Seismic data denoising method and device

Abstract

The invention discloses a seismic data de-noising method and device, and relates to the technical field of data processing, and the method comprises the steps: constructing an inverse problem through employing to-be-de-noised seismic data; the inverse problem has an optimal solution, and when the optimal solution is obtained, the seismic data to be denoised after the time difference is corrected meet the following conditions: a similarity value between any two seismic traces is smaller than a first threshold value, and a normalized cross correlation coefficient between any two seismic traces is greater than a second threshold value; based on a simulated annealing algorithm, solving the inverse problem to obtain an optimal solution; the optimal solution comprises inter-channel time difference of any seismic channel and a reference channel; correcting a nonlinear signal in the seismic data to be denoised into a linear signal by using the optimal solution, and obtaining the seismic data to be denoised after correction of the time difference; and denoising the to-be-denoised seismic data after correction of the time difference through a linear denoising method to obtain denoised seismic data. According to the method, the denoising effect of the mountain-front-zone seismic data can be improved, and the requirement of high-precision imaging is met.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method and apparatus for denoising seismic data. Background Technology

[0002] This section is intended to provide background or context for the embodiments of the invention set forth in the claims. The description herein is not an admission that it is prior art simply because it is included in this section.

[0003] Currently, the main methods for denoising seismic exploration data include conventional signal processing-based methods such as fx-domain denoising, tx-domain denoising, Fourier transform, Radon transform, Curvelet transform, and Wavelet transform. These methods utilize the differences between signals and noise in terms of frequency, apparent velocity, or spatial distribution to achieve seismic data denoising. Existing technologies also include high-dimensional data regularization and denoising based on matrix and tensor decomposition: PCA, RPCA, SSA, and CP decomposition. However, in the case of low signal-to-noise ratio data in piedmont zones, the differences in the applicability of these methods are not significant. Prestack seismic data is abstracted into a five-dimensional data volume, containing nested seismic sub-representations with specific time-distance relationships floating in random noise with different distribution characteristics. The essence of high-dimensional data denoising (regularization) is to model, predict, and reconstruct the signals contained in the five-dimensional measured data, thereby achieving the goals of denoising and regularization. However, due to the undulations of the surface and the drastic lateral variations in the velocity thickness of the low-velocity zone in the piedmont region, the assumption of linear structure (linear signal) of the same phase axis within the local time window in the seismic data is increasingly inconsistent with the actual data situation under complex noise conditions, resulting in the data processing pre-effects such as denoising or data regularization failing to meet the requirements of high-precision imaging. Summary of the Invention

[0004] This invention provides a seismic data denoising method to improve the denoising effect of seismic data and meet the requirements of high-precision imaging. The method includes:

[0005] Using the seismic data to be denoised, an inverse problem is constructed. This inverse problem has an optimal solution. When the optimal solution is obtained, the seismic data to be denoised after time difference correction satisfies the following conditions: the similarity value between any two seismic traces is less than the first threshold, and the normalized cross-correlation coefficient between any two seismic traces is greater than the second threshold.

[0006] Based on the simulated annealing algorithm, the inverse problem is solved to obtain the optimal solution; the optimal solution includes the inter-trace time difference between any seismic trace and the reference trace;

[0007] Using the optimal solution, the nonlinear signals in the seismic data to be denoised are corrected into linear signals, and the denoised seismic data after corrected time difference is obtained.

[0008] The linear denoising method is used to denoise the seismic data to be denoised after time difference correction, and the denoised seismic data is obtained.

[0009] This invention also provides a seismic data denoising device to improve the denoising effect of seismic data and meet the requirements of high-precision imaging. The device includes:

[0010] The inverse problem construction module is used to construct an inverse problem using the seismic data to be denoised. The inverse problem has an optimal solution. When the optimal solution is obtained, the seismic data to be denoised after time difference correction meets the following conditions: the similarity value between any two seismic traces is less than the first threshold, and the normalized cross-correlation coefficient between any two seismic traces is greater than the second threshold.

[0011] The inverse problem solving module is used to solve the inverse problem based on the simulated annealing algorithm to obtain the optimal solution; the optimal solution includes the inter-trace time difference between any seismic trace and the reference trace;

[0012] The denoising module is used to correct nonlinear signals in the seismic data to be denoised into linear signals using the optimal solution, so as to obtain the seismic data to be denoised after time difference correction; and to denoise the seismic data to be denoised after time difference correction using a linear denoising method, so as to obtain the denoised seismic data.

[0013] This invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-described seismic data denoising method.

[0014] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described seismic data denoising method.

[0015] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-described seismic data denoising method.

[0016] In this embodiment of the invention, an inverse problem is constructed using the seismic data to be denoised. This inverse problem has an optimal solution, which, when the optimal solution is obtained, satisfies the following conditions after time difference correction: the similarity value between any two seismic traces is less than a first threshold, and the normalized cross-correlation coefficient between any two seismic traces is greater than a second threshold. Based on the simulated annealing algorithm, the inverse problem is solved to obtain the optimal solution. The optimal solution includes the inter-trace time difference between any seismic trace and the reference trace. Using the optimal solution, the nonlinear signals in the seismic data to be denoised are corrected into linear signals, resulting in the seismic data to be denoised after time difference correction. A linear denoising method is then used to denoise the seismic data after time difference correction, resulting in denoised seismic data. In this embodiment of the invention, an inverse problem is constructed, an optimal solution is obtained, and the nonlinear signals in the seismic data to be denoised are corrected into linear signals using the optimal solution before denoising, ensuring that the signals satisfy the linear assumption. Therefore, the denoising effect is more significant after applying the linear denoising method. This method can effectively improve the denoising effect of seismic data, meet the requirements of high-precision imaging, and is particularly suitable for denoising seismic data in piedmont areas. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0018] Figure 1 This is a flowchart illustrating the seismic data denoising method in an embodiment of the present invention;

[0019] Figure 2 This is a schematic diagram of the synthesized seismic record in an embodiment of the present invention;

[0020] Figures 3 to 5 To utilize Figure 2 A schematic diagram of wavelet feature data generated from synthetic seismic records;

[0021] Figure 6 This is a flowchart of the algorithm for global optimization of simulated annealing in an embodiment of the present invention;

[0022] Figure 7 This is a specific embodiment of the seismic data denoising method in this invention;

[0023] Figures 8-11 This is a schematic diagram illustrating the noise reduction process implemented in an embodiment of the present invention;

[0024] Figure 12 Local data with Gaussian random noise added to the synthetic seismic record in Example 1;

[0025] Figure 13 The data used in Example 1 to eliminate nonlinear time differences using existing technology;

[0026] Figure 14 This refers to the data in Example 1 that uses the seismic data denoising method of this invention to eliminate nonlinear time differences;

[0027] Figure 15 for Figure 13 , Figure 14 Comparison of denoising results Figure 1 ;

[0028] Figure 16 for Figure 13 , Figure 14 Comparison of denoising results Figure 2 ;

[0029] Figure 17 The co-offset data of the synthetic seismic record in Example 1 were supplemented with Gaussian random noise;

[0030] Figure 18 The result of denoising the co-offset data of the synthetic seismic record in Example 1;

[0031] Figure 19 This is the common offset data of the synthetic seismic record before denoising in Example 2;

[0032] Figure 20 This is the denoised co-offset data of the synthetic seismic record from Example 2;

[0033] Figure 21 This is the common center point gather of the synthetic seismic record before denoising in Example 2;

[0034] Figure 22 The velocity spectrum of the synthetic seismic record before denoising in Example 2;

[0035] Figure 23 This is the denoised common centroid gather of the synthetic seismic record from Example 2;

[0036] Figure 24 The velocity spectrum of the synthesized seismic record after denoising in Example 2;

[0037] Figure 25 This is the migration imaging profile of the synthetic seismic record before denoising in Example 2;

[0038] Figure 26 This is a denoised migration imaging profile of the synthetic seismic record from Example 2;

[0039] Figure 27 This is a schematic diagram of a seismic data denoising device in an embodiment of the present invention. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0041] The acquisition, storage, use, and processing of data in this application all comply with the relevant provisions of national laws and regulations.

[0042] The applicant found that seismic exploration data processing methods mainly include conventional denoising methods based on signal processing: fx-domain denoising, tx-domain denoising, Fourier transform, Radon transform, Curvelet transform, Wavelet transform, and high-dimensional data regularization and denoising based on matrix and tensor decomposition: PCA, RPCA, SSA, CP decomposition, etc. However, when processing low signal-to-noise ratio data in mountainous areas, the assumptions of these methods do not match the actual data, resulting in unsatisfactory denoising effects.

[0043] To this end, the applicant proposed a seismic data denoising method, specifically a seismic image phase axis nonlinear time difference detection denoising method based on global optimization. This method introduces nonlinear time difference detection and elimination technology, which allows the linear signal assumption of conventional denoising methods to be satisfied in piedmont seismic data, resulting in a significant improvement in the denoising effect of piedmont seismic data and a certain improvement in the amplitude preservation of seismic signals.

[0044] Figure 1 This is a flowchart illustrating the seismic data denoising method in an embodiment of the present invention, as shown below. Figure 1 As shown, the method includes:

[0045] Step 101: Construct an inverse problem using the seismic data to be denoised; the inverse problem has an optimal solution. When the optimal solution is obtained, the seismic data to be denoised after time difference correction meets the following conditions: the similarity value between any two seismic traces is less than the first threshold, and the normalized cross-correlation coefficient between any two seismic traces is greater than the second threshold.

[0046] Step 102: Based on the simulated annealing algorithm, solve the inverse problem to obtain the optimal solution; the optimal solution includes the inter-trace time difference between any seismic trace and the reference trace;

[0047] Step 103: Using the optimal solution, the nonlinear signals in the seismic data to be denoised are corrected into linear signals to obtain the seismic data to be denoised after the corrected time difference.

[0048] Step 104: Denoise the seismic data to be denoised after time difference correction using a linear denoising method to obtain denoised seismic data.

[0049] from Figure 1 As shown in the flowchart, in this embodiment of the invention, an inverse problem is constructed using the seismic data to be denoised. This inverse problem has an optimal solution. When the optimal solution is obtained, the seismic data to be denoised after time difference correction satisfies the following conditions: the similarity value between any two seismic traces is less than a first threshold, and the normalized cross-correlation coefficient between any two seismic traces is greater than a second threshold. Based on the simulated annealing algorithm, the inverse problem is solved to obtain the optimal solution. The optimal solution includes the inter-trace time difference between any seismic trace and the reference trace. Using the optimal solution, the nonlinear signals in the seismic data to be denoised are corrected into linear signals to obtain the seismic data to be denoised after time difference correction. The seismic data to be denoised after time difference correction is then denoised using a linear denoising method to obtain the denoised seismic data. In this embodiment of the invention, an inverse problem is constructed, an optimal solution is obtained, and the nonlinear signals in the seismic data to be denoised are corrected into linear signals using the optimal solution before denoising, so that the signal satisfies the linear assumption. Therefore, the denoising effect is more obvious after applying the linear denoising method. This method can effectively improve the denoising effect of seismic data, meet the requirements of high-precision imaging, and is especially suitable for denoising seismic data in piedmont areas.

[0050] The earthquake data denoising method in the embodiments of the present invention will be explained in detail below.

[0051] This invention provides an inversion method based on simulated annealing global optimization. First, the seismic data is expressed as wavelet features. Then, the nonlinear time difference is posited as an inverse problem. This inverse problem is solved by simulated annealing to obtain the optimal nonlinear time difference sequence. The time difference of the nonlinear signal is then converted into a linear signal. Finally, a linear denoising method is applied to denoise the seismic data in the piedmont zone.

[0052] In one embodiment, constructing an inverse problem using the seismic data to be denoised may include:

[0053] Input the seismic data to be denoised; the seismic data to be denoised includes multiple wavelets;

[0054] Extracting geometric features of multiple wavelets from seismic data to be denoised;

[0055] An inverse problem is constructed using the seismic data to be denoised and the geometric characteristics of multiple wavelets.

[0056] In implementing this invention, the first step is to represent seismic data based on wavelet features. For example, a large amount of seismic data is processed by using a sliding window to extract local data in the time and spatial domains. This local data is then input, and wavelet features are used to represent it. The main method is to extract wavelet feature attributes, and through the extraction of these attributes, the seismic image is represented in a characteristic way. (Reference) Figure 2 , Figure 2This demonstrates a synthetic seismic record used to generate wavelet feature data, such as wavelet geometry.

[0057] Seismic images are composed of continuously nested band-limited wavelets. In actual seismic data, the continuity of the phase axis can be interrupted for various reasons, the essential reason for which is the absence of the continuous nesting relationship of the band-limited wavelets. Unlike ordinary images, seismic images have their own unique characteristics, being composed of composite or independent wavelets with specific time intervals and randomly distributed noise. Seismic wavelets are the most basic elements of seismic data. The most essential characteristic of a seismic wavelet is its "shape" of a main lobe plus two side lobes, and this "shape," "pattern," or "template" is distributed over the length of the seismic wavelet. The detailed parameters of the wavelet's geometric characteristics can be expressed as: main lobe width, side lobe width; trough position, crest position; crest amplitude, trough amplitude.

[0058] The width attribute of a seismic wavelet implicitly contains the length of its primary element (the "shape" of a main lobe plus two side lobes). The location and amplitude information of special points such as crests and troughs are also included within the width attribute. Appropriately "packaging" the wavelet width can reflect the important characteristics of the "shape" of a main lobe plus two side lobes distributed within the length of the seismic wavelet. Wavelet width extraction mainly involves two approaches: one is direct calculation based on the positive and negative relationships of the seismic data itself; the other is indirect approximation based on the relative positions of crests and troughs. The latter, due to the inclusion of crest and trough positions, extends the wavelet width attribute to the distribution of the wavelet morphology.

[0059] Figures 3 to 5 To utilize Figure 2 A schematic diagram of wavelet feature data generated from synthetic seismic records. Figures 3 to 5 The peak feature attributes, edge feature attributes, and wavelet width feature attributes are displayed sequentially.

[0060] Next, an inverse problem is constructed. In implementation, a metric is defined to measure the similarity or matching degree between different seismic traces. Commonly used metrics include Euclidean distance, correlation coefficient, and cosine similarity; the specific metric chosen depends on the specific requirements of the application. This embodiment of the invention uses both Euclidean distance and cosine similarity for calculation.

[0061] For example, the inverse problem of finding the optimal nonlinear inter-channel time difference can be defined as follows:

[0062]

[0063]

[0064] in This is the normalized cross-correlation coefficient;

[0065] In the formula, τ is a vector, τ=[τ1,τ2,......,τ N ], τ i x represents the time difference between the i-th earthquake and the reference trace. i ,x j denoted as the geometric characteristics of the i-th and j-th seismic trace wavelets, respectively; N is the number of seismic traces; k is the longitudinal time sampling point; t1 and t2 are the start and end sampling point times of the local window data, respectively; and ε1 and ε2 are the first and second thresholds, respectively.

[0066] The optimal solution to this inverse problem is obtained by correcting the time difference in the seismic data to be denoised, i.e., by correcting the time difference τ=[τ1,τ2,......,τ N The earthquake feature data after [ ] satisfy the following conditions: the similarity value between any two seismic traces is less than the first threshold, and the normalized cross-correlation coefficient between any two seismic traces is greater than the second threshold. The first threshold and the second threshold can be artificially set hyperparameters, which are used to determine the truncation size of the difference and correlation between seismic traces, respectively.

[0067] Subsequently, based on the simulated annealing algorithm, the inverse problem is solved to obtain the optimal solution, which includes the inter-trace time difference between any seismic trace and the reference trace, i.e., the optimal nonlinear time difference sequence.

[0068] Simulated Annealing (SA) is based on the annealing principle of solids. When a solid is at a high temperature, its internal energy is high, and the particles inside are in rapid, disordered motion. As the temperature gradually decreases, the internal energy of the solid decreases, and the particles gradually become more ordered. Finally, when the solid reaches room temperature, its internal energy reaches its minimum, at which point the particles are most stable. The simulated annealing algorithm is designed based on this principle.

[0069] Simulated annealing starts at a relatively high temperature, called the initial temperature. As the temperature parameter decreases, the solution in the algorithm tends to stabilize. However, such a stable solution may be a local optimum. At this point, the simulated annealing algorithm will jump out of such a local optimum with a certain probability in order to find the global optimum of the objective function.

[0070] In one embodiment, solving the inverse problem based on the simulated annealing algorithm to obtain the optimal solution may include:

[0071] Given an initial temperature and an initial solution τ, calculate the functions f(τ), g(τ), and h(τ), where...

[0072]

[0073]

[0074]

[0075] Execute the following loop until the preset maximum number of iterations is reached and the termination condition is met, to obtain the optimal solution:

[0076] Random perturbation generates a new τ', and the functions f(τ′), g(τ′), and h(τ′) are calculated.

[0077] If f(τ′) < f(τ), g(τ′) < g(τ), and h(τ′) > h(τ), then let τ = τ′, f(τ) = f(τ′), g(τ) = g(τ′), and h(τ) = h(τ′); otherwise, according to the Metropolis acceptance criterion, accept the solution τ′ with probability P, where,

[0078]

[0079] In the formula, T is the current temperature;

[0080] When the preset maximum number of iterations is reached, it is determined whether the termination condition is met. The termination condition includes: g(τ) < ε1 and h(τ) > ε2.

[0081] If the termination condition is met, the current τ' is taken as the optimal solution;

[0082] If the termination condition is not met, lower the temperature and reset the iteration count to zero.

[0083] For details, please refer to [link / reference]. Figure 6 , Figure 6 The flowchart of the global optimization algorithm for simulated annealing in this embodiment of the invention is as follows:

[0084] (1) Randomly generate an initial solution τ, and calculate the functions f(τ), g(τ), and h(τ);

[0085] (2) Based on the existing solution, randomly perturb to generate a new solution τ', and calculate f(τ′), g(τ′), h(τ′);

[0086] (3) If f(τ′)<f(τ), g(τ′)<g(τ), h(τ′)>h(τ),

[0087] Then let τ=τ′, f(τ)=f(τ′), g(τ)=g(τ′), h(τ)=h(τ′),

[0088] Otherwise, according to the Metropolis criterion, the solution is accepted with probability P:

[0089] (4) Has the maximum number of iterations been reached?

[0090] No: Go to step 2;

[0091] Yes: Jump to step 5;

[0092] (5) Whether the termination condition is met, g(τ)<ε1 and h(τ>ε2;

[0093] Yes: The calculation is complete, and the optimal solution is returned;

[0094] No: Reset the iteration count, T = γT, where γ is the rate of temperature decrease, and then jump to step 2.

[0095] Then, using the optimal solution, the nonlinear signals in the seismic data to be denoised are corrected into linear signals to obtain the seismic data to be denoised after time difference correction. The seismic data to be denoised after time difference correction is then denoised using a linear denoising method to obtain the denoised seismic data.

[0096] refer to Figure 7 , Figure 7 This is a specific embodiment of the seismic data denoising method in this invention. After solving for the nonlinear time difference, the nonlinear signal is corrected into a linear signal through time difference correction, and then denoised by a linear signal denoising method to obtain seismic data with a high signal-to-noise ratio. The linear denoising method is, for example, a linear signal denoising method based on Radon transform.

[0097] Figures 8-11 This is a schematic diagram illustrating the noise reduction process in an embodiment of the present invention. Figures 8-11 The following are represented in sequence: nonlinear synthetic seismic data before denoising, data with nonlinear time difference detected and eliminated, seismic data after linear denoising, and original seismic data recovered through nonlinear time difference. The horizontal and vertical axes reflect the number of sampling points.

[0098] The method for denoising seismic data in an embodiment of the present invention is described below with reference to the accompanying drawings.

[0099] Example 1:

[0100] Figure 12 Gaussian random noise was added to the local data of the synthetic seismic record in Example 1.

[0101] Figure 13 The data used in Example 1 employs existing technology to eliminate nonlinear time differences.

[0102] Figure 14 The data used in Example 1 is the data for which nonlinear time differences have been eliminated by the seismic data denoising method described in this invention.

[0103] Figure 15 for Figure 13 , Figure 14 Comparison of denoising results Figure 1 , Figure 15The upper figure shows a comparison between the nonlinear time difference (dashed line) obtained by the prior art and the actual nonlinear time difference (solid line), while the lower figure shows a comparison between the nonlinear time difference (dashed line) obtained by the method of the present invention and the actual nonlinear time difference (solid line).

[0104] Figure 16 for Figure 13 , Figure 14 Comparison of denoising results Figure 2 , Figure 16 In the figure above, the error between the nonlinear time difference obtained by the prior art and the actual nonlinear time difference is shown. In the figure below, the error between the nonlinear time difference obtained by the method of the present invention and the actual nonlinear time difference is shown.

[0105] Figure 17 The co-offset data of Gaussian random noise was added to the synthetic seismic record of Example 1.

[0106] Figure 18 The result is the denoising of the co-offset data of the synthetic seismic record in Example 1.

[0107] Example 2:

[0108] Figure 19 This is the common offset data of the synthetic seismic record before denoising in Example 2;

[0109] Figure 20 This is the denoised co-offset data of the synthetic seismic record from Example 2;

[0110] Figure 21 This is the common center point gather of the synthetic seismic record before denoising in Example 2;

[0111] Figure 22 The velocity spectrum of the synthetic seismic record before denoising in Example 2;

[0112] Figure 23 This is the denoised common centroid gather of the synthetic seismic record from Example 2;

[0113] Figure 24 The velocity spectrum of the synthesized seismic record after denoising in Example 2;

[0114] Figure 25 This is the migration imaging profile of the synthetic seismic record before denoising in Example 2;

[0115] Figure 26 This is a denoised migration imaging profile of the synthetic seismic record from Example 2.

[0116] In summary, this invention proposes a global optimization-based nonlinear time difference detection denoising method for seismic image phase axis in piedmont seismic data. By introducing nonlinear time difference detection and elimination techniques, this method ensures that the linear signal assumptions of conventional denoising methods are satisfied in piedmont seismic data, resulting in a significant improvement in the denoising effect and a certain degree of improvement in the amplitude preservation of seismic signals. The method of this invention improves the quality of mountain seismic data.

[0117] It should also be noted that the method of this invention is highly adaptable, applicable to both 2D and 3D seismic data, and has relatively low requirements for the processed data. Furthermore, it is simple to operate and easy to implement. Through algorithm integration and external integration of input parameters, only the relevant seismic data and inversion denoising parameters (initial temperature, first threshold, second threshold, etc.) need to be input into a single interface, without the need to open and modify the source code.

[0118] This invention also provides a seismic data denoising device, as described in the following embodiments. Since the principle behind this device is similar to that of the seismic data denoising method, its implementation can be referenced from the implementation of the seismic data denoising method; repeated details will not be elaborated further.

[0119] Figure 27 This is a schematic diagram of a seismic data denoising device in an embodiment of the present invention, as shown below. Figure 27 As shown, the device includes:

[0120] The inverse problem construction module 2701 is used to construct an inverse problem using the seismic data to be denoised. The inverse problem has an optimal solution. When the optimal solution is obtained, the seismic data to be denoised after time difference correction meets the following conditions: the similarity value between any two seismic traces is less than the first threshold, and the normalized cross-correlation coefficient between any two seismic traces is greater than the second threshold.

[0121] The inverse problem solving module 2702 is used to solve the inverse problem based on the simulated annealing algorithm to obtain the optimal solution; the optimal solution includes the inter-trace time difference between any seismic trace and the reference trace;

[0122] The denoising module 2703 is used to correct the nonlinear signals in the seismic data to be denoised into linear signals using the optimal solution, so as to obtain the seismic data to be denoised after time difference correction; and to denoise the seismic data to be denoised after time difference correction using a linear denoising method, so as to obtain the denoised seismic data.

[0123] In one embodiment, the inverse problem building module is specifically used for:

[0124] Input the seismic data to be denoised; the seismic data to be denoised includes multiple wavelets;

[0125] Extracting geometric features of multiple wavelets from seismic data to be denoised;

[0126] An inverse problem is constructed using the seismic data to be denoised and the geometric characteristics of multiple wavelets.

[0127] In one embodiment, the geometric features of the wavelet include any combination of the following:

[0128] Main lobe width, side lobe width, trough position, crest position, crest amplitude, trough amplitude.

[0129] In one embodiment, the inverse problem is represented as follows:

[0130]

[0131]

[0132] in This is the normalized cross-correlation coefficient;

[0133] In the formula, τ is a vector, τ=[τ1,τ2,......,τ N ], τ i x represents the time difference between the i-th earthquake and the reference trace. i ,x j denoted as the geometric characteristics of the i-th and j-th seismic trace wavelets, respectively; N is the number of seismic traces; k is the longitudinal time sampling point; t1 and t2 are the start and end sampling point times of the local window data, respectively; and ε1 and ε2 are the first and second thresholds, respectively.

[0134] In one embodiment, the inverse problem solving module is specifically used for:

[0135] Given an initial temperature and an initial solution τ, calculate the functions f(τ), g(τ), and h(τ), where...

[0136]

[0137]

[0138]

[0139] Execute the following loop until the preset maximum number of iterations is reached and the termination condition is met, to obtain the optimal solution:

[0140] Random perturbation generates a new τ', and the functions f(τ′), g(τ′), and h(τ′) are calculated.

[0141] If f(τ′) < f(τ), g(τ′) < g(τ), and h(τ′) > h(τ), then let τ = τ′, f(τ) = f(τ′), g(τ) = g(τ′), and h(τ) = h(τ′); otherwise, according to the Metropolis acceptance criterion, accept the solution τ′ with probability P, where,

[0142]

[0143] In the formula, T is the current temperature;

[0144] When the preset maximum number of iterations is reached, it is determined whether the termination condition is met. The termination condition includes: g(τ) < ε1 and h(τ) > ε2.

[0145] If the termination condition is met, the current τ' is taken as the optimal solution;

[0146] If the termination condition is not met, lower the temperature and reset the iteration count to zero.

[0147] This invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-described seismic data denoising method.

[0148] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described seismic data denoising method.

[0149] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-described seismic data denoising method.

[0150] In this embodiment of the invention, an inverse problem is constructed using the seismic data to be denoised. This inverse problem has an optimal solution, and when the optimal solution is obtained, the seismic data to be denoised after time difference correction satisfies the following conditions: the similarity value between any two seismic traces is less than a first threshold, and the normalized cross-correlation coefficient between any two seismic traces is greater than a second threshold. Based on the simulated annealing algorithm, the inverse problem is solved to obtain the optimal solution. The optimal solution includes the inter-trace time difference between any seismic trace and the reference trace. Using the optimal solution, the nonlinear signals in the seismic data to be denoised are corrected into linear signals, resulting in the seismic data to be denoised after time difference correction. A linear denoising method is then used to denoise the seismic data to be denoised after time difference correction, resulting in denoised seismic data. In this embodiment of the invention, an inverse problem is constructed, an optimal solution is obtained, and the nonlinear signals in the seismic data to be denoised are corrected into linear signals using the optimal solution before denoising, ensuring that the signal satisfies the linear assumption. Therefore, the denoising effect is more significant after applying the linear denoising method. This method can effectively improve the denoising effect of seismic data, meet the requirements of high-precision imaging, and is especially suitable for denoising seismic data in piedmont areas.

[0151] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0152] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0153] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1The function specified in one or more boxes.

[0154] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0155] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for denoising seismic data, characterized in that, include: Construct an inverse problem using the seismic data to be denoised; The inverse problem has an optimal solution. When the optimal solution is obtained, the seismic data to be denoised after correcting for time difference meets the following conditions: the similarity value between any two seismic traces is less than the first threshold, and the normalized cross-correlation coefficient between any two seismic traces is greater than the second threshold. Based on the simulated annealing algorithm, the inverse problem is solved to obtain the optimal solution; the optimal solution includes the inter-trace time difference between any seismic trace and the reference trace; Using the optimal solution, the nonlinear signals in the seismic data to be denoised are corrected into linear signals, and the denoised seismic data after corrected time difference is obtained. The linear denoising method is used to denoise the seismic data to be denoised after time difference correction, and the denoised seismic data is obtained.

2. The method as described in claim 1, characterized in that, Using the seismic data to be denoised, construct the inverse problem, including: Input the seismic data to be denoised; the seismic data to be denoised includes multiple wavelets; Extracting geometric features of multiple wavelets from seismic data to be denoised; An inverse problem is constructed using the seismic data to be denoised and the geometric characteristics of multiple wavelets.

3. The method as described in claim 2, characterized in that, The geometric characteristics of the wavelet include any combination of the following: Main lobe width, side lobe width, trough position, crest position, crest amplitude, trough amplitude.

4. The method as described in claim 2, characterized in that, The inverse problem is represented as follows: in This is the normalized cross-correlation coefficient; In the formula, τ is a vector, τ=[τ1,τ2,......,τ N ], τ i x represents the time difference between the i-th earthquake and the reference trace. i ,x j denoted as the geometric characteristics of the i-th and j-th seismic trace wavelets, N is the number of seismic traces, k is the longitudinal time sampling point, t1 and t2 are the start and end sampling point times of the local window data, and ε1 and ε2 are the first and second thresholds, respectively.

5. The method as described in claim 4, characterized in that, Based on the simulated annealing algorithm, the inverse problem is solved to obtain the optimal solution, including: Given an initial temperature and an initial solution τ, calculate the functions f(τ), g(τ), and h(τ), where... Execute the following loop until the preset maximum number of iterations is reached and the termination condition is met, to obtain the optimal solution: Random perturbation generates a new τ', and the functions f(τ′), g(τ′), and h(τ′) are calculated. If f(τ′) < f(τ), g(τ′) < g(τ), and h(τ′) > h(τ), then let τ = τ′, f(τ) = f(τ′), g(τ) = g(τ′), and h(τ) = h(τ′); otherwise, according to the Metropolis acceptance criterion, accept the solution τ′ with probability P, where, In the formula, T is the current temperature; When the preset maximum number of iterations is reached, it is determined whether the termination condition is met. The termination condition includes: g(τ) < ε1 and h(τ) > ε2. If the termination condition is met, the current τ' is taken as the optimal solution; If the termination condition is not met, lower the temperature and reset the iteration count to zero.

6. A seismic data denoising device, characterized in that, include: The inverse problem building module is used to construct inverse problems using the seismic data to be denoised; The inverse problem has an optimal solution. When the optimal solution is obtained, the seismic data to be denoised after correcting for time difference meets the following conditions: the similarity value between any two seismic traces is less than the first threshold, and the normalized cross-correlation coefficient between any two seismic traces is greater than the second threshold. The inverse problem solving module is used to solve the inverse problem based on the simulated annealing algorithm to obtain the optimal solution; the optimal solution includes the inter-trace time difference between any seismic trace and the reference trace; The denoising module is used to correct nonlinear signals in the seismic data to be denoised into linear signals using the optimal solution, so as to obtain the seismic data to be denoised after time difference correction; and to denoise the seismic data to be denoised after time difference correction using a linear denoising method, so as to obtain the denoised seismic data.

7. The apparatus as claimed in claim 6, characterized in that, Using the seismic data to be denoised, construct the inverse problem, including: Input the seismic data to be denoised; the seismic data to be denoised includes multiple wavelets; Extracting geometric features of multiple wavelets from seismic data to be denoised; An inverse problem is constructed using the seismic data to be denoised and the geometric characteristics of multiple wavelets.

8. The apparatus as claimed in claim 7, characterized in that, The geometric characteristics of the wavelet include any combination of the following: Main lobe width, side lobe width, trough position, crest position, crest amplitude, trough amplitude.

9. The apparatus as claimed in claim 7, characterized in that, The inverse problem is represented as follows: in This is the normalized cross-correlation coefficient; In the formula, τ is a vector, τ=[τ1,τ2,......,τ N ], τ i x represents the time difference between the i-th earthquake and the reference trace. i ,x j denoted as the geometric characteristics of the i-th and j-th seismic trace wavelets, N is the number of seismic traces, k is the longitudinal time sampling point, t1 and t2 are the start and end sampling point times of the local window data, and ε1 and ε2 are the first and second thresholds, respectively.

10. The apparatus as claimed in claim 9, characterized in that, The inverse problem solving module is specifically used for: Given an initial temperature and an initial solution τ, calculate the functions f(τ), g(τ), and h(τ), where... Execute the following loop until the preset maximum number of iterations is reached and the termination condition is met, to obtain the optimal solution: Random perturbation generates a new τ', and the functions f(τ′), g(τ′), and h(τ′) are calculated. If f(τ′) < f(τ), g(τ′) < g(τ), and h(τ′) > h(τ), then let τ = τ′, f(τ) = f(τ′), g(τ) = g(τ′), and h(τ) = h(τ′); otherwise, according to the Metropolis acceptance criterion, accept the solution τ′ with probability P, where, In the formula, T is the current temperature; When the preset maximum number of iterations is reached, it is determined whether the termination condition is met. The termination condition includes: g(τ) < ε1 and h(τ) > ε2. If the termination condition is met, the current τ' is taken as the optimal solution; If the termination condition is not met, lower the temperature and reset the iteration count to zero.

11. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 5.

12. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method of any one of claims 1 to 5.

13. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the method of any one of claims 1 to 5.

Images