A high-precision stratum correlation method, device and electronic equipment

Abstract

The present specification provides a high-precision stratigraphic correlation method, device and electronic equipment, which constructs a filtering factor through well logging data, and then extracts a high-frequency seismic data body and a low-frequency seismic data body from a seismic data body by using the filtering factor. Through time-varying seismic wavelet deconvolution on the high-frequency, medium-frequency and low-frequency seismic data bodies, a target data body with a "virtual resolution" and a much wider frequency band than the seismic data body can be obtained, so that the profile data of the target data body can be used to more accurately distinguish the interwell lateral characteristics, and the stratigraphic correlation is adjusted based on the interwell lateral characteristics to realize high-precision stratigraphic correlation. The high resolution and lateral variation characteristics of the target data profile are used to finely carry out stratigraphic correlation, thereby providing protection for fine exploration and development of oil and gas fields, especially development geology research.

Description

Technical Field

[0001] This application relates to the field of geological technology, and in particular to a high-precision stratigraphic correlation method, apparatus and electronic equipment. Background Technology

[0002] Stratigraphic correlation is a fundamental task in geological research. Based on understanding the vertical variation patterns of strata, it divides stratigraphic profiles into different types and levels of stratigraphic units according to various characteristics and properties. In practice, especially in comprehensive petroleum geology research, with the increasing understanding of various stratigraphic characteristics and properties, many precise and effective stratigraphic division schemes have emerged, namely, multiple stratigraphic division. Petrological characteristics, biological characteristics, isotopic ages, etc., of strata can all serve as the basis for stratigraphic division. Different bases of division result in different stratigraphic units. For example, lithostratigraphic units can be established based on lithological characteristics, chronostratigraphic units can be established based on temporal attributes, biostratigraphic (or ecological) stratigraphic units can be established based on paleontological (or paleoecological) characteristics, and magnetostratigraphic units can be established based on magnetic properties, etc. Any characteristic or property can serve as the basis for stratigraphic division; that is, the number of stratigraphic units that can be divided or established depends on the number of characteristics and properties of the strata. This is the theoretical basis of multiple stratigraphic division. A stratigraphic profile can be divided into multiple stratigraphic units based on different characteristics and properties, and the boundaries between these units are often inconsistent. To clarify the distribution patterns of strata, stratigraphic division alone is insufficient; the lateral distribution and structural features of strata must be determined through stratigraphic correlation.

[0003] Stratigraphic correlation compares stratigraphic units from different sections to determine their consistency in characteristics or properties, and whether their stratigraphic positions are comparable, thereby understanding their interrelationships and distribution patterns. Broadly defined, stratigraphic correlation includes global, large-scale correlation between different basins, regional correlation within the same sedimentary basin, and correlation of oil (gas) layers within an oilfield. Stratigraphic division studies the changes in strata over time (vertically). In fact, stratigraphic division and stratigraphic correlation are inseparable; stratigraphic division is the foundation of stratigraphic correlation, and stratigraphic correlation, in turn, promotes stratigraphic division. Through stratigraphic division and stratigraphic correlation, the formation time of strata and their position within the geological time system can be determined, thereby clarifying whether there are any missing strata in the study area and the reasons for these missing strata.

[0004] Existing stratigraphic correlation methods fuse well logging data with seismic data to obtain well-seismic sequence stratigraphic data, and then perform stratigraphic correlation based on this data. This approach is limited by the resolution of the seismic data, making it difficult to correlate fourth- and fifth-order high-frequency sequence stratigraphic sequences. Post-stack optimization of the seismic data is needed to improve resolution and imaging accuracy, highlighting the seismic response characteristics of different sequence structures.

[0005] However, because well logging data and seismic data represent strata in two different domains (time domain and depth domain), their resolutions differ significantly. For example, if the stratigraphic scale is too large, the averaging effect is strong, making it difficult to reveal stratigraphic characteristics; if the stratigraphic scale is too small, the resolution of the seismic data may not be sufficient, especially making lateral correlation difficult and making it hard to control lateral sedimentary patterns.

[0006] To address this, existing technologies typically employ various methods to improve seismic data resolution, such as compressing seismic wavelets or removing the effects of interference and tuning during seismic wave propagation underground, thereby broadening the bandwidth of the effective seismic signal, particularly by accurately broadening the high-frequency components. However, these methods for improving seismic data resolution can only extend the effective bandwidth to a limited extent, with limited frequency boosting, resulting in a still significant resolution difference between seismic and well logging data. Summary of the Invention

[0007] This specification provides a high-precision stratigraphic correlation method, apparatus, and electronic equipment to solve the problem of difficulty in lateral correlation caused by the large resolution difference between existing seismic data and well logging data.

[0008] To address the aforementioned technical problems, this specification provides a high-precision stratigraphic correlation method, comprising: acquiring a seismic data volume and a target data volume for a target area; determining well-to-well stratigraphic correlation results based on seismic wave data in the seismic data volume; acquiring a target data profile in the target data volume corresponding to the well-to-well seismic profile based on the well-to-well stratigraphic correlation results, and using the target data profile for fine inter-well stratigraphic correlation; wherein the target data volume is determined by: extracting a time-varying seismic wavelet from the seismic data volume of the target area; constructing a high-frequency filtering factor and a low-frequency filtering factor based on well logging data of the target area; extracting high-frequency information from the seismic data volume using the high-frequency filtering factor to obtain a high-frequency seismic data volume, and extracting low-frequency information from the seismic data volume using the low-frequency filtering factor to obtain a low-frequency seismic data volume; performing diffusion filtering on the seismic data volume, the high-frequency seismic data volume, and the low-frequency seismic data volume respectively; performing deconvolution on the seismic data volume, the high-frequency seismic data volume, and the low-frequency seismic data volume respectively using the time-varying seismic wavelet, and fusing the deconvolution results to obtain the target data volume.

[0009] In some embodiments, based on the well-to-well formation correlation results, a target data profile corresponding to the well-to-well seismic profile is obtained, and the target data profile is used for fine inter-well formation correlation, including: displaying the seismic data profile corresponding to the well-to-well seismic profile on a first display layer through a first display component; the intensity of the displayed color in the seismic data profile is determined according to the value of the seismic data; displaying the target data profile corresponding to the well-to-well seismic profile on a second display layer through a second display component; the intensity of the displayed color is determined according to the value of the target data; wherein the first display layer and the second display layer are overlaid and presented to the user, and the seismic data profile and the target data profile have transparent backgrounds.

[0010] In some embodiments, based on the well-to-well stratigraphic correlation results, a target data profile corresponding to the well-to-well seismic profile is obtained, and the target data profile is used for fine-grained inter-well stratigraphic correlation, including: obtaining lithological feature data volumes, paleontological fossil assemblage data volumes, sedimentary environment data volumes, isotope dating data volumes, seismic data volumes, and target data volumes for the target area; setting virtual time windows on the well-to-well seismic profiles of the lithological feature data volumes, paleontological fossil assemblage data volumes, sedimentary environment data volumes, isotope dating data volumes, seismic data volumes, and target data volumes, with the virtual time windows sliding synchronously from the same position of the data volumes, using the data in each data volume of the virtual time window at the same position as input data for the network model, and outputting stratigraphic division suggestions within the virtual time windows through the network model; and merging the stratigraphic division suggestions corresponding to the virtual time windows at each position to obtain the overall stratigraphic division suggestions for the target area on the well-to-well seismic profiles.

[0011] In some embodiments, when the virtual time window slides on the data volume, it first slides horizontally one by one, then slides vertically once after sliding horizontally to the boundary of the virtual time window, and then slides horizontally again; the distance of each horizontal slide is less than the width of the virtual time window, and the distance of the vertical slide is less than the height of the virtual time window.

[0012] In some embodiments, based on the well-to-well stratigraphic correlation results, a target data profile corresponding to the well-to-well seismic profile is obtained, and the target data profile is used to perform fine correlation of stratigraphic layers between wells. The method further includes: obtaining the user's preliminary stratigraphic division results; comparing the preliminary stratigraphic division results with the overall stratigraphic division suggestions, filtering out inconsistent stratigraphic division data; and presenting the inconsistent stratigraphic division data to the user.

[0013] In some embodiments, the network model includes: a data preprocessing module for concatenating data from lithological feature data volumes, paleontological fossil assemblage data volumes, sedimentary environment data volumes, isotope dating data volumes, seismic data volumes, and target data volumes within a virtual time window at the same location to form a multidimensional feature matrix; a convolution module for performing convolution operations on the multidimensional feature matrix using a two-dimensional convolution kernel to extract the correlation patterns between different data volumes; an LSTM module for processing the convolution results using an LSTM network model to capture stratigraphic vertical continuity and long-range dependencies, the long-range dependencies including sedimentary cycles and stratigraphic contact relationships; and an output module for determining stratigraphic division suggestions based on the output results of the LSTM module.

[0014] A second aspect of this specification provides a high-precision stratigraphic correlation device, comprising: an acquisition unit for acquiring a seismic data volume of a target area and a target data volume; a determination unit for determining well-to-well stratigraphic correlation results based on seismic wave data in the seismic data volume; and a comparison unit for acquiring a target data profile corresponding to a well-to-well seismic profile based on the well-to-well stratigraphic correlation results, and using the target data profile for fine inter-well stratigraphic correlation; the device further comprises: a first extraction unit, a construction unit, a second extraction unit, a filtering unit, and a processing unit for determining the target data volume, wherein the first extraction unit is used to extract time-varying seismic wavelets from the seismic data volume of the target area; The construction unit is used to construct high-frequency filtering factors and low-frequency filtering factors based on well logging data of the target area; the second extraction unit is used to extract high-frequency information from the seismic data volume using the high-frequency filtering factors to obtain high-frequency seismic data volume, and to extract low-frequency information from the seismic data volume using the low-frequency filtering factors to obtain low-frequency seismic data volume; the filtering unit is used to perform diffusion filtering on the seismic data volume, the high-frequency seismic data volume, and the low-frequency seismic data volume respectively; the processing unit is used to perform deconvolution on the seismic data volume, the high-frequency seismic data volume, and the low-frequency seismic data volume respectively using the time-varying seismic wavelet, and to fuse the deconvolution results to obtain the target data volume.

[0015] In some embodiments, the comparison unit includes: a first display subunit, configured to display a seismic data profile corresponding to the well-connected seismic profile on a first display layer via a first display component; the intensity of the displayed color in the seismic data profile is determined according to the value of the seismic data; and a second display subunit, configured to display a target data profile corresponding to the well-connected seismic profile on a second display layer via a second display component; the intensity of the displayed color is determined according to the value of the target data; wherein the first display layer and the second display layer are overlaid and presented to the user, and the seismic data profile and the target data profile are transparent backgrounds.

[0016] In some embodiments, the comparison unit includes: a first acquisition subunit, used to acquire lithological feature data volume, paleontological fossil assemblage data volume, sedimentary environment data volume, isotope dating data volume, seismic data volume, and target data volume of the target area; a processing subunit, used to set virtual time windows on the well-connected seismic profile of the lithological feature data volume, the paleontological fossil assemblage data volume, the sedimentary environment data volume, the isotope dating data volume, the seismic data volume, and the target data volume, the virtual time windows sliding synchronously from the same position of the data volume, using the data in each data volume in the virtual time window at the same position as the input data of the network model, and outputting stratigraphic division suggestions within the virtual time window through the network model; and a fusion subunit, used to fuse the stratigraphic division suggestions corresponding to the virtual time windows at each position to obtain the overall stratigraphic division suggestions of the target area on the well-connected seismic profile.

[0017] In some embodiments, when the virtual time window slides on the data volume, it first slides horizontally one by one, then slides vertically once after sliding horizontally to the boundary of the virtual time window, and then slides horizontally again; the distance of each horizontal slide is less than the width of the virtual time window, and the distance of the vertical slide is less than the height of the virtual time window.

[0018] In some embodiments, the comparison unit further includes: a second acquisition subunit for acquiring the user's preliminary stratigraphic division results; a filtering subunit for comparing the preliminary stratigraphic division results with the overall stratigraphic division recommendations and filtering out inconsistent stratigraphic division data; and a presentation subunit for presenting inconsistent stratigraphic division data to the user.

[0019] In some embodiments, the network model includes: a data preprocessing module for concatenating data from lithological feature data volumes, paleontological fossil assemblage data volumes, sedimentary environment data volumes, isotope dating data volumes, seismic data volumes, and target data volumes within a virtual time window at the same location to form a multidimensional feature matrix; a convolution module for performing convolution operations on the multidimensional feature matrix using a two-dimensional convolution kernel to extract the correlation patterns between different data volumes; an LSTM module for processing the convolution results using an LSTM network model to capture stratigraphic vertical continuity and long-range dependencies, the long-range dependencies including sedimentary cycles and stratigraphic contact relationships; and an output module for determining stratigraphic division suggestions based on the output results of the LSTM module.

[0020] A third aspect of this specification provides an electronic device, comprising: a memory and a processor, wherein the processor and the memory are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to implement the high-precision stratigraphic correlation method described in any of the first aspects.

[0021] A fourth aspect of this specification provides a computer storage medium storing computer program instructions that, when executed by a processor, implement the high-precision stratigraphic correlation method described in any of the first aspects.

[0022] The fifth aspect of this specification provides a computer program product comprising a computer program that, when executed by a processor, implements the high-precision stratigraphic correlation method described in any of the first aspects.

[0023] The high-precision stratigraphic correlation method provided in this manual constructs filter factors from well logging data, and then uses these filter factors to extract high-frequency and low-frequency seismic data volumes from the seismic data volume. By performing deconvolution on the high, medium, and low-frequency seismic data volumes using time-varying seismic wavelets, a target data volume with a "visual resolution" and bandwidth far exceeding that of the seismic data volume can be obtained. Thus, the profile data of the target data volume can be used to more accurately distinguish the lateral features between wells, and the sequence division scheme can be adjusted based on the lateral features between wells to achieve high-precision stratigraphic correlation. The high resolution and lateral variation characteristics of the target data profile are used to conduct detailed stratigraphic correlation, providing a guarantee for the fine exploration and development of oil and gas fields, especially for development geological research. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of this application 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 recorded in this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0025] Figure 1 This document provides a flowchart illustrating a high-precision stratigraphic correlation method.

[0026] Figure 2 A flowchart illustrating the process of determining the target data volume;

[0027] Figure 3 A schematic diagram of the sequence stratigraphic division scheme for well A in the target region DY;

[0028] Figure 4 A schematic diagram of the stratigraphic correlation profile of wells AB in the target area DY;

[0029] Figure 5 A schematic diagram of the well-connected seismic profile for the target area DY;

[0030] Figure 6 This is a schematic diagram of the target data profile of the DY wells in the target area.

[0031] Figure 7 This is a schematic diagram showing the overlay of target data profiles and seismic data profiles from well DY in the target area.

[0032] Figure 8 This is a comparison diagram of the seismic data profile of the DY well in the target area and the target data profile.

[0033] Figure 9 A flowchart illustrating a process for obtaining target data profiles corresponding to well-to-well seismic profiles based on well-to-well stratigraphic correlation results, and using these target data profiles for fine-grained inter-well stratigraphic correlation.

[0034] Figure 10 This is a schematic diagram of the network model structure;

[0035] Figure 11 This is a schematic diagram of another process for fine inter-well stratigraphic comparison, which is based on the stratigraphic correlation results of the well-connected wells to obtain the target data profile corresponding to the seismic profile of the well-connected wells.

[0036] Figure 12 This is a schematic diagram of a high-precision stratigraphic correlation device provided in this specification;

[0037] Figure 13 This is a schematic block diagram of an electronic device provided in this specification. Detailed Implementation

[0038] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this application.

[0039] Oil exploration typically includes stages such as collection, processing, interpretation, and reservoir prediction.

[0040] Data Acquisition: Using techniques such as seismic and well logging, physical information of underground geological bodies is collected, such as seismic wave reflection data and formation resistivity, to provide basic data for subsequent analysis and processing.

[0041] Processing: The massive amounts of collected data undergo a series of processing steps, including denoising, filtering, and correction, to improve data quality, enhance effective information, suppress interference signals, and make the data more conducive to subsequent geological interpretation.

[0042] Explanation: Combining knowledge from multiple disciplines such as geology and geophysics, the processed data is analyzed and interpreted to identify stratigraphic structure, tectonic features, lithological changes, etc., and to infer the possible location and distribution range of underground oil and gas reservoirs.

[0043] Reservoir prediction: Using techniques such as seismic attribute analysis and well logging constrained inversion, parameters such as lithology, physical properties, and oil-bearing properties of reservoirs are predicted to clarify the distribution range and quality differences of reservoirs, providing key reservoir information for oil and gas field development.

[0044] In the "interpretation" phase, it is necessary to synthesize various data to analyze and understand the subsurface geological structure and stratigraphic characteristics. Stratigraphic correlation is an important part of this work. By comparing stratigraphic characteristics between different wells or between seismic data and well data, such as lithology, electrical properties, and characteristics of reflected seismic waves, it is possible to determine the sequence relationship, sedimentary cycles, and lateral variations of the strata. This helps to identify the stratigraphic structure, tectonic features, and the distribution range of oil and gas reservoirs. For example, by comparing logging curves from different wells, analyzing the changes in lithology and physical properties of the strata, and combining this with the reflection characteristics of seismic data, the lateral continuity and variation patterns of the strata can be determined, providing a basis for the interpretation and evaluation of oil and gas reservoirs.

[0045] Existing technologies typically employ various methods to improve seismic data resolution, such as compressing seismic wavelets or removing the effects of interference and tuning during seismic wave propagation underground, thereby broadening the bandwidth of effective seismic signals, particularly by accurately broadening high-frequency components. However, these methods can only extend the effective bandwidth to a limited extent, with limited frequency boosting, resulting in a significant resolution difference between seismic and well logging data, making cross-sectional comparisons difficult.

[0046] In contrast, this specification provides a high-precision stratigraphic correlation method that improves the accuracy of stratigraphic correlation by introducing target data.

[0047] like Figure 1 As shown, this high-precision stratigraphic correlation method includes the following steps S10 to S30.

[0048] S10: Obtain the seismic data volume of the target area and the target data volume.

[0049] In seismic exploration, seismic waves are artificially generated, such as through explosive detonation or the use of controlled seismic sources to produce elastic waves. These waves propagate through the subsurface medium and undergo reflection, refraction, and transmission when they encounter interfaces between strata of different properties. Seismic detectors located on the surface or underground receive these reflected and refracted waves, convert them into electrical signals, and then record them to form seismic wave data.

[0050] A seismic data volume typically organizes and arranges seismic wave data collected within a specific area during seismic exploration according to spatial location and temporal order. It provides a complete description of seismic wave information in three-dimensional space (usually a two-dimensional surface location plus depth). Within the data volume, each data point corresponds to the seismic wave characteristics at a specific location and time underground, such as amplitude and frequency. It can be imagined as a three-dimensional "data cube" containing a massive amount of seismic wave data information.

[0051] Seismic data volumes can also be data volumes obtained after processing seismic wave data. These processing methods can include compressing seismic wavelets, or removing the effects of interference, tuning, and other effects on underground strata during the propagation of seismic waves underground, thereby broadening the frequency band of effective seismic signals, especially broadening the high-frequency components more accurately.

[0052] The target data volume is a three-dimensional dataset formed by organizing and storing target data from each location within the underground strata. Within a three-dimensional spatial grid, each grid point corresponds to a target data point, and these target data points collectively constitute the target data volume.

[0053] The target data volume can be accessed through Figure 2 S11 to S15 are shown.

[0054] S11: Extract time-varying seismic wavelets from the seismic data volume of the target region.

[0055] Specifically, the process involves extracting seismic traces from the well location and then performing equalization on these traces; extracting adjacent traces near the well and performing equalization on these traces; taking time window centers at equal intervals for each equalized seismic trace; processing each time window signal sequentially to obtain the corresponding time window wavelet; performing energy normalization on each time window wavelet after extraction; directly normalizing the time window wavelet corresponding to the first time window signal, and applying windowing normalization to the time window wavelets corresponding to the remaining time window signals, using the first time window as the window function; and finally, smoothing the normalized time window wavelets to obtain time-varying seismic wavelet information.

[0056] S12: Construct high-frequency filter factors and low-frequency filter factors based on well logging data of the target area.

[0057] S12 can first preprocess the logging data of the target area, such as removing outliers, filling missing values, and standardizing the data. Then, it performs a Fourier transform on the preprocessed logging data to obtain the energy distribution of the logging data at different frequencies (spectral analysis results), and presents the spectral analysis results to the user. The user will determine and input a first frequency value and a second frequency value based on the spectral analysis results. The first frequency value is the boundary between low and mid-frequency, and the second frequency value is the boundary between mid and high-frequency. Based on the first and second frequency values ​​input by the user, the boundaries and center frequencies of low and high frequencies in the logging data are determined. Low-frequency filter factors and high-frequency filter factors are constructed based on the boundaries and center frequencies of low and high frequencies.

[0058] Both low-frequency and high-frequency filter factors are values ​​that vary with frequency. The high-frequency filter factor is used to enhance high-frequency data and attenuate low-frequency and mid-frequency data, while the low-frequency filter factor enhances low-frequency data and attenuates high-frequency and mid-frequency data. For example, if the frequency is high, the high-frequency filter factor is greater than 1, and the low-frequency filter factor is less than 1; if the frequency is low, the high-frequency filter factor is less than 1, and the low-frequency filter factor is greater than 1; if the frequency is mid-frequency, both the high-frequency and low-frequency filter factors are less than 1.

[0059] Filter factors come in various forms, such as Gaussian filter factors and Barthold filter factors. Taking the Gaussian filter factor as an example, for high-frequency filtering, its expression can be: Among them, H high (f) is the high-frequency filtering factor, and f is the frequency variable. c This is the center frequency of the high-frequency filter, typically taken as a value near the boundary frequency. σ is a parameter controlling the filter width; the smaller the σ value, the sharper the filtering effect, and the more severely high-frequency components are attenuated. For low-frequency filtering, the expression for the Gaussian filter factor is similar: Among them, H low (f) is the low-frequency filter factor, f c It is the center frequency of the low-frequency filter, and is generally chosen to reflect the low-frequency structure of a large scale. σ is a parameter that controls the filter width. The smaller the value of σ, the sharper the filtering effect and the more severe the attenuation of low-frequency components.

[0060] S13: High-frequency information is extracted from the seismic data volume using the high-frequency filtering factor to obtain a high-frequency seismic data volume, and low-frequency information is extracted from the seismic data volume using the low-frequency filtering factor to obtain a low-frequency seismic data volume.

[0061] S13 multiplies the frequency domain data of the seismic data volume with corresponding high-frequency filter factors. Specifically, it multiplies the frequency domain values ​​of the seismic data corresponding to the same frequency value with the high-frequency filter factor, and then transforms the result to the time domain to obtain the high-frequency seismic data volume. Similarly, it multiplies the frequency domain data of the seismic data volume with corresponding low-frequency filter factors. This process is repeated to obtain the low-frequency seismic data volume.

[0062] S14: Perform diffusion filtering on the seismic data volume, the high-frequency seismic data volume, and the low-frequency seismic data volume respectively.

[0063] In seismic data processing, diffusion filtering can redistribute the energy of seismic waves in time and space. Diffusion filtering has various effects on high, medium, and low-frequency seismic data, exhibiting different characteristics at different frequencies, as detailed below:

[0064] 1. High-frequency seismic data

[0065] Suppressing random noise: High-frequency seismic data often contains a significant amount of random noise. Diffusion filtering can effectively suppress this noise and improve the signal-to-noise ratio by smoothing the data. For example, in seismic exploration, the original high-frequency seismic record may contain some minor interference signals. After diffusion filtering, these noises are significantly suppressed, making the details of the seismic signal clearer.

[0066] Highlighting stratigraphic details: While diffusion filtering can smooth high-frequency signals to some extent, under appropriate parameters, it can suppress noise while preserving subtle stratigraphic features, such as reflections from thin strata and minor faults. This helps to more accurately identify and describe the details of subsurface geological structures, providing a better data foundation for refined geological interpretation.

[0067] 2. Medium-frequency seismic data

[0068] Enhancing the continuity of effective signals: Mid-frequency seismic data typically contains key reflection information from subsurface strata. Diffusion filtering can improve the continuity of these reflection signals, enhancing the continuity and traceability of phase axes. For example, in seismic profiles, mid-frequency data after diffusion filtering shows smoother and more continuous reflection phase axes from the strata, facilitating stratigraphic tracking and comparison by geologists.

[0069] Balancing Amplitude Differences: Different strata have varying reflection coefficients, leading to differences in the amplitude of reflected signals in mid-frequency seismic data. Diffusion filtering can balance these amplitude differences to some extent, resulting in a more uniform overall amplitude distribution in the seismic data. This is beneficial for subsequent analysis and interpretation of the seismic data, avoiding interpretation biases caused by excessive amplitude differences.

[0070] Since seismic data is mainly concentrated in the intermediate frequency range, the effect of diffusion filtering on seismic data is the same as the effect of diffusion filtering on intermediate frequency seismic data.

[0071] 3. Low-frequency seismic data

[0072] Suppressing low-frequency interference: Low-frequency seismic data may contain low-frequency interference signals, such as instrument noise and ground vibration. Diffusion filtering can effectively suppress these low-frequency interferences and improve the quality of low-frequency signals. For example, in some seismic acquisition environments, there is low-frequency mechanical vibration interference, which can be reduced through diffusion filtering, allowing the low-frequency seismic data to more accurately reflect the low-frequency characteristics of underground geological structures.

[0073] Highlighting large-scale structural information: Low-frequency seismic data primarily reflects large-scale geological structural information. Diffusion filtering, while suppressing interference, can further highlight the characteristics of these large-scale structures, such as large folds and fault zones. This is of great significance for understanding the overall morphology and structural pattern of subsurface geological structures macroscopically, providing strong data support for regional geological structural analysis.

[0074] S15: The time-varying seismic wavelet is used to perform deconvolution on the seismic data volume, the high-frequency seismic data volume, and the low-frequency seismic data volume respectively, and the deconvolution results are fused to obtain the target data volume.

[0075] In S15, high-, medium-, and low-frequency seismic data volumes contain geological information at different scales. High-frequency data volumes help identify subtle changes in strata and thin strata, while low-frequency data volumes reflect large-scale geological structures. Using time-varying seismic wavelets for deconvolution allows for optimization of different frequency components, compressing the length of the seismic wavelets, and improving the resolution of the seismic record. This makes stratigraphic interfaces and geological bodies that are difficult to distinguish in conventional seismic records easier to identify, contributing to more accurate geological interpretation and hydrocarbon reservoir description.

[0076] Seismic waves undergo attenuation and dispersion during propagation underground, causing changes in the shape and frequency characteristics of the seismic wavelet. Time-varying seismic wavelet deconvolution can compensate for these changes based on the physical laws of seismic wave propagation and the characteristics of actual seismic data, eliminating the effects of propagation and allowing the deconvolved seismic data to more accurately reflect underground geological conditions. For example, by appropriately enhancing the high-frequency components, the attenuation of high-frequency seismic waves during propagation can be compensated for, thereby better highlighting detailed information about underground strata.

[0077] By using time-varying seismic wavelets to deconvolve high, medium, and low frequency data volumes, the responses of different geological bodies at different frequencies can be separated, which helps to identify and distinguish different types of geological bodies, such as faults, fractures, and lithological abrupt change zones, and provides richer information for stratigraphic correlation.

[0078] The above-mentioned steps S11 to S15 construct filter factors from well logging data, and then use these filter factors to extract high-frequency and low-frequency seismic data volumes from the seismic data volume. This enhances the high-frequency and low-frequency signals in the seismic data, thereby broadening the frequency band of the target data volume. After deconvolving the high, medium, and low-frequency seismic data volumes using time-varying seismic wavelets, the time-varying seismic wavelets can be removed from the seismic data, resulting in a target data volume that reflects stratigraphic interfaces and geological bodies. Therefore, the "apparent resolution" of the target data volume profile is much higher than that of conventional seismic data profiles, clearly reflecting stratigraphic transformation characteristics and highlighting the longitudinal and lateral variations of the stratigraphy. It can be used to conduct high-precision sequence stratigraphic correlation.

[0079] S20: Determine the well-to-well stratigraphic correlation results based on the seismic wave data in the seismic data volume.

[0080] Before conducting well-to-well correlation, single-well sequence stratigraphy can be performed first. Based on this single-well sequence stratigraphy, a well-to-well stratigraphic correlation profile can be established, ensuring consistency and comparability of the same stratigraphic sequence across different wells. Then, a well-to-well seismic correlation framework is established using seismic wave data. Based on the well-to-well stratigraphic correlation, the amplitude and frequency variations of seismic phase axes are used to correlate strata and sand bodies between wells.

[0081] S30: Based on the well-to-well stratigraphic correlation results, obtain the target data profile in the target data volume corresponding to the well-to-well seismic profile, and use the target data profile to perform fine correlation of the stratigraphic layers between wells.

[0082] Due to the differences between well logging and seismic data, well-seismic mismatch often exists in stratigraphic correlation. This necessitates further adjustments to the sequence stratigraphy scheme by incorporating seismic variation characteristics. However, due to the resolution differences between well and seismic data (i.e., well logging data has high resolution while seismic data has low resolution), directly using seismic data for stratigraphic correlation (especially in some small-scale stratigraphic correlations) is insufficient. The resolution of the target data volume is much higher than that of the seismic data volume, allowing for further adjustments to the sequence stratigraphy scheme using the corresponding profile data within the target data volume.

[0083] The high-precision stratigraphic correlation method provided in this manual constructs filter factors from well logging data, and then uses these filter factors to extract high-frequency and low-frequency seismic data volumes from the seismic data volume. By performing deconvolution on the high, medium, and low-frequency seismic data volumes using time-varying seismic wavelets, a target data volume with a "visual resolution" and bandwidth far exceeding that of the seismic data volume can be obtained. Thus, the profile data of the target data volume can be used to more accurately distinguish the lateral features between wells, and the sequence division scheme can be adjusted based on the lateral features between wells to achieve high-precision stratigraphic correlation. The high resolution and lateral variation characteristics of the target data profile are used to conduct detailed stratigraphic correlation, providing a guarantee for the fine exploration and development of oil and gas fields, especially for development geological research.

[0084] Existing stratigraphic correlation methods involve: first, core logging curves are calibrated, assigning sedimentological meaning to the curves and identifying the stratigraphic stacking pattern; then, different levels of base-level cycles are delineated within a single well, and multi-well correlation is performed. Finally, the amplitude, frequency, phase, continuity, and geometric morphology of reflected waves at different stratigraphic interfaces are analyzed to establish a high-resolution sequence stratigraphic framework for the study area. However, this method requires separating reflected waves from the seismic record. Due to the influence of reflected waves, refracted waves, bypass waves, and interference noise at different stratigraphic interfaces, the separation process is complex and inaccurate, with a narrow frequency band. The accuracy of stratigraphic correlation is far lower than the high-precision stratigraphic correlation method provided in this specification.

[0085] The following example, using the target area DY in a bay basin, illustrates the high-precision stratigraphic correlation method provided in this manual. The target stratigraphic area DY has thin sand bodies, poor continuity, and rapid inter-well variations, making lateral correlation difficult. The correlation results of the stratigraphy and sand bodies directly affect the analysis of reservoir distribution patterns and development geological studies. The high-precision stratigraphic correlation method provided in this manual for stratigraphic correlation in the target area DY includes the following steps one through three.

[0086] Step 1: Single-well sequence stratigraphy

[0087] First, single-well sequence division was carried out: taking Well A as an example, based on logging curves, lithology, and cycle characteristics, the target area DY was divided into three sections and eight oil groups: the first section is a third-order sequence, which is an overall reverse cycle with multiple high-frequency cycles inside, and is further subdivided into three oil groups according to the cycle transition surface; the second section is a third-order sequence, which is a complete cycle with one positive and one negative cycle, and the upper part has the largest lacustrine flooding surface of the second-order sequence, which can be used as a correlation marker layer for the whole area, and is further subdivided into three oil groups according to the cycle transition surface; the third section is an overall positive cycle, and is subdivided into upper and lower oil groups. Figure 3 A schematic diagram of the sequence division scheme for well A in the target region DY.

[0088] Step 2: Well-to-well stratigraphic correlation

[0089] Based on single-well sequence stratigraphy, a well-connected stratigraphic correlation profile was established, ensuring consistency and comparability of the same stratigraphic set across different wells. Well-connected correlation reveals that the overall stratigraphic development characteristics of the DY interface cycles and lithological assemblages in the target area are similar, showing good consistency. However, several differences exist: in the upper Ed3 sub-member, well A exhibits a positive cycle, while well B shows a complete cycle; in the lower Ed3 sub-member, well A shows both positive and negative complete cycles, while well B shows two superimposed positive cycles. Furthermore, stratigraphic variations between different wells require further investigation using seismic data. Figure 4 This is a schematic diagram of the stratigraphic correlation profile of wells AB in the target area DY.

[0090] Step 3: Well-seismic stratigraphic correlation

[0091] Using conventional seismic data, a well-to-well seismic correlation framework was established. Based on the well-to-well stratigraphic correlation, the amplitude and frequency variation characteristics of the seismic phase axis were used to further conduct fine correlation of strata and sand bodies between wells. Figure 5 This is a schematic diagram of the seismic profile of the target area DY. The seismic profile shows that all eight oil formations in the target area DY are developed throughout the region, exhibiting good overall consistency.

[0092] The discrepancies between well logging and seismic data often lead to well-seismic mismatch issues in stratigraphic correlation. For example... Figure 5 The stratigraphic division schemes based on well logging information from the two wells do not match perfectly on the seismic profiles. This necessitates further adjustments to the sequence division scheme by incorporating the seismic variation characteristics. However, due to the difference in well-seismic resolution (high well logging resolution, low seismic resolution), directly using seismic data for stratigraphic correlation is insufficient, especially in some small-scale stratigraphic correlations. Therefore, a higher-resolution data is needed to assist in conducting fine stratigraphic correlation.

[0093] Step Four:

[0094] Utilizing the high-resolution vertical and horizontal characteristics of the target data volume profile, further detailed correlation of strata and sand bodies between wells is conducted based on well-to-well stratigraphic correlation to finalize the stratigraphic correlation scheme. By combining well logging curves and lithological characteristics with the lateral variations of the target data, the lateral stratigraphic and lithological variation characteristics between wells are inferred. The target data volume profile prominently displays the lateral variation characteristics of sand bodies, accurately locating the sand body extinction points, whereas conventional seismic profiles are limited by resolution, and the phase axes cannot effectively reflect sand body variations. Figure 5 and Figure 6 As shown, where Figure 6 This is a schematic diagram of the target data profile of the DY well in the target area.

[0095] From the overlay diagram of the target data profile and the seismic data profile, it can be seen that the target data profile removes the sidelobe interference of the seismic wavelet, and can better reflect the changing characteristics of the lithological interface. The vertical resolution and lateral variation of the target data are more accurate than those of the seismic profile, and can be used as auxiliary data for fine stratigraphic correlation between wells. Figure 7 This is a schematic diagram showing the overlay of target data profiles and seismic data profiles for the DY well network in the target area.

[0096] Based on the well logging curves and lithological characteristics, and their corresponding changes on conventional seismic profiles and target data profiles, it can be seen that one set of seismic reflections on the conventional seismic profile corresponds to two sets of reflections on the target data profile, and the waveform characteristics of the seismic reflections correspond well with the well logging GR curves. Figure 8 This is a comparison chart of the seismic data profiles of well DY in the target area and the target data profile. Figure 8 (a) in the image is a conventional seismic profile. Figure 8 (b) in the figure represents the target data profile. In (a), the green box shows one set of wave crest reflections (black), while (b) shows two sets of reflection interfaces (black). Based on the well logging GR curve (blue curve in the figure) and lithological data (yellow for sandstone, dark gray for mudstone), multiple thin sandstone layers exist. Changes in the GR curve (lithological change interfaces) correspond to stratigraphic interfaces, but due to insufficient resolution of conventional seismic profiles, the correspondence is poor; however, it is better aligned with the target data profile.

[0097] In some embodiments, S30 includes: displaying a seismic data profile corresponding to the well-connected seismic profile on a first display layer via a first display component; and displaying a target data profile corresponding to the well-connected seismic profile on a second display layer via a second display component; wherein the first display layer and the second display layer are overlaid and presented to the user.

[0098] Because both the seismic data profile and the target data profile have transparent backgrounds, they are clearly visible when overlaid. Different colors can be used to visually represent the seismic and target data, making it easier to distinguish them from the overlay display. For example... Figure 7 As shown, this allows for fine-grained comparison of stratigraphic layers between wells based on the overlay display results of seismic data profiles and target data profiles, such as adjusting the sequence stratigraphy scheme.

[0099] In some embodiments, such as Figure 9 As shown, S30 includes the following S31 to S33.

[0100] S31: Acquire lithological characteristic data, paleontological fossil assemblage data, sedimentary environment data, isotope dating data, seismic data, and target data for the target area.

[0101] The lithological data volume records the physical properties, mineral composition, and structural characteristics of rocks within the target area. The paleontological fossil assemblage data volume records the types, quantities, preservation states, and assemblage characteristics of fossils in the strata of the target area, reflecting information about biological communities during geological history. The sedimentary environment data volume records the geographical environment, hydrodynamic conditions, and sediment supply during the deposition of strata in the target area. The isotopic dating data volume records the formation age of rocks or minerals determined by measuring the decay patterns of radioactive isotopes in the strata.

[0102] S32: A virtual time window is set on the well-connected seismic profile of the lithological feature data volume, the paleontological fossil assemblage data volume, the sedimentary environment data volume, the isotope dating data volume, the seismic data volume, and the target data volume. The virtual time window slides synchronously from the same position of the data volume. The data in each data volume in the virtual time window at the same position is used as the input data of the network model. The stratigraphic division suggestions within the virtual time window are output by the network model.

[0103] On the well-connected seismic profile, the virtual time window can slide along a "bow" shaped path or a document line break path.

[0104] In some embodiments, such as Figure 10 As shown, the network model includes a data preprocessing module, a convolution module, an LSTM module, and an output module.

[0105] The data preprocessing module is used to stitch together data from lithological features, paleontological fossil assemblages, sedimentary environments, isotope dating, seismic data, and target data within the same virtual time window to form a multidimensional feature matrix.

[0106] The data volume in a virtual time window can be treated as a two-dimensional data volume. By concatenating the two-dimensional data of various data volumes, a three-dimensional data volume can be obtained. For example, if the size of the virtual time window is M (length) × N (height), then the multi-dimensional feature matrix formed by concatenation is M × N × 6, where M and N are positive integers.

[0107] The convolution module is used to perform convolution operations on the multidimensional feature matrix using a two-dimensional convolution kernel to extract the correlation patterns between different data volumes.

[0108] The size of the two-dimensional convolution kernel can be N×6. The convolution operation is performed by the kernel and the M×N×6 multidimensional feature matrix. The convolution result is the association pattern between different data volumes.

[0109] The LSTM module is used to process convolution results using an LSTM network model to capture vertical continuity and long-range dependencies of formations, including sedimentary cycles and formation contact relationships.

[0110] The output module is used to determine stratigraphic partitioning suggestions based on the output of the LSTM network model.

[0111] The output module can output the probability that a location in the profile belongs to each of the various reference strata, and finally output the identifier of the stratum with the highest probability value. The reference strata can be defined by the user based on well logging data, i.e., the single-well sequence stratigraphy result. The well-connected data profile of the aforementioned seismic data volume can include single-well sequence stratigraphy results to facilitate the extraction of various reference strata.

[0112] S33: Merge the stratigraphic division suggestions corresponding to the virtual time windows at each location to obtain the overall stratigraphic division suggestions for the target area on the well-connected seismic profile.

[0113] When the virtual time window slides on the data volume, it first slides horizontally one by one, then slides vertically once after sliding horizontally to the boundary of the virtual time window, and then slides horizontally again; the distance of each horizontal slide is less than the width of the virtual time window, and the distance of the vertical slide is less than the height of the virtual time window.

[0114] Different stratigraphic classification suggestions may be determined for the same location when the virtual time window is in different locations. When merging the stratigraphic classification suggestions corresponding to the virtual time windows of various locations, the average value of the different stratigraphic classification suggestions can be calculated, or the maximum value can be taken.

[0115] The steps S31 to S33 above involve using a network model to learn the correlation between the ensemble of multimodal data (i.e., lithological characteristic data, paleontological fossil assemblage data, sedimentary environment data, isotopic dating data, seismic data volumes, and target data) on the same cross-section and the stratigraphic division results. By directly inputting the cross-sectional data of these data volumes into the network model, overall stratigraphic division suggestions can be obtained quickly and accurately. Using these overall stratigraphic division suggestions for stratigraphic correlation can significantly reduce the workload for users.

[0116] Among the aforementioned lithological feature data volume, paleontological fossil assemblage data volume, sedimentary environment data volume, isotopic dating data volume, seismic data volume, and target data volume, the target data volume has a wider bandwidth and higher resolution, which can significantly improve the accuracy of the overall stratigraphic division suggestions output by the network model.

[0117] It should be noted that lithological feature data volumes, paleontological fossil assemblage data volumes, sedimentary environment data volumes, isotope dating data volumes, and seismic data volumes usually have low resolution and may have missing data. It is usually impossible to obtain accurate stratigraphic division results by processing these data volumes using a network model. Therefore, the technical solutions described in S31 to S33 are not available in the existing technology.

[0118] Furthermore, such as Figure 11 As shown, S30 may also include S34 to S36 as follows.

[0119] S34: Obtain the user's preliminary stratigraphic division results.

[0120] S35: Compare the preliminary stratigraphic division results with the overall stratigraphic division recommendations, and filter out inconsistent stratigraphic division data.

[0121] S36: Presenting inconsistent stratigraphic data to users.

[0122] Users may first determine a preliminary stratigraphic division result based on single-well sequence stratigraphic division, well-to-well stratigraphic correlation, and well-seismic stratigraphic correlation, such as Figure 4 As shown. This preliminary stratigraphic division result is the result of human intervention, and single-well sequence division is usually quite accurate. However, the inter-well sequence division in the preliminary stratigraphic division result is determined based on seismic data, which has a low resolution, so inaccuracies in inter-well sequence division are inevitable.

[0123] The preliminary stratigraphic division results can be compared with the overall stratigraphic division suggestions heard by the network model. Inconsistent stratigraphic division data can be filtered out and presented to the user. A user confirmation component can also be provided in the interactive interface so that the user acknowledges the inconsistent stratigraphic division data and determines whether the preliminary stratigraphic division results or the overall stratigraphic division suggestions should prevail. Users can also choose to ignore the inconsistent stratigraphic division data, i.e., use the preliminary stratigraphic division results as the standard.

[0124] By combining the results of manual division with those of the network model through S34 and S36, the workload of manual division can be reduced and the accuracy of the final stratigraphic correlation can be improved.

[0125] This specification also provides a high-precision stratigraphic correlation device, which can be used to implement the above-mentioned high-precision stratigraphic correlation method. For example... Figure 12 As shown, it includes an acquisition unit 10, a determination unit 20, a comparison unit 30, a first extraction unit 40, a construction unit 50, a second extraction unit 60, a filtering unit 70, and a processing unit 80.

[0126] The acquisition unit 10 is used to acquire the seismic data volume and target data volume of the target area.

[0127] Unit 20 is used to determine the well-to-well stratigraphic correlation results based on the seismic wave data in the seismic data volume.

[0128] The comparison unit 30 is used to obtain the target data profile in the target data volume corresponding to the well seismic profile based on the well-connected formation comparison results, and to use the target data profile to perform fine comparison of formations between wells.

[0129] The first extraction unit 40 is used to extract time-varying seismic wavelets from the seismic data volume of the target area.

[0130] The construction unit 50 is used to construct high-frequency filtering factors and low-frequency filtering factors based on well logging data of the target area.

[0131] The second extraction unit 60 is used to extract high-frequency information from the seismic data volume using the high-frequency filtering factor to obtain a high-frequency seismic data volume, and to extract low-frequency information from the seismic data volume using the low-frequency filtering factor to obtain a low-frequency seismic data volume.

[0132] The filtering unit 70 is used to perform diffusion filtering on the seismic data volume, the high-frequency seismic data volume, and the low-frequency seismic data volume, respectively.

[0133] The processing unit 80 is used to perform deconvolution on the seismic data volume, the high-frequency seismic data volume, and the low-frequency seismic data volume respectively using the time-varying seismic wavelet, and fuse the deconvolution results to obtain the target data volume.

[0134] In some embodiments, the comparison unit includes a first display subunit and a second display subunit.

[0135] The first display subunit is used to display the seismic data profile corresponding to the well-connected seismic profile on the first display layer through the first display component; the depth of the color displayed in the seismic data profile is determined according to the value of the seismic data.

[0136] The second display subunit is used to display the target data profile corresponding to the well-connected seismic profile on the second display layer through the second display component; the depth of the display color is determined according to the value of the target data.

[0137] The first and second display layers are superimposed and presented to the user, with the seismic data profile and target data profile having a transparent background.

[0138] In some embodiments, the comparison unit includes a first acquisition subunit, a processing subunit, and a fusion subunit.

[0139] The first acquisition subunit is used to acquire lithological feature data, paleontological fossil assemblage data, sedimentary environment data, isotope dating data, seismic data, and target data for the target area.

[0140] The processing subunit is used to set virtual time windows on the well-connected seismic profiles of the lithological feature data volume, the paleontological fossil assemblage data volume, the sedimentary environment data volume, the isotope dating data volume, the seismic data volume, and the target data volume. The virtual time windows slide synchronously from the same position in the data volume, and use the data in each data volume in the virtual time window at the same position as the input data of the network model. The network model outputs stratigraphic division suggestions within the virtual time window.

[0141] The fusion sub-unit is used to merge the stratigraphic division suggestions corresponding to the virtual time windows at each location to obtain the overall stratigraphic division suggestions for the target area on the well-connected seismic profile.

[0142] In some embodiments, when the virtual time window slides on the data volume, it first slides horizontally one by one, then slides vertically once after sliding horizontally to the boundary of the virtual time window, and then slides horizontally again; the distance of each horizontal slide is less than the width of the virtual time window, and the distance of the vertical slide is less than the height of the virtual time window.

[0143] In some embodiments, the comparison unit further includes a second acquisition subunit, a filtering subunit, and a presentation subunit.

[0144] The second acquisition subunit is used to acquire the user's preliminary stratigraphic division results.

[0145] The filtering subunit is used to compare the preliminary stratigraphic division results with the overall stratigraphic division recommendations and filter out inconsistent stratigraphic division data.

[0146] The presentation sub-unit is used to present inconsistent stratigraphic data to the user.

[0147] In some embodiments, the network model includes a data preprocessing module, a convolution module, an LSTM module, and an output module.

[0148] The data preprocessing module is used to stitch together data from lithological features, paleontological fossil assemblages, sedimentary environments, isotope dating, seismic data, and target data within the same virtual time window to form a multidimensional feature matrix.

[0149] The convolution module is used to perform convolution operations on the multidimensional feature matrix using a two-dimensional convolution kernel to extract the correlation patterns between different data volumes.

[0150] The LSTM module is used to process convolution results using an LSTM network model to capture vertical continuity and long-range dependencies of formations, including sedimentary cycles and formation contact relationships.

[0151] The output module is used to determine stratigraphic subdivision recommendations based on the output of the LSTM module.

[0152] This invention also provides an electronic device, such as... Figure 13 As shown, the electronic device may include a processor 1301 and a memory 1302, wherein the processor 1301 and the memory 1302 may be connected via a bus or other means. Figure 13 Taking the example of a connection between China and Israel via a bus.

[0153] Processor 1301 may be a central processing unit (CPU). Processor 1301 may also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or combinations thereof.

[0154] Memory 1302, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the high-precision stratigraphic correlation method in this embodiment of the invention (e.g., Figure 12 The diagram shows an acquisition unit 10, a determination unit 20, a comparison unit 30, a first extraction unit 40, a construction unit 50, a second extraction unit 60, a filtering unit 70, and a processing unit 80. The processor 1301 executes various functional applications and data processing by running non-transitory software programs, instructions, and modules stored in the memory 1302, thereby realizing the high-precision stratigraphic correlation method in the above method embodiment.

[0155] The memory 1302 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created by the processor 1301, etc. Furthermore, the memory 1302 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory 1302 may optionally include memory remotely located relative to the processor 1301, and these remote memories may be connected to the processor 1301 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0156] The one or more modules are stored in the memory 1302, and when executed by the processor 1301, the above-mentioned high-precision stratigraphic correlation method is performed.

[0157] The specific details of the above-mentioned electronic device can be understood by referring to the relevant descriptions and effects in the method embodiments, and will not be repeated here.

[0158] This specification also provides a computer storage medium storing computer program instructions, which, when executed by a processor, implement the steps of the above-described high-precision stratigraphic correlation method.

[0159] This specification also provides a computer program product comprising a computer program that, when executed by a processor, implements the steps of the above-described high-precision stratigraphic correlation method.

[0160] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc.; the storage medium can also include combinations of the above types of memory.

[0161] The various embodiments in this specification are described in a progressive manner. For the same or similar parts between the various embodiments, please refer to each other. The focus of each embodiment is to describe the differences from other embodiments.

[0162] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions.

[0163] For ease of description, the above devices are described separately by function as various units. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware.

[0164] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute certain parts of the methods of various embodiments of this application.

[0165] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc.

[0166] This application can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0167] Although this application has been described through embodiments, those skilled in the art will know that this application has many modifications and variations without departing from the spirit of this application, and it is intended that the appended claims cover such modifications and variations without departing from the spirit of this application.

Claims

1. A high-precision stratigraphic correlation method, characterized in that, include: Acquire the seismic data volume of the target area and the target data volume; Determine the well-to-well stratigraphic correlation results based on the seismic wave data in the seismic data volume; Based on the well-to-well stratigraphic correlation results, target data profiles are obtained from the target data volumes corresponding to the well-to-well seismic profiles. These target data profiles are then used for fine-grained inter-well stratigraphic correlation. Furthermore, based on the lateral variation characteristics of the sand bodies highlighted in the target data profiles, the extermination points of the sand bodies are determined. The target data volumes are determined in the following manner: Extract time-varying seismic wavelets from the seismic data volume of the target region; Construct high-frequency and low-frequency filter factors based on well logging data from the target area; The high-frequency information is extracted from the seismic data volume using the high-frequency filtering factor to obtain a high-frequency seismic data volume, and the low-frequency information is extracted from the seismic data volume using the low-frequency filtering factor to obtain a low-frequency seismic data volume; Diffusion filtering is performed on the seismic data volume, the high-frequency seismic data volume, and the low-frequency seismic data volume respectively; The time-varying seismic wavelet is used to perform deconvolution on the seismic data volume, the high-frequency seismic data volume, and the low-frequency seismic data volume respectively. The deconvolution results are then fused to obtain the target data volume. Fine-scale inter-well formation correlation is performed using the target data profile, including: Acquire lithological feature data, paleontological fossil assemblage data, sedimentary environment data, isotope dating data, seismic data, and target data for the target area; Virtual time windows are set on the well-connected seismic profiles of the lithological feature data volume, the paleontological fossil assemblage data volume, the sedimentary environment data volume, the isotope dating data volume, the seismic data volume, and the target data volume. The virtual time windows slide synchronously from the same position in the data volume. The data in each data volume in the virtual time window at the same position are used as the input data of the network model. The network model outputs stratigraphic division suggestions within the virtual time window. By merging the stratigraphic classification suggestions corresponding to the virtual time windows at each location, the overall stratigraphic classification suggestions for the target area on the well-connected seismic profile are obtained; The network model includes: The data preprocessing module is used to stitch together data from lithological features, paleontological fossil assemblages, sedimentary environments, isotope dating, seismic data, and target data within the same virtual time window to form a multidimensional feature matrix. The convolution module is used to perform convolution operations on the multidimensional feature matrix using a two-dimensional convolution kernel to extract the correlation patterns between different data volumes. The LSTM module is used to process convolution results using an LSTM network model to capture vertical continuity and long-range dependencies of formations, including sedimentary cycles and formation contact relationships. The output module is used to determine stratigraphic subdivision suggestions based on the output of the LSTM module.

2. The method according to claim 1, characterized in that, Based on the well-to-well stratigraphic correlation results, target data profiles corresponding to the well-to-well seismic profiles are obtained. These target data profiles are then used for fine-grained inter-well stratigraphic correlation, including: The first display component displays the seismic data profile corresponding to the well-connected seismic profile on the first display layer; the intensity of the color displayed in the seismic data profile is determined according to the value of the seismic data. The second display component displays the target data profile corresponding to the well-connected seismic profile on the second display layer; the depth of the display color is determined according to the value of the target data. The first and second display layers are superimposed and presented to the user, with the seismic data profile and target data profile having a transparent background.

3. The method according to claim 1, characterized in that, When the virtual time window slides on the data volume, it first slides horizontally one by one, then slides vertically once after sliding horizontally to the boundary of the virtual time window, and then slides horizontally again; the distance of each horizontal slide is less than the width of the virtual time window, and the distance of the vertical slide is less than the height of the virtual time window.

4. The method according to claim 1, characterized in that, Based on the well-to-well stratigraphic correlation results, target data profiles corresponding to the well-to-well seismic profiles are obtained. Fine-scale inter-well stratigraphic correlation is then performed using these target data profiles. The process also includes: Obtain the user's preliminary stratigraphic division results; The preliminary stratigraphic division results are compared with the overall stratigraphic division recommendations, and inconsistent stratigraphic division data are filtered out. Inconsistent stratigraphic data is presented to users.

5. A high-precision stratigraphic correlation device, characterized in that, include: The acquisition unit is used to acquire the seismic data volume and target data volume of the target area. The unit is defined to determine the well-to-well stratigraphic correlation results based on the seismic wave data in the seismic data volume. The comparison unit is used to obtain the target data profile in the target data body corresponding to the well seismic profile based on the well-connected formation comparison results, use the target data profile to perform fine comparison of formations between wells, and determine the sand body annihilation point based on the lateral variation characteristics of the sand body highlighted by the target data profile. The apparatus further includes a first extraction unit, a construction unit, a second extraction unit, a filtering unit, and a processing unit for determining the target data volume; wherein, The first extraction unit is used to extract time-varying seismic wavelets from the seismic data volume of the target area; The construction unit is used to construct high-frequency filtering factors and low-frequency filtering factors based on well logging data of the target area. The second extraction unit is used to extract high-frequency information from the seismic data volume using the high-frequency filtering factor to obtain a high-frequency seismic data volume, and to extract low-frequency information from the seismic data volume using the low-frequency filtering factor to obtain a low-frequency seismic data volume; The filtering unit is used to perform diffusion filtering on the seismic data volume, the high-frequency seismic data volume, and the low-frequency seismic data volume, respectively. The processing unit is used to perform deconvolution on the seismic data volume, the high-frequency seismic data volume, and the low-frequency seismic data volume respectively using the time-varying seismic wavelet, and fuse the deconvolution results to obtain the target data volume; The comparison unit includes: The first acquisition subunit is used to acquire lithological feature data, paleontological fossil assemblage data, sedimentary environment data, isotope dating data, seismic data, and target data of the target area; The processing subunit is used to set virtual time windows on the well-connected seismic profiles of the lithological feature data volume, the paleontological fossil assemblage data volume, the sedimentary environment data volume, the isotope dating data volume, the seismic data volume, and the target data volume. The virtual time windows slide synchronously from the same position in the data volume, and use the data in each data volume in the virtual time window at the same position as the input data of the network model. The network model outputs stratigraphic division suggestions within the virtual time window. The fusion sub-unit is used to merge the stratigraphic division suggestions corresponding to the virtual time windows at various locations to obtain the overall stratigraphic division suggestions for the target area on the well-connected seismic profile. The network model includes: The data preprocessing module is used to stitch together data from lithological features, paleontological fossil assemblages, sedimentary environments, isotope dating, seismic data, and target data within the same virtual time window to form a multidimensional feature matrix. The convolution module is used to perform convolution operations on the multidimensional feature matrix using a two-dimensional convolution kernel to extract the correlation patterns between different data volumes. The LSTM module is used to process convolution results using an LSTM network model to capture vertical continuity and long-range dependencies of formations, including sedimentary cycles and formation contact relationships. The output module is used to determine stratigraphic subdivision suggestions based on the output of the LSTM module.

6. An electronic device, characterized in that, include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to implement the high-precision stratigraphic correlation method according to any one of claims 1 to 4.

7. A computer storage medium, characterized in that, The computer storage medium stores computer program instructions, which, when executed by a processor, implement the high-precision stratigraphic correlation method according to any one of claims 1 to 4.

8. A computer program product, characterized in that, It includes a computer program that, when executed by a processor, implements the high-precision stratigraphic correlation method according to any one of claims 1 to 4.

Images