Method and system for rapid integration of oil and gas exploration data based on bidirectional cascading

Abstract

This application provides a method and system for rapid integration of oil and gas exploration data based on bidirectional cascade. The method includes: S1: acquiring the original exploration data stream and performing normalization processing to form a standardized initial data volume; S2: extracting horizontally distributed seismic reflection interface feature points and vertically distributed well logging lithological inflection points to construct a bidirectional cascade topological skeleton; S3: performing waveform similarity propagation of seismic attributes horizontally and recursive derivation of rock physical parameters of well logging attributes vertically, and generating an intermediate integrated dataset; S4: acquiring gravity and magnetic exploration data and generating a geological structural background field, performing spatial registration and weight superposition, and correcting stratigraphic boundary drift; S5: backtracking and updating the point-to-point constraint relationships in the bidirectional cascade topological skeleton to generate high-precision integrated oil and gas exploration data after closed-loop verification. Its effect is to effectively improve the accuracy of the integrated data, thereby meeting the actual needs of high-precision oil and gas exploration.

Description

Technical Field

[0001] This invention relates to data processing technology in the field of oil and gas exploration, and in particular to a method and system for rapid integration of oil and gas exploration data based on bidirectional cascading. Background Technology

[0002] In oil and gas exploration, it is necessary to integrate seismic data, well logging data, and gravity and magnetic exploration data from geophysical acquisition to form a comprehensive and accurate exploration data model, providing data support for oil and gas resource exploration. Current technologies for integrating oil and gas exploration data often employ a step-by-step processing approach. Seismic and well logging data are first standardized separately, and then multi-source data are fused through simple overlay. Some methods introduce gravity and magnetic data for auxiliary correction, but no deep correlation mechanism between seismic and well logging data has been established.

[0003] In conventional data integration methods, seismic and well logging data are processed separately, lacking a connecting structure that links the two. This results in a failure to accurately match the horizontally distributed seismic waveform characteristics with the vertically distributed rock physical properties, leading to a disconnect between horizontal and vertical information in the integrated data. Furthermore, the resolution differences between individual data sources mean that simple overlay cannot effectively correct for stratigraphic boundary drift, and the lack of an effective feedback and verification mechanism makes it difficult to dynamically adjust for deviations during the integration process. Ultimately, this results in insufficient accuracy of the integrated data, failing to meet the practical needs of high-precision oil and gas exploration. Summary of the Invention

[0004] In view of this, the primary objective of this invention is to provide a rapid integration method for oil and gas exploration data based on bidirectional cascading. By dynamically adjusting the deviations during the integration process, the accuracy of the integrated data can be effectively improved, thereby meeting the actual needs of high-precision oil and gas exploration.

[0005] To achieve the above objectives, the specific technical solution adopted by the present invention is as follows:

[0006] A rapid integration method for oil and gas exploration data based on bidirectional cascading, the key of which includes the following steps:

[0007] S1: Acquire the raw exploration data stream and normalize it according to the stratigraphic depth slices to form a standardized initial data volume;

[0008] S2: Based on the standardized initial data volume, extract the horizontally distributed seismic reflection interface feature points and the vertically distributed well logging lithology inflection points to construct a two-way cascaded topological skeleton. The two-way cascaded topological skeleton defines the point-to-point constraint relationship between the seismic profile and the well logging curve.

[0009] S3: Apply the bidirectional cascaded topology skeleton to the standardized initial data volume, perform waveform similarity propagation of seismic attributes in the horizontal direction, perform rock physical parameter recursion of well logging attributes in the vertical direction, and generate an intermediate integrated dataset that integrates horizontal waveform features and vertical rock physical attributes.

[0010] S4: Acquire gravity and magnetic exploration data and generate a geological structure background field. Perform spatial registration and weight superposition on the intermediate integrated dataset to correct the stratigraphic boundary drift caused by the resolution limitations of a single data source such as earthquake or well logging.

[0011] S5: Based on the corrected stratigraphic boundary drift results, backtrack and update the point-to-point constraint relationships in the bidirectional cascaded topology skeleton to generate high-precision integrated oil and gas exploration data after closed-loop verification.

[0012] Optionally, the raw exploration data stream includes a raw seismic data stream and a raw well logging data stream. The raw seismic data stream includes seismic wave travel time and amplitude envelope, and the raw well logging data stream includes well logging sonic curves.

[0013] Optionally, the original exploration data stream is resampled and normalized at a sampling interval of 2 milliseconds and a depth interval of 5 meters to form the standardized initial data volume.

[0014] Optionally, the specific steps for performing waveform similarity propagation of seismic attributes in the lateral direction are as follows:

[0015] S301: Traverse each transverse seismic profile node in the bidirectional cascaded topology skeleton and read the seismic wave time series corresponding to the transverse seismic profile node;

[0016] S302: Calculate the peak value of the waveform cross-correlation function between adjacent seismic profile nodes to determine the waveform propagation time delay;

[0017] S303: Based on the time delay and the preset waveform attenuation factor, the amplitude values ​​of adjacent seismic profile nodes are weighted and smoothed to eliminate lateral splicing traces.

[0018] S304: Compare the amplitude values ​​of adjacent seismic profile nodes after processing with the original amplitude envelope, retain waveform segments with consistent amplitude change trends, and remove distorted waveforms caused by noise interference.

[0019] S305: Stitch together all seismic profile data after lateral waveform similarity propagation processing to output a seismic attribute dataset with enhanced lateral continuity.

[0020] Optionally, the specific steps for performing the recursive derivation of rock physical parameters of well logging attributes in the vertical direction are as follows:

[0021] S311: Along each longitudinal logging curve node in the bidirectional cascaded topology skeleton, read the sonic transit time and density logging values ​​corresponding to the longitudinal logging curve node;

[0022] S312: Based on a pre-established porosity conversion lookup table, convert the sonic transit time and density logging values ​​into equivalent formation porosity values;

[0023] S313: Based on the formation porosity value and combined with the preset empirical relationship of saturation, the corresponding formation oil and gas saturation value is recursively calculated.

[0024] S314: Combine the formation porosity value with the formation hydrocarbon saturation value into a multidimensional rock physics vector and map it onto the corresponding formation depth coordinates;

[0025] S315: Concatenate the multidimensional rock physics vectors on all formation depth coordinates to output a vertical rock physics property continuity logging correction dataset.

[0026] Optionally, according to Perform formation porosity numerical conversion, where: This represents the formation porosity value. Indicates the regional lithology correction factor. This indicates the acoustic transit time logging value read. Indicates the acoustic transit time of the regional rock skeleton. The acoustic transit time of formation fluids This indicates the density logging value read. Indicates the density of the regional rock skeleton. This indicates the density of the formation fluid.

[0027] Optionally, step S4 specifically includes:

[0028] S401: Obtain the raw gravity exploration data and perform latitude correction, height correction, intermediate layer correction, topographic correction and normal field correction in sequence to obtain Bouguer gravity anomaly values ​​that reflect the differences in subsurface density.

[0029] S402: Perform gridded interpolation of the Bouguer gravity anomaly values ​​in a planar coordinate system to draw a Bouguer gravity anomaly planar map, which uses color or contour lines to represent the spatial distribution of gravity anomaly intensity.

[0030] S403: Obtain raw magnetic exploration data and perform diurnal variation correction and normal field correction to obtain magnetic anomaly values ​​that reflect the differences in the magnetic properties of underground rocks;

[0031] S404: Perform gridded interpolation of the magnetic anomaly values ​​in a planar coordinate system, calculate the gradient vector of the magnetic anomaly in the horizontal direction at each grid point, and generate a magnetic anomaly vector map containing the magnetic anomaly size and horizontal direction information at each grid point.

[0032] S405: Read the Bouguer gravity anomaly plan view and the magnetic anomaly vector map, and perform a Fourier transform;

[0033] S406: Extract low-frequency tectonic trend components and perform inverse Fourier transform to generate the geological tectonic background field.

[0034] Optionally, step S4 further includes:

[0035] S407: Calculate the tectonic curvature value of each grid point in the geological tectonic background field, and generate weighting coefficients based on the tectonic curvature value;

[0036] S408: Spatial registration is performed based on the coordinates of the preliminary predicted stratigraphic boundaries on the plane extracted from the intermediate integrated dataset and the weighting coefficients.

[0037] Optionally, according to Calculate the curvature value ,in:

[0038] set up The geological tectonic background field obtained by inverse Fourier transform is shown in spatial coordinates. The electric field strength at that location, Indicates the geological tectonic background field in First-order partial derivative in the direction, Indicates the geological tectonic background field in First-order partial derivative in the direction, Indicates the geological tectonic background field in Second-order partial derivatives in the direction, Indicates the geological tectonic background field in Second-order partial derivatives in the direction, Indicates the geological tectonic background field in and Mixed second-order partial derivatives in the direction.

[0039] Based on the above method, this invention also provides a rapid oil and gas exploration data integration system based on bidirectional cascading, used to realize the rapid oil and gas exploration data integration method based on bidirectional cascading described above. Its key features include a data input and preprocessing layer, a core processing engine layer, a multi-source data constraint and correction layer, a closed-loop verification and output layer, and an advanced application module layer, wherein:

[0040] The data input and preprocessing layer is equipped with a data standardization module, which is used to acquire the raw exploration data stream and normalize it according to the stratigraphic depth slice to form a standardized initial data volume.

[0041] The core processing engine layer includes a feature extraction module, a bidirectional cascaded topology skeleton, a cascaded data processing module, and a data fusion module. The feature extraction module extracts laterally distributed seismic reflection interface feature points and vertically distributed well logging lithological inflection points based on the standardized initial data volume. The bidirectional cascaded topology skeleton defines the point-to-point constraint relationship between seismic profiles and well logging curves. The cascaded data processing module performs waveform similarity propagation of seismic attributes laterally and recursively calculates rock physical parameters of well logging attributes vertically. The data fusion module generates an intermediate integrated dataset that fuses lateral waveform features and vertical rock physical attributes.

[0042] The multi-source data constraint and correction layer is equipped with a gravity and magnetic data processing module and a spatial registration and weight superposition module. The gravity and magnetic data processing module is used to acquire gravity and magnetic exploration data and generate a geological structural background field. The spatial registration and weight superposition module is used to perform spatial registration and weight superposition on the intermediate integrated dataset to obtain boundary correction results.

[0043] The closed-loop verification and output layer is equipped with a skeleton reverse update module. The skeleton reverse update module backtracks and updates the point-to-point constraint relationship in the bidirectional cascaded topology skeleton according to the boundary correction result, and generates high-precision integrated oil and gas exploration data after closed-loop verification.

[0044] The advanced application module layer is used to implement fluid identification marking and dynamic feedback for well vibration calibration.

[0045] Based on the above description of the method and system, the significant advantages of the present invention are:

[0046] It can quickly achieve bidirectional cascaded integration of raw seismic data streams and raw well logging data streams. Through real-time feedback and updates, it can effectively correct stratigraphic boundary drift, eliminate lateral data splicing traces, enhance the lateral continuity of seismic survey data, and improve the accuracy of the data integration model.

[0047] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0048] The accompanying drawings of this invention are described below.

[0049] Figure 1 A flowchart of the rapid integration method for oil and gas exploration data based on bidirectional cascading provided by the present invention;

[0050] Figure 2 A flowchart illustrating the process of propagating the similarity of lateral seismic attribute waveforms;

[0051] Figure 3 A flowchart for recursively extrapolating the rock physical parameters of vertical logging attributes;

[0052] Figure 4 A flowchart of the multi-source data constraint and correction process;

[0053] Figure 5 The system architecture diagram for rapid integration of oil and gas exploration data based on bidirectional cascading provided by this invention;

[0054] Figure 6 This is an internal architecture diagram of the hierarchical structure of the system in a specific embodiment;

[0055] Figure 7 A chart analyzing the proportion of low-confidence data in slices at various depths;

[0056] Figure 8 This is a waveform similarity analysis diagram of multiple seismic profiles. Detailed Implementation

[0057] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0058] Example 1:

[0059] like Figure 1 The method for rapid integration of oil and gas exploration data based on bidirectional cascading, as shown, includes the following steps:

[0060] S1: Obtain the raw exploration data stream from the geophysical acquisition end. This raw exploration data stream includes the raw seismic data stream and the raw well logging data stream. The raw seismic data stream includes seismic wave travel time and amplitude envelope, while the raw well logging data stream includes well logging sonic curves. The acquired raw data streams then need to be normalized according to a unified formation depth slice to eliminate differences in scale and dimensions between data from different sources, thereby forming a standardized initial data volume.

[0061] S2: Based on the standardized initial data volume, extract the horizontally distributed seismic reflection interface feature points and the vertically distributed well logging lithology inflection points. Use these feature points to construct a two-way cascaded topological skeleton that can simultaneously run through the seismic data volume and the well logging data volume. The core function of this skeleton is to clearly define the point-to-point constraint relationship between the seismic profile nodes and the well logging curve nodes.

[0062] S3: Apply the constructed bidirectional cascaded topology skeleton to the standardized initial data volume, perform waveform similarity propagation of seismic attributes in the horizontal direction, perform rock physical parameter recursion of well logging attributes in the vertical direction, and generate an intermediate integrated dataset that integrates horizontal waveform features and vertical rock physical attributes.

[0063] S4: To improve the geological reliability of the integrated data, it is necessary to introduce the geological structural background field obtained by converting gravity and magnetic exploration data. This background field is used to perform spatial registration and weight superposition on the intermediate integrated dataset, thereby correcting the spatial drift of the stratigraphic boundary caused by the resolution limitations of a single data source such as seismic or well logging.

[0064] S5: Based on the stratigraphic boundary correction results obtained after spatial registration, backtrack to the initially constructed bidirectional cascaded topological skeleton, and iteratively update the point-to-point constraint relationships defined within it. Through this closed-loop verification process, a more accurate integrated oil and gas exploration data is finally generated.

[0065] The following example, using the data integration and processing of a 3D seismic survey area in a basin in western China and a key exploration well within it, illustrates the implementation method of waveform similarity propagation. In the specific implementation, the raw exploration data stream acquired from the geophysical acquisition end includes a 3D seismic data volume covering a full coverage area of ​​300 square kilometers and a well logging curve from a deep exploration well designated Well-A. During implementation, the raw exploration data stream is resampled and normalized according to a uniform 2-millisecond sampling interval and 5-meter depth interval to form a standardized initial data volume. Based on the standardized initial data volume, on two adjacent seismic profiles, Inline500 and Inline501, a strong wave peak is picked at 2000 milliseconds on the time axis as a seismic reflection interface feature point, and at the corresponding depth of Well-A well, an anomalous point of acoustic time difference is picked as a well logging lithology inflection point. This constructs a bidirectional cascaded topological skeleton segment connecting the two seismic profile nodes and one well logging curve node. Figure 2 It can be seen that:

[0066] The specific steps for performing waveform similarity propagation of seismic attributes in the horizontal direction are as follows:

[0067] S301: Traverse each transverse seismic profile node in the bidirectional cascaded topology skeleton and read the seismic wave time series corresponding to the transverse seismic profile node;

[0068] S302: Calculate the peak value of the waveform cross-correlation function between adjacent seismic profile nodes to determine the waveform propagation time delay;

[0069] S303: Based on the time delay and the preset waveform attenuation factor, the amplitude values ​​of adjacent seismic profile nodes are weighted and smoothed to eliminate lateral splicing traces.

[0070] S304: Compare the amplitude values ​​of adjacent seismic profile nodes after processing with the original amplitude envelope, retain waveform segments with consistent amplitude change trends, and remove distorted waveforms caused by noise interference.

[0071] S305: Stitch together all seismic profile data after lateral waveform similarity propagation processing to output a seismic attribute dataset with enhanced lateral continuity.

[0072] This example uses the Inline500 profile node as an example. It reads the seismic waveform time series from the Inline500 profile node within a time window of 1990 to 2010 milliseconds, using this series as the reference waveform. Seismic waveform time series within the same time window range are read from the adjacent Inline501 profile node, and this series is used as the comparison waveform. The preset sliding step size is 1 sampling point. Using 1 sampling point as the sliding step size, the position of the comparison waveform relative to the reference waveform is cyclically moved. At each sliding position, the cross-correlation coefficient between the reference waveform and the compared waveform after the sliding is calculated at the corresponding sampling point amplitude value. After traversing all sliding positions, all calculated cross-correlation coefficients are used to form the waveform cross-correlation function, and the peak value with the largest value in the waveform cross-correlation function is identified. The sliding step size corresponding to the peak value is recorded as 3 sampling points, the seismic data sampling time interval is 2 milliseconds, and the waveform propagation delay is calculated as 6 milliseconds. The cross-correlation coefficient calculation formula corresponding to the peak value of the waveform cross-correlation function adopts a standardized cross-correlation coefficient calculation form, and the calculation formula is:

[0073]

[0074] Where: characters Indicates the sliding step size Cross-correlation number at time, character Indicates the reference waveform at the 1st The amplitude value of each sampling point, character Indicates the comparison waveform at the 1st The amplitude value of each sampling point, character Represents the mean of the reference waveform amplitude sequence, character This indicates a comparison of the mean of waveform amplitude sequences; the character... This represents the total number of sampling points within the time window. This formula can be understood as quantifying the similarity between two waveform sequences at different relative positions.

[0075] In some embodiments, the amplitude values ​​of the Inline500 and Inline501 profile nodes in the overlapping region are weighted and smoothed based on the calculated 6-millisecond delay and a preset waveform attenuation factor of 0.8. The purpose of the weighted smoothing is to eliminate abrupt changes in lateral amplitude caused by acquisition footprints or processing traces. The amplitude value of the Inline501 profile node after weighted smoothing is compared with the amplitude envelope extracted from the original data. Waveform segments whose amplitude variation trend is consistent with the original amplitude envelope (i.e., synchronously enhanced within the range of 1995 to 2005 milliseconds) are retained, while distorted waveforms with a sharp amplitude jump around 2010 milliseconds that does not conform to the background envelope trend are removed. After completing the above lateral waveform similarity propagation processing for all seismic profile nodes, the data of the Inline500, Inline501, and all subsequent processed profile nodes are stitched together according to the survey line order to output a seismic attribute dataset with enhanced lateral continuity.

[0076] The bidirectional cascaded topology framework, centered on the integration needs of oil and gas exploration data, primarily comprises four modules: a seismic profile node module, a well logging curve node module, a point-to-point constraint relationship module, and a closed-loop verification module. Specifically: the seismic profile node module stores the stratigraphic interpretation coordinates and waveform characteristics of the seismic data; the well logging curve node module stores the depth-corrected physical property parameters of the well logging data; the point-to-point constraint relationship module initially establishes rigid connections based on the uncorrected stratigraphic boundary coordinates, including Euclidean distance parameters between nodes; and the closed-loop verification module compares the boundary coordinates before and after correction and triggers constraint updates. The model training steps are: node initialization, establishing initial rigid constraints, external field correction, and reverse backtracking update. The preset tolerance range of the elastic constraints in the parameter settings is based on the interpretation of small faults and the maximum expected displacement of deflection in the regional micro-geological structure, ensuring that the model adapts to the fine-tuning needs of the actual scenario.

[0077] In the specific implementation of this invention, the waveform similarity propagation of lateral seismic attributes and the recursion of rock physical parameters of vertical logging attributes are not performed independently, but are deeply integrated through a two-way cascaded topological framework. This two-way cascaded topological framework defines the point-to-point constraint relationship between seismic profile nodes and logging curve nodes, forming a rigid channel for lateral and vertical information exchange. During data processing, the waveform similarity of adjacent profiles on which lateral waveform propagation is based is constrained by the vertical recursion results of rock physical parameters at the corresponding depth, such as porosity and saturation. For example, in a section where logging shows high porosity, the smoothing weight of lateral waveform propagation can be adjusted accordingly to better protect the amplitude anomaly characteristics of the reservoir. Conversely, the rationality and trend of the lateral variation of the acoustic transit time and density logging values ​​on which the vertical rock physical parameter recursion depends are also verified and guided by the continuity of lateral seismic waveforms, such as the lateral consistency revealed by the time delay determined by the peak value of the cross-correlation function. For example, in areas with excellent transverse waveform continuity, vertically derived rock physical parameters, such as drastic abrupt changes in saturation, will be considered abnormal and need to be verified in conjunction with the geological tectonic background. Transverse waveform characteristics and vertical physical parameters are coupled and iterated through a two-way cascaded topological framework to achieve mutual calibration and constraint.

[0078] pass Figure 3 As can be seen, the specific steps for recursively extrapolating rock physical parameters of well logging attributes in the vertical direction are as follows:

[0079] S311: Along each longitudinal logging curve node in the bidirectional cascaded topology skeleton, read the sonic transit time and density logging values ​​corresponding to the longitudinal logging curve node;

[0080] S312: Based on a pre-established porosity conversion lookup table, convert the sonic transit time and density logging values ​​into equivalent formation porosity values;

[0081] S313: Based on the formation porosity value and combined with the preset empirical relationship of saturation, the corresponding formation oil and gas saturation value is recursively calculated.

[0082] S314: Combine the formation porosity value with the formation hydrocarbon saturation value into a multidimensional rock physics vector and map it onto the corresponding formation depth coordinates;

[0083] S315: Concatenate the multidimensional rock physics vectors on all formation depth coordinates to output a vertical rock physics property continuity logging correction dataset.

[0084] The following section uses exploration data processing in a deep-water block in the South China Sea as an example to illustrate the implementation method of vertical recursive petrophysical parameter extrapolation. In practice, a two-way cascaded topology framework has been established in the target area. This framework includes a vertical connection path that runs through seismic profiles and multiple exploration wells. The vertical logging curve node of a key appraisal well, Well-BH-01, is selected as the operational object. Specifically, along the vertical logging curve node corresponding to Well-BH-01 in the two-way cascaded topology framework, the sonic transit time logging sequence and density logging sequence for that node, with a sampling interval of 0.125 meters, are read from the standardized initial data volume within the depth range of 3500 meters to 3600 meters.

[0085] In some embodiments, based on a porosity conversion lookup table pre-established through laboratory core measurements and statistical analysis, the sonic transit time logging values ​​and density logging values ​​at each depth point are converted into equivalent formation porosity values. The porosity conversion process is performed according to a variant of the Willy time-averaging formula, and the calculation formula is as follows:

[0086]

[0087] Where: characters This represents the calculated formation porosity value, represented by the character. Indicates the regional lithology correction factor, character This indicates the read acoustic transit time logging value, character. Characters representing the acoustic transit time of the regional rock skeleton Characters representing the acoustic transit time of formation fluids. This indicates the density logging value read, character. Indicates the density of the regional rock skeleton, character This indicates the density of the formation fluid.

[0088] Regional lithology correction factor This was determined through laboratory core measurements and statistical analysis. For different exploration areas, core samples of typical lithologies from that area were collected, and the sonic transit time of the samples was measured. ,density Formation fluid acoustic transit time Formation fluid density and the actual porosity of the sample Actual measurement; substitute the measured values ​​into the porosity conversion formula, and then reverse the process to calculate the porosity using the formula. Consistent with actual measurements Values; statistical analysis of multiple samples The values ​​are averaged or weighted averaged according to lithological classification to obtain the values ​​for the region. It is used for subsequent formation porosity numerical conversion.

[0089] In some embodiments, based on the calculated formation porosity value of 0.182, and combined with the Archie formula saturation empirical relation established beforehand for the sandstone reservoir in this deep-water block, the formation hydrocarbon saturation value at the corresponding depth point of 3520.5 meters is recursively calculated. It can be understood that the saturation empirical relation correlates porosity, resistivity, and saturation. At the depth point of 3520.5 meters, the input deep lateral resistivity logging value is 12.5 ohm-meters, and the input formation water resistivity is 0.05 ohm-meters. Using the Archie formula, the formation hydrocarbon saturation value is recursively calculated to be 0.65. In a specific implementation, the formation porosity value of 0.182 and the formation hydrocarbon saturation value of 0.65 are combined to form a multidimensional rock physics vector [0.182, 0.65] containing both porosity and saturation components, and this multidimensional rock physics vector is accurately mapped to the depth coordinate of 3520.5 meters. For each sampling point of the vertical logging curve node in Well-BH-01 within the depth range of 3500m to 3600m, repeat the above steps of reading, transforming, recursively calculating, combining, and mapping. Optionally, perform the exact same vertical recursive process for all other logging curve nodes involved in the bidirectional cascaded topology skeleton within the work area. Concatenate all multidimensional rock physics vectors on the formation depth coordinates in ascending order of depth to output a vertically continuous logging correction dataset of rock physics properties.

[0090] pass Figure 4 As can be seen, step S4 specifically includes:

[0091] S401: Obtain raw gravity exploration data from the geophysical acquisition end. The raw gravity exploration data consists of gravity acceleration observations from a series of measurement points. Latitude correction, altitude correction, intermediate layer correction, topography correction, and normal field correction are performed on the raw gravity exploration data in sequence to eliminate the influence of latitude, altitude, surrounding topography, and the theoretical gravity field of the Earth ellipsoid at the measurement points, and to obtain Bouguer gravity anomaly values ​​that reflect the differences in underground density.

[0092] S402: Perform grid interpolation on the processed Bouguer gravity anomaly values ​​in a plane coordinate system to draw a Bouguer gravity anomaly plane map that represents the spatial distribution of gravity anomaly intensity with color or contour lines.

[0093] S403: The original magnetic exploration data consists of the total intensity or component observations of the geomagnetic field at a series of measurement points. The original magnetic exploration data is corrected for diurnal variation and normal field to eliminate the influence of diurnal variation of the geomagnetic field and normal background field, and magnetic anomaly values ​​reflecting the differences in the magnetic properties of underground rocks are obtained.

[0094] S404: The magnetic anomaly values ​​are interpolated in a planar coordinate system and the gradient vector of the magnetic anomaly in the horizontal direction at each grid point is calculated. Finally, a magnetic anomaly vector diagram containing the magnetic anomaly size and horizontal direction information of each grid point is generated.

[0095] S405: Read the Bouguer gravity anomaly plan view and the magnetic anomaly vector map, and perform a Fourier transform;

[0096] S406: Extract low-frequency tectonic trend components and perform inverse Fourier transform to generate the geological tectonic background field;

[0097] In practice, the generated Bouguer gravity anomaly planar map and magnetic anomaly vector map are read, and the two-dimensional grid data representing the physical field distribution in these two maps are used as input. Fourier transform is used to transform the Bouguer gravity anomaly planar map and magnetic anomaly vector map from the spatial domain to the frequency domain. The Fourier transform converts the image or field distribution in the spatial domain into a series of frequency components, where low-frequency components correspond to large-scale, macroscopic regional tectonic trends, while high-frequency components correspond to small-scale, local geological details or noise. In practice, a low-pass filter is set in the frequency domain. This filter allows spectral components below a set cutoff frequency to pass through, while truncating or attenuating spectral components above the cutoff frequency, thereby extracting the low-frequency tectonic trend components reflecting the regional tectonic background. An inverse Fourier transform is applied to the extracted low-frequency tectonic trend components to transform them back from the frequency domain to the spatial domain, generating a macroscopic geological tectonic background field that retains macroscopic regional tectonic information while filtering out local details and noise.

[0098] S407: Calculate the tectonic curvature value of each grid point in the geological tectonic background field, and generate weighting coefficients based on the tectonic curvature value;

[0099] S408: Spatial registration is performed based on the coordinates of the preliminary predicted stratigraphic boundaries on the plane extracted from the intermediate integrated dataset and the weighting coefficients.

[0100] In some embodiments, the tectonic curvature value of each grid point in the geological tectonic background field is calculated, and the tectonic curvature value is used as a weighting coefficient. Based on the calculated first and second-order partial derivative values, the tectonic curvature value of each grid point in the geological tectonic background field is calculated by substituting them into a preset curvature calculation formula. Calculate the curvature value The formula is as follows:

[0101]

[0102] Where: characters This represents the geological tectonic background field obtained after inverse Fourier transform in spatial coordinates. Field strength value at the location, character Indicates the geological tectonic background field in First-order partial derivative of direction, character Indicates the geological tectonic background field in First-order partial derivative of direction, character Indicates the geological tectonic background field in Second-order partial derivatives of direction, character Indicates the geological tectonic background field in Second-order partial derivatives of direction, character Indicates the geological tectonic background field in and Mixed second-order partial derivatives in the direction.

[0103] The calculated construction curvature value As a weighting coefficient The formula for calculating the weighting coefficient is:

[0104]

[0105] Where: function It is a monotonically decreasing mapping function, whose function is to convert larger construction curvature values... Mapped to smaller weight coefficients To construct a smaller curvature value Mapped to larger weighting coefficients This results in a greater weighting of regions with gentle slopes.

[0106] In practice, the set of coordinate positions of the preliminary predicted stratigraphic boundaries on the plane is extracted from the intermediate integrated dataset. This set of coordinate positions contains a series of two-dimensional coordinate points. In practice, the coordinates of each stratigraphic boundary point will be... Construction curvature weighting coefficient Multiplying the coordinates by their weights is geometrically equivalent to applying an adjustment that "attracts" the coordinates towards the higher-weighted regions. The adjusted coordinates align with the higher-weighted regions overall, forcing the stratigraphic boundary to fit the position with the minimum tectonic curvature, thus completing spatial registration. Optionally, spatial registration can be iteratively performed using a global optimization algorithm, with the objective function defined as finding the optimal weighted fit between the adjusted stratigraphic boundary coordinate set and the tectonic curvature distribution in the geological background field. Alternatively, the specific implementation of multiplying the weighting coefficients by the coordinates can employ linear superposition or an iterative optimization method based on energy minimization.

[0107] Combination Figure 5 and Figure 6As can be seen, based on the above method, this embodiment also provides a rapid integration system for oil and gas exploration data based on bidirectional cascading, including a data input and preprocessing layer, a core processing engine layer, a multi-source data constraint and correction layer, a closed-loop verification and output layer, and an advanced application module layer, wherein:

[0108] The data input and preprocessing layer is equipped with a data standardization module, which is used to acquire the raw exploration data stream and normalize it according to the stratigraphic depth slice to form a standardized initial data volume.

[0109] The core processing engine layer includes a feature extraction module, a bidirectional cascaded topology skeleton, a cascaded data processing module, and a data fusion module. The feature extraction module extracts laterally distributed seismic reflection interface feature points and vertically distributed well logging lithological inflection points based on the standardized initial data volume. The bidirectional cascaded topology skeleton defines the point-to-point constraint relationship between seismic profiles and well logging curves. The cascaded data processing module performs waveform similarity propagation of seismic attributes laterally and recursively calculates rock physical parameters of well logging attributes vertically. The data fusion module generates an intermediate integrated dataset that fuses lateral waveform features and vertical rock physical attributes.

[0110] The multi-source data constraint and correction layer is equipped with a gravity and magnetic data processing module and a spatial registration and weight superposition module. The gravity and magnetic data processing module is used to acquire gravity and magnetic exploration data and generate a geological structural background field. The spatial registration and weight superposition module is used to perform spatial registration and weight superposition on the intermediate integrated dataset to obtain boundary correction results.

[0111] The closed-loop verification and output layer is equipped with a skeleton reverse update module. The skeleton reverse update module backtracks and updates the point-to-point constraint relationship in the bidirectional cascaded topology skeleton based on the boundary correction result, generating high-precision integrated oil and gas exploration data after closed-loop verification. The generation of high-precision integrated oil and gas exploration data after closed-loop verification is based on the formation boundary drift result corrected in step S4, and is achieved by backtracking and updating the point-to-point constraint relationship in the bidirectional cascaded topology skeleton. Specifically, by comparing the stratigraphic boundary coordinates before correction with the spatially registered and corrected stratigraphic boundary coordinates, the displacement vectors of each boundary point in the lateral and longitudinal directions are calculated. All nodes in the two-way cascaded topology framework associated with the drifting stratigraphic boundary are traversed, including the seismic profile nodes defining the boundary and well logging curve nodes located near the boundary, and the rigid connections between these nodes based on the coordinates before correction are released. Based on the corrected new stratigraphic boundary coordinates, the Euclidean shortest spatial distance between each associated seismic profile node and well logging curve node is recalculated, and an elastic point-to-point constraint relationship is re-established according to this distance. This relationship allows the connection between nodes to be fine-tuned within a preset tolerance range as subsequent micro-geological structure interpretations are performed. The reconstructed constraint relationship parameters, including the new spatial distance and elastic tolerance range, are written into the data structure or configuration file of the two-way cascaded topology framework, completing a closed-loop verification process from data integration, external field correction to internal framework update, ultimately generating high-precision integrated oil and gas exploration data.

[0112] The advanced application module layer is used to implement fluid identification marking and dynamic feedback for well vibration calibration.

[0113] In practical implementation, based on the corrected stratigraphic boundary drift results, the point-to-point constraint relationships in the bidirectional cascaded topology skeleton are backtracked and updated. This includes the following steps: comparing the stratigraphic boundary coordinates before correction with the spatially registered corrected stratigraphic boundary coordinates; and calculating the displacement vectors of each boundary point in the horizontal (x-direction) and vertical (y-direction) directions. In practice, all nodes in the bidirectional cascaded topology framework associated with the drifting stratigraphic boundary are traversed. These nodes include seismic profile nodes that define the boundary and well logging curve nodes located near the boundary. The original rigid point-to-point connections between these nodes, based on the coordinates before correction, are released. In practice, the Euclidean shortest spatial distance between each associated seismic profile node and well logging curve node is recalculated based on the corrected, new stratigraphic boundary coordinates. In practice, the elastic point-to-point constraint relationship between the seismic profile node and well logging curve node is re-established according to the recalculated shortest spatial distance. This elastic point-to-point constraint relationship allows the connection between nodes to be fine-tuned within a certain preset tolerance range as subsequent micro-geological structure interpretations are performed, rather than a strictly fixed distance connection. In practice, the reconstructed elastic point-to-point constraint relationship parameters, which include the new spatial distance and elastic tolerance range, are written into the data structure or configuration file of the bidirectional cascaded topology framework, thus completing a closed-loop verification process from data integration, external field correction to internal framework update. Understandably, the closed-loop verification mechanism enables the integration process to self-correct the core cascaded topological framework based on the external correction result of the geological structural background field, thereby improving the rationality of the model. Optionally, the tolerance range of the flexible point-to-point constraint relationship can be quantitatively preset based on factors such as the complexity of the regional geological structure and the degree of certainty of existing knowledge.

[0114] In one embodiment of the present invention, taking a data integration model of a tight sandstone gas reservoir exploration area in the Ordos Basin as an example, the implementation method of fluid identification and labeling and data reliability evaluation for high-precision oil and gas exploration integration data is described. In specific implementation, the high-precision oil and gas exploration integration data has been generated through a two-way cascade integration process, and the model includes vertically continuously distributed rock physical parameters. In specific implementation, the recursive results of the vertically continuous rock physical parameters in the high-precision oil and gas exploration integration data are retrieved, and the formation oil and gas saturation values ​​at each sampling point within the target layer (depth 2800 meters to 3000 meters) are extracted. The extracted formation oil and gas saturation values ​​range from 0.1 to 0.8.

[0115] In practice, after extracting the formation hydrocarbon saturation values, a reliability evaluation of these values ​​is necessary. All logging curve nodes used in constructing the recursive results of the aforementioned rock physical parameters are retrieved, and the quality level of the original logging data corresponding to each node is verified. The quality level is comprehensively evaluated based on the signal-to-noise ratio, frequency jump, and depth matching error of the logging curve, and is divided into four levels: A, B, C, and D. If the logging data quality level corresponding to a certain logging curve node is C or D, i.e., lower than the preset B level standard, then all formation hydrocarbon saturation values ​​calculated from this logging curve node are marked as low-confidence data. The number of data points marked as low-confidence in the high-precision integrated oil and gas exploration data on a 2850-meter depth slice is counted. The proportion of low-confidence data points to the total number of data points in that depth slice is calculated to obtain the low-confidence data percentage. When the proportion of low-confidence data on the 2850-meter depth slice reaches 35%, exceeding the preset 30% warning threshold, all fluid identification and labeling operations on the 2850-meter depth slice are suspended. This step is understood to control the impact of low-quality input data on the final interpretation conclusions. Only formation hydrocarbon saturation values ​​from Class A and Class B quality logging curve nodes that were not marked as low-confidence are used in subsequent fluid identification and labeling procedures.

[0116] In some embodiments, fluid identification and labeling are performed based on the formation hydrocarbon saturation values ​​after confidence evaluation screening. A classification threshold of 0.5 is set for hydrocarbon saturation; formation intervals with hydrocarbon saturation values ​​higher than 0.5 are marked as suspected gas-bearing areas, while those with values ​​lower than 0.5 are marked as water-bearing or dry layers. In a specific implementation, a seismic attribute dataset, processed by lateral waveform similarity propagation, is read from high-precision integrated oil and gas exploration data. This dataset contains amplitude attribute volumes. On a 2855-meter depth slice of the amplitude attribute volume, with an amplitude background value set to 10000, local anomalies with absolute amplitude values ​​greater than 15000 are screened. The spatial coordinates of these local anomalies are represented in the form of grid points, for example, (205, 88). Optionally, spatial clustering is performed on the screened local anomalies, grouping adjacent anomaly grid points into one anomaly unit. The spatial coordinate set of the local anomaly units is then intersected with the spatial coordinate set of the previously marked suspected gas-bearing areas. Spatial coordinates of points that simultaneously satisfy a formation hydrocarbon saturation value greater than 0.5 and are located within the amplitude anomaly range are retained. In some embodiments, the block coordinates retained after the intersection operation include (205, 88), (205, 89), and (206, 88). These retained blocks are assigned a specific fluid identification code, for example, the numeric code "2" represents "gas layer". In specific implementations, the fluid identification code is embedded as a new attribute channel into the high-precision integrated oil and gas exploration data, stored alongside existing porosity, saturation, and other attributes. The formula for calculating the saturation-weighted confidence average is:

[0117]

[0118] Where: characters This represents the average oil and gas saturation after considering credibility weights. Indicates the first The original oil and gas saturation values ​​for each data point, character Indicates the first The credibility weight coefficient for each data point (Level A is assigned 1.0, Level B is assigned 0.7, Level C is assigned 0.3, and Level D is assigned 0), character This represents the total number of data points involved in the calculation.

[0119] In one embodiment of the present invention, taking a data integration model of a tight sandstone gas reservoir exploration area in the Ordos Basin as an example, the implementation method of fluid identification and labeling and data reliability evaluation for high-precision oil and gas exploration integration data is described. In specific implementation, the high-precision oil and gas exploration integration data has been generated through a two-way cascade integration process, and the model includes vertically continuously distributed rock physical parameters. In specific implementation, the recursive results of the vertically continuous rock physical parameters in the high-precision oil and gas exploration integration data are retrieved, and the formation oil and gas saturation values ​​at each sampling point within the target layer (depth 2800 meters to 3000 meters) are extracted. The extracted formation oil and gas saturation values ​​range from 0.1 to 0.8.

[0120] In practice, after extracting the formation hydrocarbon saturation values, a reliability evaluation of these values ​​is necessary. All logging curve nodes used in constructing the recursive results of the aforementioned rock physical parameters are retrieved, and the quality level of the original logging data corresponding to each node is verified. The quality level is comprehensively evaluated based on the signal-to-noise ratio, frequency jump, and depth matching error of the logging curve, and is divided into four levels: A, B, C, and D. If the logging data quality level corresponding to a certain logging curve node is C or D, i.e., lower than the preset B level standard, then all formation hydrocarbon saturation values ​​calculated from this logging curve node are marked as low-confidence data. The number of data points marked as low-confidence in the high-precision integrated oil and gas exploration data on a 2850-meter depth slice is counted. The proportion of low-confidence data points to the total number of data points in that depth slice is calculated to obtain the low-confidence data percentage. When the proportion of low-confidence data on the 2850-meter depth slice reaches 35%, exceeding the preset 30% warning threshold, all fluid identification and labeling operations on the 2850-meter depth slice are suspended. This step is understood to control the impact of low-quality input data on the final interpretation conclusions. Only formation hydrocarbon saturation values ​​from Class A and Class B quality logging curve nodes that were not marked as low-confidence are used in subsequent fluid identification and labeling procedures.

[0121] In some embodiments, fluid identification and labeling are performed based on the formation hydrocarbon saturation values ​​selected through confidence evaluation. A grading threshold of 0.5 is set for hydrocarbon saturation; formation intervals with hydrocarbon saturation values ​​higher than 0.5 are marked as suspected gas-bearing areas, while those with values ​​lower than 0.5 are marked as water-bearing or dry layers. In a specific implementation, a seismic attribute dataset, processed by lateral waveform similarity propagation, is read from high-precision integrated oil and gas exploration data. This dataset contains amplitude attribute volumes. On a 2855-meter depth slice of the amplitude attribute volume, an amplitude background value of 10000 is set, and local anomalies with absolute amplitude values ​​greater than 15000 are selected. Optionally, spatial clustering is performed on the selected local anomalies, grouping adjacent anomaly grid points into one anomaly unit. The spatial coordinate set of the local anomaly units is intersected with the spatial coordinate set of the previously marked suspected gas-bearing areas. Spatial coordinate points that simultaneously satisfy a formation hydrocarbon saturation value greater than 0.5 and are located within the amplitude anomaly range are retained. These preserved blocks are assigned a specific fluid identification code, for example, the numerical code "2" represents "gas layer". In practice, the fluid identification code is embedded as a new attribute channel into the high-precision oil and gas exploration integrated data, stored alongside existing attributes such as porosity and saturation. The confidence-weighted saturation value is calculated using the following formula:

[0122]

[0123] Where: characters This represents the overall saturation value after credibility weighting, in characters. Indicates the first The weight coefficients of each data point involved in the calculation and evaluated for credibility, characters Indicates the first The original oil and gas saturation values ​​for each data point, character This represents the total number of valid data points involved in the calculation. Refer to Table 1, which defines the rules for assigning reliability weight coefficients to different well logging data quality levels.

[0124] Weighting coefficient The selection criteria are determined based on the quality grade of the original logging data corresponding to the nodes of the logging curves used when constructing the recursive results of rock physical parameters. The quality grade is comprehensively evaluated based on the signal-to-noise ratio, frequency jump, and depth matching error of the logging curves, and is divided into four levels: A, B, C, and D. Level A corresponds to high signal-to-noise ratio, no frequency jump, and accurate depth matching. A value of 1.0 is assigned; Class B corresponds to a moderate signal-to-noise ratio, slight cycle skipping, and good depth matching. A value of 0.7 is assigned; Grade C corresponds to a low signal-to-noise ratio, cycle skipping, and depth errors. The value is assigned to 0.3; Class D corresponds to a very low signal-to-noise ratio, severe cycle skipping, and poor depth matching. The value is assigned to 0.0. The preset credibility weight coefficient corresponding to the quality level is directly used to calculate the credibility weighted saturation value.

[0125] Table 1: Correspondence between Well Logging Data Quality Grade and Reliability Weighting Coefficient

[0126]

[0127] See Figure 7 This is a chart analyzing the proportion of low-confidence data at different depth slices, showing the percentage of low-confidence data across different depth slices in the oil and gas exploration data integration process. It is used to assess data quality and guide subsequent fluid identification work. The warning threshold represents a preset 30% warning threshold; when the proportion exceeds this value, fluid identification operations for that depth slice must be suspended. The key anomaly is at the 2850-meter depth slice, where the low-confidence data proportion reaches 35%, significantly higher than the warning threshold; therefore, the fluid identification process for this slice is suspended. From 2820 meters to 2850 meters, the proportion of low-confidence data steadily increases from approximately 12% to 35%, peaking at 2850 meters. From 2850 meters to 2870 meters, the proportion gradually decreases from 35% to approximately 22%, remaining within a manageable range. By visualizing the proportion of low-confidence data, depth ranges with poor data quality can be quickly identified, preventing low-quality data from affecting the final interpretation conclusions.

[0128] In one embodiment of the present invention, taking a block in an oilfield in eastern China that has entered the development stage as an example, the implementation method of dynamic feedback adjustment based on well-seismic calibration results is described. In a specific implementation, a production well that has encountered industrial oil flow and has complete core and testing data is selected as the calibration well, which is numbered Well-DY-1. In a specific implementation, the measured reservoir thickness and physical property parameters of the calibration well Well-DY-1 in the main oil-bearing section (depth 2350 m to 2370 m) are extracted. The measured reservoir thickness is 15.2 m, and the measured average porosity is 0.18. From the high-precision oil and gas exploration integrated data that has completed preliminary modeling, the predicted reservoir thickness and predicted physical property parameters corresponding to the Well-DY-1 well location are extracted. The model predicts a reservoir thickness of 16.8 m and a model predicts an average porosity of 0.21.

[0129] In some embodiments, the percentage error between the measured reservoir thickness of 15.2 meters and the predicted reservoir thickness of 16.8 meters is calculated, with the percentage error being (16.8-15.2) / 15.2 ≈ 10.5%. Simultaneously, the deviation between the measured average porosity of 0.18 and the predicted average porosity of 0.21 is calculated, with the deviation being 0.21-0.18 = 0.03. It can be understood that the percentage error and the deviation measure the difference between the model prediction and the downhole measurement results. The calculated percentage error of 10.5% and the deviation of 0.03 are used as input variables and fed into the parameter adaptive adjustment function built into the bidirectional cascaded topology framework. The parameter adaptive adjustment function will adjust the weighting factors of waveform similarity propagation and the empirical coefficients of recursive rock physical parameters based on these inputs.

[0130] In practical implementation, the empirical coefficients for recursively estimating the weighting factors of waveform similarity propagation and rock physical parameters include specific steps. A two-dimensional lookup table is constructed, indexed by error percentage and deviation. In the lookup table, error percentage is used as the x-axis and deviation as the y-axis. Each cell defines a target weighting factor interval that matches the range of the error percentage and deviation combination. The target weighting factor interval matching the current input error percentage of 10.5% and deviation of 0.03 is retrieved from the lookup table; the retrieved interval is [0.50, 0.55]. A linear interpolation algorithm is used to calculate a transition weighting factor between the current waveform similarity propagation weighting factor of 0.60 and the lower limit of the target weighting factor interval of 0.50. The formula for calculating the transition weighting factor is:

[0131]

[0132] Where: characters This represents the calculated transition weight factor, character. This indicates the waveform similarity propagation weight factor currently in use, character This represents the lower limit of the target weight factor interval retrieved from the two-dimensional lookup table, character. This represents an adjustment step size factor between 0 and 1, used to control the magnitude of the adjustment.

[0133] Propagation weight factor The selection rules combine initial presets with a dynamic feedback adjustment mechanism. Initial It can be preset according to the complexity of regional geological structures and the requirements for waveform similarity propagation. For example, in lateral waveform similarity propagation, it can be set in conjunction with a preset waveform attenuation factor. Subsequent updates are made through dynamic feedback from well-seismic calibration results. Specifically, production wells with encountered industrial oil flows and complete coring data are selected as calibration wells. The measured reservoir thickness and physical properties of the main oil-bearing sections of the calibration wells are extracted and compared with the predicted parameters of the corresponding well locations in the high-precision integrated oil and gas exploration data. The error percentage and deviation are then calculated. The error percentage and deviation are input into a built-in adaptive parameter adjustment function within the bidirectional cascaded topology framework. This function is based on a two-dimensional lookup table indexed by both parameters, where the error percentage is on the x-axis and the deviation is on the y-axis. Each cell defines the target weight factor interval for the matching combination. After retrieving the target interval corresponding to the current error percentage and deviation, a linear interpolation algorithm is used to calculate the transition weight factor. The formula is equal minus Multiply by the parentheses reduce ,in The adjustment step size factor is between 0 and 1. The lower limit of the target interval, after updating That is, new This enables dynamic adjustment of the propagation weight factor.

[0134] See Figure 8 This is a waveform similarity analysis chart of multiple seismic profiles, showing the amplitude response of five adjacent seismic profiles within a 0-2 second timeframe. It is a typical visualization of the lateral waveform similarity propagation process in oil and gas exploration. The waveforms of each profile are highly consistent in overall trend, especially after 1.0 second, where the peaks and valleys are almost perfectly aligned, indicating that they reflect similar subsurface geological reflection interfaces. There is a noticeable vertical offset in the amplitude baseline of each profile; this is a visual adjustment to clearly display multiple curves in the same image and does not represent actual amplitude differences. Minor high-frequency jitter is superimposed on the waveforms, a common noise in seismic data, requiring smoothing and filtering in subsequent processing. By calculating the cross-correlation function between adjacent profiles, the time delay of waveform propagation can be accurately determined, a crucial step in eliminating lateral splicing traces and enhancing the lateral continuity of seismic data. Waveform similarity and consistency can also be used to evaluate the acquisition quality and processing effectiveness of seismic data.

[0135] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for fast integration of oil and gas exploration data based on bidirectional cascading, characterized in that, Includes the following steps: S1: Acquire the raw exploration data stream and normalize it according to the stratigraphic depth slices to form a standardized initial data volume; S2: Based on the standardized initial data volume, extract the horizontally distributed seismic reflection interface feature points and the vertically distributed well logging lithology inflection points to construct a two-way cascaded topological skeleton. The two-way cascaded topological skeleton defines the point-to-point constraint relationship between the seismic profile and the well logging curve. S3: Apply the bidirectional cascaded topology skeleton to the standardized initial data volume, perform waveform similarity propagation of seismic attributes in the horizontal direction, perform rock physical parameter recursion of well logging attributes in the vertical direction, and generate an intermediate integrated dataset that integrates horizontal waveform features and vertical rock physical attributes. S4: Acquire gravity and magnetic exploration data and generate a geological structure background field. Perform spatial registration and weight superposition on the intermediate integrated dataset to correct the stratigraphic boundary drift caused by the resolution limitations of a single data source such as earthquake or well logging. S5: Based on the corrected stratigraphic boundary drift results, backtrack and update the point-to-point constraint relationship in the bidirectional cascaded topology skeleton to generate a high-precision integrated oil and gas exploration data model after closed-loop verification. Step S4 specifically includes: S401: Obtain the raw gravity exploration data and perform latitude correction, height correction, intermediate layer correction, topographic correction and normal field correction in sequence to obtain Bouguer gravity anomaly values ​​that reflect the differences in subsurface density. S402: Perform gridded interpolation of the Bouguer gravity anomaly values ​​in a planar coordinate system to draw a Bouguer gravity anomaly planar map, which uses color or contour lines to represent the spatial distribution of gravity anomaly intensity. S403: Obtain raw magnetic exploration data and perform diurnal variation correction and normal field correction to obtain magnetic anomaly values ​​that reflect the differences in the magnetic properties of underground rocks; S404: Perform gridded interpolation of the magnetic anomaly values ​​in a planar coordinate system, calculate the gradient vector of the magnetic anomaly in the horizontal direction at each grid point, and generate a magnetic anomaly vector map containing the magnitude and horizontal direction information of the magnetic anomaly at each grid point. S405: Read the Bouguer gravity anomaly plan view and the magnetic anomaly vector map, and perform a Fourier transform; S406: Extract low-frequency tectonic trend components and perform inverse Fourier transform to generate the geological tectonic background field; S407: Calculate the tectonic curvature value of each grid point in the geological tectonic background field, and generate weighting coefficients based on the tectonic curvature value; S408: Spatial registration is performed based on the coordinates of the preliminary predicted stratigraphic boundaries on the plane extracted from the intermediate integrated dataset and the weighting coefficients; and according to calculating a curvature value wherein: set up The geological tectonic background field obtained by inverse Fourier transform is shown in spatial coordinates. The electric field strength at that location, Indicates the geological tectonic background field in First-order partial derivative in the direction, Indicates the geological tectonic background field in First-order partial derivative in the direction, Indicates the geological tectonic background field in Second-order partial derivatives in the direction, Indicates the geological tectonic background field in Second-order partial derivatives in the direction, Indicates the geological tectonic background field in and Mixed second-order partial derivatives in the direction.

2. The rapid integration method for oil and gas exploration data based on bidirectional cascading as described in claim 1, characterized in that, The raw exploration data stream includes a raw seismic data stream and a raw well logging data stream. The raw seismic data stream includes seismic wave travel time and amplitude envelope, and the raw well logging data stream includes well logging sonic curves.

3. The rapid integration method for oil and gas exploration data based on bidirectional cascading as described in claim 2, characterized in that, The original exploration data stream is resampled and normalized at a sampling interval of 2 milliseconds and a depth interval of 5 meters to form the standardized initial data volume.

4. The rapid integration method for oil and gas exploration data based on bidirectional cascading as described in claim 1, 2, or 3, characterized in that, The specific steps for performing waveform similarity propagation of seismic attributes in the horizontal direction are as follows: S301: Traverse each transverse seismic profile node in the bidirectional cascaded topology skeleton and read the seismic wave time series corresponding to the transverse seismic profile node; S302: Calculate the peak value of the waveform cross-correlation function between adjacent seismic profile nodes to determine the waveform propagation time delay; S303: Based on the time delay and the preset waveform attenuation factor, the amplitude values ​​of adjacent seismic profile nodes are weighted and smoothed to eliminate lateral splicing traces. S304: Compare the amplitude values ​​of adjacent seismic profile nodes after processing with the original amplitude envelope, retain waveform segments with consistent amplitude change trends, and remove distorted waveforms caused by noise interference. S305: Stitch together all seismic profile data after lateral waveform similarity propagation processing to output a seismic attribute dataset with enhanced lateral continuity.

5. The rapid integration method for oil and gas exploration data based on bidirectional cascading as described in claim 4, characterized in that, The specific steps for recursively extrapolating the rock physical parameters of well logging attributes in the vertical direction are as follows: S311: Along each longitudinal logging curve node in the bidirectional cascaded topology skeleton, read the sonic transit time and density logging values ​​corresponding to the longitudinal logging curve node; S312: Based on a pre-established porosity conversion lookup table, convert the sonic transit time and density logging values ​​into equivalent formation porosity values; S313: Based on the formation porosity value and combined with the preset empirical relationship of saturation, the corresponding formation oil and gas saturation value is recursively calculated. S314: Combine the formation porosity value with the formation hydrocarbon saturation value into a multidimensional rock physics vector and map it onto the corresponding formation depth coordinates. S315: Concatenate the multidimensional rock physics vectors on all formation depth coordinates to output a vertical rock physics property continuity logging correction dataset.

6. The rapid integration method for oil and gas exploration data based on bidirectional cascading as described in claim 5, characterized in that, according to Perform formation porosity numerical conversion, where: This represents the formation porosity value. Indicates the regional lithology correction factor. This indicates the acoustic transit time logging value read. Indicates the acoustic transit time of the regional rock skeleton. The acoustic transit time of formation fluids This indicates the density logging value read. Indicates the density of the regional rock skeleton. This indicates the density of the formation fluid.

7. A rapid integration system for oil and gas exploration data based on bidirectional cascading, used to implement the rapid integration method for oil and gas exploration data based on bidirectional cascading as described in any one of claims 1-6, characterized in that, It includes a data input and preprocessing layer, a core processing engine layer, a multi-source data constraint and correction layer, a closed-loop verification and output layer, and an advanced application module layer, among which: The data input and preprocessing layer is equipped with a data standardization module, which is used to acquire the raw exploration data stream and normalize it according to the stratigraphic depth slice to form a standardized initial data volume. The core processing engine layer includes a feature extraction module, a bidirectional cascaded topology skeleton, a cascaded data processing module, and a data fusion module. The feature extraction module extracts laterally distributed seismic reflection interface feature points and vertically distributed well logging lithological inflection points based on the standardized initial data volume. The bidirectional cascaded topology skeleton defines the point-to-point constraint relationship between seismic profiles and well logging curves. The cascaded data processing module performs waveform similarity propagation of seismic attributes laterally and recursively calculates rock physical parameters of well logging attributes vertically. The data fusion module generates an intermediate integrated dataset that fuses lateral waveform features and vertical rock physical attributes. The multi-source data constraint and correction layer is equipped with a gravity and magnetic data processing module and a spatial registration and weight superposition module. The gravity and magnetic data processing module is used to acquire gravity and magnetic exploration data and generate a geological structural background field. The spatial registration and weight superposition module is used to perform spatial registration and weight superposition on the intermediate integrated dataset to obtain boundary correction results. The closed-loop verification and output layer is equipped with a skeleton reverse update module. The skeleton reverse update module backtracks and updates the point-to-point constraint relationship in the bidirectional cascaded topology skeleton according to the boundary correction result, and generates a high-precision integrated oil and gas exploration data model after closed-loop verification. The advanced application module layer is used to implement fluid identification marking and dynamic feedback for well vibration calibration.

Images