510145941.9, a method for enhancing the energy of effective weak signals of seismic data with high resolution is involved. The method needs to process the stacked or offset seismic data one by one, and perform the following steps on the single-channel seismic data: 1) perform Hilbert transform on the single-channel seismic data to obtain the envelope, time shift and instantaneous attributes of the single-channel seismic data, and establish a Morlet wavelet library in combination with the average wavelet extracted by well-seismic calibration; 2) based on the wavelet library established in step 1), use the matching pursuit algorithm to decompose the effective wavelets constituting the single-channel seismic data; 3) improve the spectral whitening algorithm, and perform adaptive spectrum extension and energy compensation processing on the effective wavelets obtained by decomposition; 4) superimpose the effective wavelets processed in step 3) to obtain the single-channel seismic data with improved resolution and enhanced effective weak signal energy.

[0006] In the Chinese patent application with application number: 201810028520.1, a weak signal extraction algorithm based on compressed sensing is involved, and the weak signal extraction algorithm based on compressed sensing includes: step 1, inputting noisy seismic data dobs, measurement matrix Ф, and dictionary basis matrix Ψ; step 2, collective empirical mode decomposition, i.e., CEEMD decomposition, decomposing dobs into K fixed mode function IMF components; step 3, performing autocorrelation analysis of IMF components; step 4, applying compressed sensing CS denoising method to process IMFk respectively; step 5, reconstructing the IMF components and other components after CS denoising to obtain denoised seismic signals.

[0007] Luo Hongmei et al. (Progress in Geophysics, 2019) performed wavelet multi-level decomposition on the seismic data, then performed empirical mode decomposition on the decomposition results to obtain the intrinsic mode function, and performed orthogonal function normalization processing, and finally reconstructed the seismic data to improve the resolution while strengthening weak signals.

[0008] The second category is the strong reflection layer suppression method, which uses empirical mode decomposition, wavelet packet transform, matching pursuit, spectral decomposition and other methods to approximate the strong reflection layer by matching wavelets, and then removes the strong reflection layer event axis, thereby achieving the purpose of weak signal enhancement.

[0009] In the Chinese patent application with application number: 201210247077.X, a method for enhancing weak seismic signals is involved, which includes: step a, establishing an initial surrounding rock reflection from a target seismic trace; step b, calculating a new surrounding rock reflection based on the initial surrounding rock reflection; step c, judging whether the new surrounding rock reflection is stable, and when the new surrounding rock reflection is stable, outputting the new surrounding rock reflection as a surrounding rock signal; and step d, subtracting the surrounding rock signal from the seismic data of the seismic trace to obtain a weak signal.

[0010] In the Chinese patent application with application number: 201510290771.3, a model-based strong reflection separation method and device are involved, the method comprising: establishing a strong reflection model according to actual well data; performing three-instant parameter analysis on the strong reflection event axis to obtain initial sub-wave control parameters; optimizing the control parameters to obtain the sub-wave that best matches the strong reflection event axis; subtracting the best matching wavelet from the original seismic data to obtain data after strong reflection separation, and drawing model test conclusions; using the test conclusions for actual data analysis to perform strong reflection separation; performing well bypass analysis and adjusting parameters; determining the final matching wavelet parameters to obtain the strong reflection separation results for the entire area.

[0011] In the Chinese patent application with application number: 201610581152.4, a method for stripping a strong shielding layer based on the analysis of the seismic response characteristics of coal seams is involved, which includes the following steps: spectrum analysis of seismic data to determine the main frequency and frequency band range of seismic signals; layer time tracking of the strong shielding layer; calculation of instantaneous attribute values ​​of seismic data as the initial scale factor of matching tracking; obtaining wavelet amplitude and performing wavelet matching analysis; and stripping the strong shielding layer using a multi-channel dynamic matching tracking method. After stripping the strong shielding using this method, the hidden effective weak signal can be highlighted, providing more effective data for subsequent seismic reservoir prediction.

[0012] In the Chinese patent application with application number: 202010013148.4, a phase-controlled seismic strong shielding time-frequency information extraction and stripping method is involved, including: step 1, well seismic calibration and layer tracking of seismic data; step 2, logging phase and wellside seismic phase analysis of typical wells, and rough division of phase belts; step 3, extracting seismic attributes within the upper and lower time window ranges of the determined layers, screening and combining the characteristic attributes with good performance, and determining an accurate sedimentary phase belt map; step 4, performing generalized S transform on the seismic data of the target processing layer to extract algorithm parameters; step 5, determining the variable sub-wave information involved in the algorithm; step 6, matching the best sub-wave according to the waveform characteristics of the multi-phase belt for strong layer stripping.

[0013] At present, both of these two types of weak signal enhancement methods have shortcomings. The first type of weak signal direct enhancement method: due to the adverse effects of strong reflection strata such as igneous rocks, gypsum layers, and coal seams, the energy of the reflected signal of the underlying strata is weak, and the signal ratio of the seismic data is low. When the decomposed signal is subjected to noise suppression or resolution improvement processing, the stronger energy signal can achieve ideal results, but it is not suitable for weak signal enhancement under strong shielding conditions. The second type of strong reflection layer suppression method: the method of signal decomposition of seismic data has good adaptability. Under normal circumstances, strong reflections are suppressed to achieve the purpose of enhancing weak signals; however, the conventional signal decomposition method also has shortcomings, that is, strong reflection suppression of full-band data will affect the fidelity and amplitude preservation of the data, which is not conducive to subsequent reservoir description and fluid identification.

[0014] The above existing technologies are all significantly different from the present invention and fail to solve the technical problem we want to solve. Therefore, we have invented a new method for enhancing weak signal frequency bands of seismic data. Summary of the invention

[0015] The purpose of the present invention is to provide a method for enhancing weak signal frequency bands of seismic data, which can solve the problem that the energy of the reflected signal of the underlying stratum is weak and the signal ratio of the seismic data is low, and finally realize the weak signal frequency band enhancement processing.

[0016] The purpose of the present invention can be achieved by the following technical measures: a method for enhancing the weak signal frequency band of seismic data, the method comprising:

[0017] Step 1, decomposing a series of frequency-band component data by wavelet packet frequency-band method;

[0018] Step 2, identifying the strong shielding component;

[0019] Step 3, performing empirical mode decomposition on the determined strong reflection dominant frequency band component data to calculate the instantaneous amplitude, instantaneous main frequency, and instantaneous phase of the strong reflection;

[0020] Step 4, using a main lobe wavelet approximation method controlled by three factors to fit the optimal main lobe wavelet;

[0021] Step 5, subtracting the optimal main lobe wavelet obtained by approximation from the data of the strong reflection dominant frequency band component;

[0022] Step 6: synthesize all sub-band component data to obtain seismic data after weak signal enhancement.

[0023] The purpose of the present invention can also be achieved by the following technical measures:

[0024] The method for enhancing weak signal frequency bands of seismic data also includes, before step 1, collecting seismic data and determining an analysis time window.

[0025] In step 1, after determining the seismic data processing time window, the original seismic data is decomposed into a series of sub-band component data in different frequency ranges by using the wavelet packet sub-band method. The wavelet packet sub-band formula is as follows:

[0026]

[0027] Among them: XBB i(t) is the decomposed frequency band data, i = 1, 2, ..., N, including N frequency band component data in total, d(t) is the original seismic data, t is the time series, m(A, f, φ) is the decomposed wavelet, A is the wavelet amplitude value, f is the wavelet frequency value, and φ is the wavelet phase value.

[0028] In step 2, the strong shielding component is identified, that is, the energy ratio of strong reflection and weak reflection in the frequency band component data is calculated, and the dominant frequency band component data of strong reflection is selected for processing, while other frequency band component data are not processed, so as to ensure the relative relationship of the reflection coefficient to the greatest extent.

[0029] In step 2, the strong shielding component identification formula is:

[0030] J=Max[EnUp i (t) / EnDown i (t)],i=1,2,……,N

[0031] Where: J is the selected strong reflection dominant frequency band component data number, EnUp i (t) is the frequency band data XBB i (t) strong reflected signal energy, EnDown i (t) is the frequency band data XBB i The weak reflection signal energy of (t), i = 1, 2, ..., N, includes N sub-band component data in total.

[0032] In step 2, one or two strong shielding components are selected for processing according to the characteristics of the seismic data, and the strong reflection dominant band component data is XBB J (t), J is the data number of the selected strong reflection dominant frequency band component, and the data of other sub-band components are not processed to ensure the relative relationship of the reflection coefficient to the greatest extent.

[0033] In step 3, during the strong reflection suppression process, empirical mode decomposition is performed on the determined strong reflection dominant frequency band component data. The empirical mode decomposition formula is:

[0034]

[0035] Among them: XBB J (t) is the data of the strong reflection dominant frequency band component, imf j (t) is the intrinsic mode function component, j = 1, 2, ..., M, including M intrinsic mode function components, re m (t) is the residual component of empirical mode decomposition.

[0036] In step 3, the first intrinsic mode function component imf of the strong reflection dominant frequency band component data is selected 1(t), calculate the instantaneous amplitude, instantaneous main frequency, and instantaneous phase initial value of the eigenmode function component.

[0037] In step 4, according to the calculated strong reflection instantaneous amplitude, instantaneous main frequency, and instantaneous phase initial value, the main lobe wavelet approximation method controlled by three factors is used to fit the optimal main lobe wavelet. The initial main lobe wavelet expression is:

[0038] in: is the instantaneous amplitude of the component data, is the instantaneous frequency of the component data, is the instantaneous phase of the component.

[0039] In step 4, under the control of the three factors of instantaneous amplitude, instantaneous main frequency and instantaneous phase, the optimal main lobe wavelet is determined according to the characteristics of the seismic data, and the optimal main lobe wavelet obtained by fitting and the strong reflection dominant frequency band component are close to the minimum sum of the minimum squares.

[0040] In step 5, the optimal main lobe wavelet obtained by approximation is subtracted from the strong reflection dominant frequency band component data to complete the suppression of the strong reflection component data that needs to be processed.

[0041] The weak signal frequency band enhancement method of seismic data in the present invention uses the wavelet packet frequency band method to identify strong shielding components in different frequency band data, and only suppresses the dominant frequency band component data with strong reflection, thereby ensuring the relative relationship of the reflection coefficient to the greatest extent. In addition, the main lobe wavelet approximation method controlled by three factors, namely, instantaneous amplitude, instantaneous main frequency and instantaneous phase, is adopted to subtract the optimal main lobe wavelet from the strong shielding component data, thereby realizing weak signal frequency band enhancement processing under the premise of high fidelity and amplitude preservation.

[0042] The present invention decomposes the original seismic data into a series of sub-band component data in different frequency ranges by using the wavelet packet sub-band method, automatically identifies the strong shielding components in the different sub-band data, automatically selects the dominant frequency band component data with strong reflection for processing, and does not process the other sub-band component data, thereby ensuring the relative relationship of the reflection coefficients to the greatest extent. The processed seismic data has high fidelity and amplitude preservation.

[0043] Empirical mode decomposition is performed on the determined strong shielding component data, and the main lobe wavelet approximation method controlled by three factors, namely, instantaneous amplitude, instantaneous main frequency and instantaneous phase, is adopted to subtract the optimal main lobe wavelet from the strong shielding component data without destroying the weak reflection of the underlying strata of the strong shielding. This can better overcome the adverse effects of strong reflection strata such as igneous rock, gypsum layer and coal seam, and more specifically solve the difficult problems of weak reflection signal energy of the underlying strata and low signal ratio of seismic data, and ultimately realize weak signal frequency band enhancement processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a flow chart of a specific embodiment of the method for enhancing weak signal frequency bands of seismic data of the present invention;

[0045] Figure 2 A schematic diagram of a seismic profile before processing in a specific embodiment of the present invention;

[0046] Figure 3 A schematic diagram of a strong shielding component seismic profile in a specific embodiment of the present invention;

[0047] Figure 4 It is a schematic diagram of intrinsic mode function component data after empirical mode decomposition of strong shielding components in a specific embodiment of the present invention;

[0048] Figure 5 It is a schematic diagram of a processed seismic profile in a specific embodiment of the present invention. DETAILED DESCRIPTION

[0049] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0050] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations and / or combinations thereof.

[0051] The weak signal frequency band enhancement method of seismic data of the present invention firstly decomposes the original seismic data into a series of frequency band component data of different frequency ranges by wavelet packet frequency band method; secondly, the strong shielding component is identified, that is, the energy ratio of strong reflection and weak reflection in the frequency band component data is calculated, and the dominant frequency band component data of strong reflection is automatically selected for processing, and other frequency band component data are not processed, so as to ensure the relative relationship of reflection coefficients to the greatest extent; thirdly, in the process of strong reflection suppression processing, empirical mode decomposition is performed in the determined strong shielding component data, the instantaneous amplitude, instantaneous main frequency and instantaneous phase of the strong reflection are calculated, and the main lobe wavelet approximation method controlled by three factors is adopted to fit and obtain the optimal main lobe wavelet; finally, the optimal main lobe wavelet is subtracted from the strong shielding component data, and after the strong reflection suppression of the strong shielding component data to be processed is completed, all the frequency band component data are synthesized, so as to achieve the purpose of suppressing strong reflection and enhancing weak signals.

[0052] Embodiment 1:

[0053] In a specific embodiment 1 of the present invention, Figure 1 As shown, Figure 1 The flowchart of the weak signal frequency band enhancement method of seismic data of the present invention is as follows. The weak signal frequency band enhancement method of seismic data comprises the following steps:

[0054] Step 1: After determining the seismic data processing time window, the original seismic data is decomposed into a series of sub-band component data in different frequency ranges by using the wavelet packet sub-band method. The wavelet packet sub-band formula is as follows:

[0055]

[0056] Among them: XBB i (t) is the decomposed frequency band data, i = 1, 2, ..., N, including N frequency band component data in total, d(t) is the original seismic data, t is the time series, m(A, f, φ) is the decomposed wavelet, A is the wavelet amplitude value, f is the wavelet frequency value, and φ is the wavelet phase value.

[0057] Step 2: Calculate the energy ratio of strong reflection and weak reflection in the frequency band component data, automatically select the dominant frequency band component data of strong reflection for processing, and identify the strong shielding component. The strong shielding component identification formula is:

[0058] J=Max[EnUp i (t) / EnDown i (t)],i=1,2,……,N

[0059] Where: J is the selected strong reflection dominant frequency band component data number, EnUp i (t) is the frequency band data XBB i (t) strong reflected signal energy, EnDown i (t) is the frequency band data XBB i The weak reflection signal energy of (t), i = 1, 2, ..., N, includes N sub-band component data in total.

[0060] According to the characteristics of seismic data, one or two strong shielding components can be selected for processing. The data of the strong reflection dominant band component is XBB. J (t), J is the data number of the selected strong reflection dominant frequency band component, and the data of other sub-band components are not processed to ensure the relative relationship of the reflection coefficient to the greatest extent.

[0061] Step 3: During the strong reflection suppression process, empirical mode decomposition is performed on the determined strong reflection dominant frequency band component data. The empirical mode decomposition formula is:

[0062]

[0063] Among them: XBB J (t) is the data of the strong reflection dominant frequency band component, imf j (t) is the intrinsic mode function component, j = 1, 2, ..., M, including M intrinsic mode function components, re m (t) is the residual component of the empirical mode decomposition. The first intrinsic mode function component imf of the strong reflection dominant frequency band component data is selected 1 (t), calculate the instantaneous amplitude, instantaneous main frequency, and instantaneous phase initial value of the eigenmode function component.

[0064] Step 4: According to the calculated strong reflection instantaneous amplitude, instantaneous main frequency, and instantaneous phase initial value, the main lobe wavelet approximation method controlled by three factors is used to fit the optimal main lobe wavelet. The initial main lobe wavelet expression is:

[0065] in: is the instantaneous amplitude of the component data, is the instantaneous frequency of the component data, is the instantaneous phase of the component. Under the control of the three factors of instantaneous amplitude, instantaneous main frequency and instantaneous phase, the optimal main lobe wavelet is determined according to the characteristics of seismic data, and the optimal main lobe wavelet obtained by fitting and the strong reflection dominant frequency band component are close to the minimum sum of the minimum squares.

[0066] Step 5: Subtract the approximated optimal main lobe wavelet from the strong reflection dominant frequency band component data to complete the suppression of the strong reflection component data that needs to be processed.

[0067] Step 6: Synthesize all the sub-band component data to achieve the purpose of suppressing strong reflections and enhancing weak signals.

[0068] Embodiment 2:

[0069] In a specific embodiment 2 of the present invention, a method for enhancing weak signals in seismic data frequency bands includes:

[0070] Step 1: Read the seismic profile to be processed, such as Figure 2 As shown, Figure 2 To process the schematic diagram of the previous seismic profile, there are problems in the seismic profile such as weak energy of the underlying stratum reflection signal and low signal-to-noise ratio of the seismic data, which is not conducive to the detailed description of the underlying oil and gas reservoirs.

[0071] Step 2: Decompose a series of frequency-band component data by using the wavelet packet frequency-band method, and the frequency-band component data is 10 in total. The wavelet packet frequency-band formula is as follows:

[0072]

[0073] Among them: XBB i (t) is the decomposed frequency band data, i = 1, 2, ..., N, including N frequency band component data in total, d(t) is the original seismic data, t is the time series, m(A, f, φ) is the decomposed wavelet, A is the wavelet amplitude value, f is the wavelet frequency value, and φ is the wavelet phase value.

[0074] Step 3: Identify the strong shielding component. The identification result is as follows: Figure 3 As shown, Figure 3 It is a schematic diagram of a strong shielding component seismic profile in a specific embodiment of the present invention. The main frequency of the strong shielding component seismic data is 15 Hz.

[0075] Step 4: Perform empirical mode decomposition on the determined strong reflection dominant frequency band component data. Figure 4 The first intrinsic mode function component data after empirical mode decomposition is used to calculate the instantaneous amplitude, instantaneous main frequency and instantaneous phase of the intrinsic mode function component data.

[0076] Step 5: Use the main lobe wavelet approximation method controlled by three factors to fit the optimal main lobe wavelet.

[0077] Step 6, subtract the approximated optimal main lobe wavelet from the strongly reflected dominant frequency band component data.

[0078] Step 7: synthesize all sub-band component data to obtain seismic data after weak signal enhancement. Figure 5 For the processed seismic profile, the problems of weak reflection signal energy of the underlying strata and low signal-to-noise ratio of seismic data have been better solved.

[0079] Embodiment 3:

[0080] In the specific embodiment 3 of the present invention, the weak signal frequency band enhancement method of seismic data includes the following steps: Figure 1 ):

[0081] Step 1: Select the seismic profile before processing ( Figure 2 ), after determining the seismic data processing time window, the original seismic data is decomposed into a series of sub-band component data in different frequency ranges by wavelet packet sub-band method, and the incremental step length of different frequency ranges is 5Hz or 10Hz.

[0082] Step 2: Calculate the energy ratio of strong reflection and weak reflection in the frequency band component data, automatically select the dominant frequency band component data of strong reflection for processing, and identify the strong shielding component. The strong shielding component identification formula is:

[0083] J=Max[EnUp i (t) / EnDown i (t)],i=1,2,……,N

[0084] Where: J is the selected strong reflection dominant frequency band component data number, EnUp i (t) is the frequency band data XBB i (t) strong reflected signal energy, EnDown i (t) is the frequency band data XBB i The weak reflection signal energy of (t), i = 1, 2, ..., N, includes N sub-band component data in total.

[0085] According to the characteristics of seismic data, the strong shielding component is selected for processing, and the strong reflection dominant band component data is XBB J (t), J is the selected strong reflection dominant frequency band component data number, numbered J = 3 ( Figure 3 ), the other sub-band component data are not processed and are numbered J=1, 2, 4, ..., 10, to ensure the relative relationship of the reflection coefficients to the greatest extent.

[0086] Step 3: During the strong reflection suppression process, empirical mode decomposition is performed on the determined strong reflection dominant frequency band component data. The empirical mode decomposition formula is:

[0087]

[0088] Among them: XBB J (t) is the data of the strong reflection dominant frequency band component, imf j (t) is the intrinsic mode function component, j = 1, 2, ..., M, including M intrinsic mode function components, re m (t) is the residual component of the empirical mode decomposition. The first intrinsic mode function component imf of the strong reflection dominant frequency band component data is selected 1 (t)( Figure 4 ).

[0089] Step 4: According to the calculated strong reflection instantaneous amplitude, instantaneous main frequency, and instantaneous phase initial value, the main lobe wavelet approximation method controlled by three factors is used to fit the optimal main lobe wavelet. The initial main lobe wavelet expression is:

[0090] in: is the instantaneous amplitude of the component data, is the instantaneous frequency of the component data, is the instantaneous phase of the component.

[0091] Step 5: Subtract the approximated optimal main lobe wavelet from the strong reflection dominant frequency band component data to complete the suppression of the strong reflection component data.

[0092] Step 6: Synthesize the suppressed strong shielding component data with other sub-band component data to obtain the seismic data after weak signal enhancement ( Figure 5 ), to solve the problem of weak reflection signal energy of underlying strata and low signal-to-noise ratio of seismic data, thereby improving the characterization of underlying oil and gas reservoirs and reducing reservoir description errors.

[0093] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions recorded in the aforementioned embodiments or replace some of the technical features therein with equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0094] Except for the technical features described in the specification, all other technical features are known technologies to those skilled in the art.

Claims

1. Frequency band enhancement method for weak signals in seismic data, It is characterized in that The weak signal frequency band enhancement method of seismic data includes: Step 1, decomposing a series of frequency-band component data by wavelet packet frequency-band method; Step 2, identifying the strong shielding component; Step 3, performing empirical mode decomposition on the determined strong reflection dominant frequency band component data to calculate the instantaneous amplitude, instantaneous main frequency, and instantaneous phase of the strong reflection; Step 4, using a main lobe wavelet approximation method controlled by three factors to fit the optimal main lobe wavelet; Step 5, subtracting the optimal main lobe wavelet obtained by approximation from the data of the strong reflection dominant frequency band component; Step 6: synthesize all sub-band component data to obtain seismic data after weak signal enhancement.

2. The method for enhancing weak signals of seismic data by frequency band according to claim 1, It is characterized in that The method for enhancing weak signal frequency bands of seismic data also includes, before step 1, collecting seismic data and determining an analysis time window.

3. The method for enhancing weak signals of seismic data by frequency band division according to claim 2, It is characterized in that In step 1, after determining the seismic data processing time window, the original seismic data is decomposed into a series of sub-band component data in different frequency ranges by using the wavelet packet sub-band method. The wavelet packet sub-band formula is as follows: Among them: XBB i (t) is the decomposed frequency band data, i = 1, 2, ..., N, including N frequency band component data in total, d(t) is the original seismic data, t is the time series, m(A, f, φ) is the decomposed wavelet, A is the wavelet amplitude value, f is the wavelet frequency value, and φ is the wavelet phase value.

4. The method for enhancing weak signals of seismic data by frequency band division according to claim 1, It is characterized in that In step 2, the strong shielding component is identified, that is, the energy ratio of strong reflection and weak reflection in the frequency band component data is calculated, and the dominant frequency band component data of strong reflection is selected for processing, while other frequency band component data are not processed, so as to ensure the relative relationship of the reflection coefficient to the greatest extent.

5. The method for enhancing weak signals of seismic data by frequency bands according to claim 4, It is characterized in that In step 2, the strong shielding component identification formula is: J=Max[EnUp i (t) / EnDown i (t)],i=1,2,……,N Where: J is the selected strong reflection dominant frequency band component data number, EnUp i (t) is the frequency band data XBB i (t) strong reflected signal energy, EnDown i (t) is the frequency band data XBB i The weak reflection signal energy of (t), i = 1, 2, ..., N, includes N sub-band component data in total.

6. The method for enhancing weak signals of seismic data by frequency bands according to claim 5, It is characterized in that In step 2, one or two strong shielding components are selected for processing according to the characteristics of the seismic data, and the strong reflection dominant band component data is XBB J (t), J is the data number of the selected strong reflection dominant frequency band component, and the data of other sub-band components are not processed to ensure the relative relationship of the reflection coefficient to the greatest extent.

7. The method for enhancing weak signals of seismic data by frequency bands according to claim 1, It is characterized in that In step 3, during the strong reflection suppression process, empirical mode decomposition is performed on the determined strong reflection dominant frequency band component data. The empirical mode decomposition formula is: Among them: XBB J (t) is the data of the strong reflection dominant frequency band component, imf j (t) is the intrinsic mode function component, j = 1, 2, ..., M, including M intrinsic mode function components, re m (t) is the residual component of empirical mode decomposition.

8. The method for enhancing weak signals of seismic data by frequency bands according to claim 7, It is characterized in that In step 3, the first intrinsic mode function component imf of the strong reflection dominant frequency band component data is selected 1 (t), calculate the instantaneous amplitude, instantaneous main frequency, and instantaneous phase initial value of the eigenmode function component.

9. The method for enhancing weak signals of seismic data by frequency bands according to claim 1, It is characterized in that In step 4, according to the calculated strong reflection instantaneous amplitude, instantaneous main frequency, and instantaneous phase initial value, the main lobe wavelet approximation method controlled by three factors is used to fit the optimal main lobe wavelet. The initial main lobe wavelet expression is: in: is the instantaneous amplitude of the component data, is the instantaneous frequency of the component data, is the instantaneous phase of the component.

10. The method for enhancing weak signals of seismic data by frequency bands according to claim 9, It is characterized in that In step 4, under the control of the three factors of instantaneous amplitude, instantaneous main frequency and instantaneous phase, the optimal main lobe wavelet is determined according to the characteristics of the seismic data, and the optimal main lobe wavelet obtained by fitting and the strong reflection dominant frequency band component are close to the minimum sum of the minimum squares.

11. The method for enhancing weak signals of seismic data by frequency bands according to claim 1, It is characterized in that In step 5, the optimal main lobe wavelet obtained by approximation is subtracted from the strong reflection dominant frequency band component data to complete the suppression of the strong reflection component data that needs to be processed.

Images