High-precision seismic data time-frequency spectrum generation method based on sidelobe suppression method

Abstract

The invention belongs to the technical field of seismic data processing, and particularly relates to a high-precision seismic data time-frequency spectrum generation method based on a sidelobe suppression method, and the method comprises the following steps: 1, calculating an analysis signal of a seismic channel for seismic data; 2, solving the instantaneous frequency of the seismic data; 3, solving a stabilized seismic dominant frequency of the seismic data; 4, constructing a discrete matrix related to the Kaiser window; 5, multiplying the discrete matrix by the seismic trace to obtain a time-frequency spectrum; and step 6, displaying and analyzing the seismic data time-frequency spectrum. Aiming at the problem that a Gaussian window applied by a conventional W transformation method cannot flexibly adjust a balance relation between a main lobe width and sidelobe attenuation, the suppression of sidelobe intensity in a W transformation process is enhanced by introducing a Kaize time window, and the frequency spectrum leakage crosstalk between different signal components in a time spectrum is suppressed, so that the suppression of the sidelobe intensity in the W transformation process is enhanced. And the resolution of the time-frequency spectrum frequency domain is improved.

Description

Technical Field

[0001] This invention belongs to the field of seismic data processing technology, and particularly relates to a high-precision seismic data time-spectrum generation method based on sidelobe suppression. Background Technology

[0002] In the field of seismic data processing, time-frequency analysis is a key technology for revealing underground geological structures and reservoir characteristics. Its core lies in generating high-precision time-frequency spectra to clearly reflect the distribution characteristics of seismic signals in the time and frequency domains.

[0003] In the prior art, a precise S-transform method for generating high-precision time spectrum (patent number: CN202010445275.1) has been applied. This method generates time spectrum through steps such as calculating the analytical signal channel by channel, stabilizing the time-varying instantaneous frequency, estimating the time-varying and space-varying seismic main frequency, and fine S-transform.

[0004] Meanwhile, the W transform, as a high-resolution time-frequency analysis method, has certain advantages in high-resolution acquisition of the time spectrum in the low-frequency region of seismic data due to its use of window functions (such as Gaussian windows) that are symmetric about the dominant frequency of seismic data. This helps to improve the ability to extract low-frequency information and identify anomalies.

[0005] However, existing technologies still have significant drawbacks:

[0006] First, the W transform is a linear time-frequency analysis method. Due to the Gabor uncertainty principle, the time domain resolution and frequency domain resolution are inversely proportional, and it is impossible to achieve the optimal state at the same time. This makes it difficult to meet the high-resolution time-frequency spectrum requirements for deep-to-ultra-deep reservoir description and fluid prediction.

[0007] Secondly, although the Gaussian window used in the W transform is theoretically free of spectral leakage and has optimal time-frequency concentration, it needs to be truncated in actual seismic data processing. The truncated Gaussian window has strong side lobes, which leads to severe crosstalk between different frequency components of seismic data and directly causes a decrease in frequency domain resolution.

[0008] Third, the Gaussian window has inherent limitations. It cannot simultaneously adjust the width of the main lobe and the compression degree of the side lobes. Furthermore, the side lobes decay slowly and have a long "tail," which means that frequency regions far from the main lobe are still affected by spectral energy leakage, further restricting the improvement of time-spectral resolution.

[0009] To address the shortcomings of the existing technologies, there is an urgent need for a high-precision seismic data time-frequency generation method that can effectively suppress sidelobes, balance time-frequency resolution, and improve processing efficiency. Summary of the Invention

[0010] The purpose of this invention is to address the aforementioned technical problems by providing a high-precision seismic data time-spectrum generation method based on sidelobe suppression.

[0011] In view of this, the present invention provides a high-precision seismic data time-spectrum generation method based on sidelobe suppression, comprising the following steps:

[0012] Step 1: Calculate the analytical signal of the seismic trace from the seismic data;

[0013] Step 2: Determine the instantaneous frequency of the seismic data;

[0014] Step 3: Determine the stable seismic dominant frequency of the seismic data;

[0015] Step 4: Construct the discrete matrix associated with the Kaiser window;

[0016] Step 5: Multiply the discrete matrix with the seismic trace to obtain the time spectrum;

[0017] Step Six: Display and analyze the time spectrum of the earthquake data.

[0018] Preferably, the Kaiser window is a window function whose main lobe width and side lobe energy intensity can be adjusted by a non-negative parameter. When the non-negative parameter is 0, it degenerates into a rectangular window. Furthermore, by adjusting this parameter, the main lobe width and side lobe energy intensity can be inversely adjusted to suppress the side lobe energy.

[0019] Preferably, when constructing the discrete matrix related to the Kaiser window, the Kaiser window W transform is used, which is to extract the spectrum of the seismic signal after applying the Kaiser window.

[0020] Preferably, the time window of the Kaiser window W transform is a function related to the instantaneous frequency. This function enables the time window to be scaled in the time-frequency plane to balance the resolution of the time spectrum in the time domain and the frequency domain.

[0021] Preferably, the function related to the instantaneous frequency is constructed based on the stable seismic dominant frequency obtained in step three, the difference between the instantaneous frequency obtained in step two and the frequency to be analyzed, and the parameters used to control the length of the time window. Adjusting the time window through this function can improve the resolution of the Kaiser window in the low-frequency region.

[0022] Preferably, in step three, the stable seismic dominant frequency is obtained by performing a conventional W-transform on the seismic data.

[0023] Preferably, steps four and five further include: converting the Kaiser window W transform into a product of a matrix and a vector, where the vector is the original seismic data, the matrix is ​​the coefficient matrix corresponding to the Kaiser window W transform, and the product result is the output time spectrum.

[0024] Preferably, during the Kaiser window W transform, a block processing method is used to perform parallel computation on the multichannel seismic records: the multichannel seismic records are divided into multiple local multichannel seismic records, a local matrix is ​​constructed for each local multichannel seismic record and the Kaiser window W transform is performed to obtain the local time spectrum, and all local time spectra are summed to obtain the multichannel time spectrum.

[0025] Preferably, the product of the matrix and the vector is accelerated by calling the cublasgemv equation of CUDA, thereby improving computational efficiency.

[0026] Preferably, the Kaiser window W transform is performed based on a GPU device.

[0027] The beneficial effects of this invention are:

[0028] This invention proposes a Kaiser window W transform for generating high-resolution seismic data time spectrum. It addresses the problem that the Gaussian window used in conventional W transform methods cannot flexibly adjust the balance between the main lobe width and side lobe attenuation. By introducing the Kaiser time window, the suppression of side lobe intensity during the W transform process is enhanced, and the resolution of the time spectrum frequency domain is improved by suppressing spectral leakage crosstalk between different signal components in the time spectrum.

[0029] The invention proposes a Kaiser window-based W transform to address the problem of wide sidelobes and poor frequency resolution in the low-frequency region. It proposes a Kaiser window that is related to the instantaneous frequency and improves the resolution of the Kaiser window in the low-frequency region by constructing a time window that is symmetrical with the instantaneous main frequency.

[0030] This invention utilizes multi-GPU parallel computing to effectively improve the efficiency of time-frequency analysis methods for seismic data based on the Kaiser window W transform. By leveraging the multi-core characteristics of GPUs, the time required for time-frequency analysis is shortened, thereby enhancing the applicability of time-frequency analysis methods to large-scale 3D seismic data. Attached Figure Description

[0031] Figure 1 This is a schematic diagram of the overall process of the present invention;

[0032] Figure 2 This is a schematic diagram of the time-domain and frequency-domain response of the Kaiser window of the present invention;

[0033] Figure 3 This is a schematic diagram comparing the time-domain and frequency-domain responses of the Gaussian window and the Kaiser window of this invention;

[0034] Figure 4 This is a schematic diagram of the multichannel seismic record analysis process of the present invention. Detailed Implementation

[0035] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0036] It should be noted that all directional and positional terms used in this invention, such as "up," "down," "left," "right," "front," "back," "vertical," "horizontal," "inner," "outer," "top," "lower," "lateral," "longitudinal," and "center," are only used to explain the relative positional relationships and connections between components in a specific state (as shown in the accompanying drawings). They are merely for the convenience of describing the invention and do not require the invention to be constructed and operated in a specific orientation; therefore, they should not be construed as limitations on the invention. Furthermore, descriptions involving "first," "second," etc., are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated.

[0037] In the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0038] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0039] like Figures 1-4 As shown, the high-precision seismic data time-spectrum generation method based on sidelobe suppression includes the following steps:

[0040] Step 1: Calculate the analytical signal of the seismic trace from the seismic data;

[0041] Step 2: Determine the instantaneous frequency of the seismic data;

[0042] Step 3: Determine the stable seismic dominant frequency of the seismic data;

[0043] Step 4: Construct the discrete matrix associated with the Kaiser window;

[0044] Step 5: Multiply the discrete matrix with the seismic trace to obtain the time spectrum;

[0045] Step Six: Display and analyze the time spectrum of the earthquake data.

[0046] As a preferred example of this application, the Kaiser window is a window function whose main lobe width and side lobe energy intensity can be adjusted by nonnegative parameters. It is a commonly used window function in the field of digital signal processing and can be mathematically represented as follows:

[0047]

[0048] Where I0() is the Bessel function of the first kind, n is the number of data sampling points, N is the total data length, and β is a non-negative number that controls the main lobe width and side lobe energy intensity. The Kaiser window can degenerate into a rectangular window if and only if β = 0.

[0049] like Figure 2 As shown, Figure 2 a and 2b represent the time-domain and frequency-domain responses of the Kaiser window, respectively. It can be seen that as β increases, the Kaiser window widens in the time domain and decreases the main lobe width in the frequency domain. In addition, it can be seen that as β increases, the sidelobe energy is suppressed to a large extent, indicating that the Kaiser window can not only adjust the resolution in the time and frequency domains, but also obtain higher frequency resolution by suppressing the sidelobes. At the same time, it can be seen that the main lobe width and the sidelobe energy are inversely proportional.

[0050] In addition, such as Figure 3 As shown, Figure 3 The comparison of the time domain (a) and frequency domain (b) responses of Gaussian window and Kaiser window shows that, with almost identical time domain responses, Kaiser window has higher suppression of sidelobes in the frequency domain while maintaining almost identical main lobe intensity. This proves that Kaiser window has a certain advantage over truncated Gaussian window in terms of sidelobe suppression.

[0051] Therefore, this application constructs a discrete matrix using a Kaiser window. By utilizing the Kaiser window's ability to adjust the main lobe width and side lobe energy intensity through parameters, it effectively suppresses side lobe energy, reduces spectral leakage crosstalk between different signal components, and thus improves the frequency resolution of the seismic data's time spectrum.

[0052] As a preferred example of this application, when constructing the discrete matrix related to the Kaiser window, the Kaiser window W transform is used. Compared with the conventional W transform, the Kaiser window W transform utilizes the excellent sidelobe suppression capability of the Kaiser window. The Kaiser window W transform of the seismic signal s(t) can be mathematically defined as:

[0053]

[0054] As can be seen from the above formula, the Kaiser window W transform of seismic data can be regarded as taking the spectrum of the seismic signal after applying a Kaiser window;

[0055] Unlike the W transform, the time window of the Kaiser window W transform method is:

[0056]

[0057] Where β(f) is the instantaneous frequency correlation function, and can be defined as:

[0058]

[0059] Among them, f dom f is the dominant frequency of seismic data obtained in the conventional W transform. diff To differentiate between the instantaneous frequency and the frequency to be analyzed, k is used to change the resolution of the time-frequency domain or frequency domain of the time spectrum by controlling the time window length. Using a technique similar to the W transform, a Kaiser window symmetrical to the dominant frequency of the seismic data can be obtained. By adjusting β(f), the time-frequency window can be scaled in the time-frequency plane, thereby achieving a balance between the time-frequency and frequency domain resolutions of the time spectrum.

[0060] Therefore, this application uses the Kaiser window W transform to construct a discrete matrix. By taking the spectrum of the windowed seismic signal and combining the sidelobe suppression advantage of the Kaiser window, spectral leakage is reduced and the frequency resolution of the time spectrum is improved. Compared with the conventional W transform, it can better highlight the time and frequency characteristics of the seismic signal.

[0061] By associating the time window of the Kaiser window W transform with the instantaneous frequency, the time window can adapt to the instantaneous changes of the seismic signal, laying the foundation for subsequent targeted adjustment of the time-frequency resolution and improving the ability of the time spectrum to capture dynamic changes in the signal.

[0062] This application constructs parameters related to instantaneous frequency based on the stable seismic dominant frequency and instantaneous frequency difference, making time window adjustment more targeted and effectively optimizing different frequency bands;

[0063] By adjusting parameters related to instantaneous frequency, the time-frequency window can be scaled in the time-frequency plane, which can flexibly balance the resolution of the time spectrum in the time and frequency domains and meet the needs of different seismic data processing scenarios for time-frequency feature analysis.

[0064] The stable seismic dominant frequency is obtained by conventional W transform, providing a reliable dominant frequency reference for the subsequent construction of time windows related to instantaneous frequencies, ensuring the accuracy and effectiveness of time window adjustment, and guaranteeing the resolution of the time spectrum.

[0065] As a preferred example of this application, steps four to five further include: converting the Kaiser window W transform into a product of a matrix and a vector, wherein the vector is the original seismic data, the matrix is ​​the coefficient matrix corresponding to the Kaiser window W transform, and the product result is the output time spectrum.

[0066] To implement the Kaiser window W-transform method, it is necessary to propose a Gaussian Kaiser window W-transform for discrete seismic data. The discrete Kaiser window W-transform can be mathematically represented as:

[0067]

[0068] Since the Kaiser window W transform cannot be implemented using the Fast Fourier Transform, its time-frequency analysis efficiency is limited and cannot meet the efficiency requirements of time-frequency analysis of large-scale 3D seismic data. Therefore, this application proposes a Kaiser window W transform based on GPU devices, utilizing the multi-core processing capabilities of GPUs to improve the efficiency of seismic data time-frequency analysis methods. Furthermore, drawing inspiration from the Discrete Fourier Transform, we decompose the above formula into a product of matrices and vectors, which can be expressed as:

[0069] S(t,f;τ)=Kd;

[0070] Where d represents the original seismic data, and K is the coefficient matrix corresponding to the Kaiser window W transform. Using the above formula, the CUDA-provided cublasgemv equation can be used to accelerate time-frequency analysis of seismic data, as follows: Figure 4 As shown;

[0071] This application decomposes the discrete Kaiser window W transform into a product of matrices and vectors, which facilitates efficient computation using computing tools, provides a feasible computational method for time-frequency analysis of large-scale seismic data, and improves processing efficiency.

[0072] By utilizing multi-GPU parallel computing and block processing, the multi-core computing power of GPUs can be fully utilized to perform parallel calculations on multi-channel seismic records, significantly shortening the time for time-frequency analysis and improving the applicability of the method to large-scale 3D seismic data.

[0073] The embodiments of this application have been described above with reference to the accompanying drawings. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. This application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A high-precision seismic data time-spectrum generation method based on sidelobe suppression, characterized in that: Includes the following steps: Step 1: Calculate the analytical signal of the seismic trace from the seismic data; Step 2: Determine the instantaneous frequency of the seismic data; Step 3: Determine the stable seismic dominant frequency of the seismic data; Step 4: Construct the discrete matrix associated with the Kaiser window; Step 5: Multiply the discrete matrix with the seismic trace to obtain the time spectrum; Step Six: Display and analyze the time spectrum of the earthquake data.

2. The high-precision seismic data time-spectrum generation method based on sidelobe suppression according to claim 1, characterized in that: The Kaiser window is a window function whose main lobe width and side lobe energy intensity can be adjusted by a non-negative parameter. When the non-negative parameter is 0, it degenerates into a rectangular window. By adjusting this parameter, the main lobe width and side lobe energy intensity can be inversely adjusted to suppress the side lobe energy.

3. The high-precision seismic data time-spectrum generation method based on sidelobe suppression according to claim 2, characterized in that: When constructing the discrete matrix related to the Kaiser window, the Kaiser window W transform is used, which is to extract the spectrum of the seismic signal after applying the Kaiser window.

4. The high-precision seismic data time-spectrum generation method based on sidelobe suppression according to claim 3, characterized in that: The time window of the Kaiser window W transform is a function related to the instantaneous frequency. This function enables the time window to be scaled in the time-frequency plane to balance the resolution of the time spectrum in the time and frequency domains.

5. The high-precision seismic data time-spectrum generation method based on sidelobe suppression according to claim 4, characterized in that: The instantaneous frequency-related function is constructed based on the stabilized seismic dominant frequency obtained in step three, the difference between the instantaneous frequency obtained in step two and the frequency to be analyzed, and the parameters used to control the time window length. Adjusting the time window through this function can improve the resolution of the Kaiser window in the low-frequency region.

6. The high-precision seismic data time-spectrum generation method based on sidelobe suppression according to claim 5, characterized in that: In step three, the stable earthquake dominant frequency is obtained by performing a conventional W-transform on the earthquake data.

7. The high-precision seismic data time-spectrum generation method based on sidelobe suppression according to claim 2, characterized in that: Steps four and five further include: converting the Kaiser window W transform into a product of a matrix and a vector, where the vector is the original seismic data, the matrix is ​​the coefficient matrix corresponding to the Kaiser window W transform, and the product result is the output time spectrum.

8. The high-precision seismic data time-spectrum generation method based on sidelobe suppression according to claim 7, characterized in that: During the Kaiser window W transform, a block-based processing method is used to perform parallel computation on the multichannel seismic records: the multichannel seismic records are divided into multiple local multichannel seismic records, a local matrix is ​​constructed for each local multichannel seismic record and the Kaiser window W transform is performed to obtain the local time spectrum, and all local time spectra are summed to obtain the multichannel time spectrum.

9. The high-precision seismic data time-spectrum generation method based on sidelobe suppression according to claim 8, characterized in that: The product of the matrix and the vector is accelerated by calling the cublasgemv equation of CUDA, thereby improving computational efficiency.

10. The high-precision seismic data time-spectrum generation method based on sidelobe suppression according to claim 8, characterized in that: The Kaiser window W transform is performed based on a GPU device.

Images