Underground rock stratum identification method, system and equipment based on three-dimensional seismic data

Abstract

The invention discloses an underground rock stratum identification method, system and equipment based on three-dimensional seismic data, and relates to the technical field of rock stratum identification. The method comprises the following steps: acquiring three-dimensional seismic data of a target area by using seismic exploration equipment; preprocessing the data to obtain standard three-dimensional seismic data; backstepping underground rock attribute information based on the propagation characteristic data by using a three-dimensional seismic inversion technology; performing multi-attribute analysis on the rock attribute information to obtain rock stratum identification features; and building a rock stratum feature recognition model, optimizing recognition features through the model, determining a recognition result, and carrying out three-dimensional modeling based on the recognition result to generate a rock stratum three-dimensional model. The technical problems that a traditional rock stratum recognition method is insufficient in complex geological structure analysis capacity and low in inversion result precision are solved, and the technical effects that the rock stratum space distribution characteristics are described with high precision and the complex geological structure recognition reliability is improved are achieved by fusing the three-dimensional seismic inversion technology and the multi-attribute intelligent analysis technology.

Description

Technical Field

[0001] The present invention relates to the technical field of rock formation identification, and in particular to an underground rock formation identification method, system and equipment based on three-dimensional seismic data. Background Art

[0002] With the increasing demand for resource exploration and underground engineering, accurate identification of underground rock formations and their structures has become an important task in geological exploration. Traditional rock formation identification methods usually rely on two-dimensional seismic data or local area data acquisition, which cannot fully reveal the full picture of complex underground structures and have low accuracy, making it difficult to meet the exploration needs in large-scale or complex geological environments.

[0003] At present, three-dimensional seismic data has become the main means of obtaining underground rock information in geological exploration. It can provide higher-resolution seismic wave propagation data and help identify the distribution characteristics and hierarchical structure of different underground rock layers. However, how to extract effective information from massive three-dimensional seismic data, accurately invert the physical properties of underground rocks, and analyze complex rock layer structures is still a challenging technical problem. Existing technical solutions mostly rely on manual or traditional inversion methods, which have problems such as insufficient recognition accuracy, low efficiency, and poor adaptability to complex geological environments. Summary of the invention

[0004] The present application provides a method, system and equipment for identifying underground rock formations based on three-dimensional seismic data, which solves the technical problems of insufficient analysis capability of traditional rock formation identification methods for complex geological structures and low accuracy of inversion results.

[0005] In a first aspect of the present application, a method for identifying underground rock formations based on three-dimensional seismic data is provided, the method comprising:

[0006] Seismic exploration equipment is used to collect and acquire three-dimensional seismic data of a target area, wherein the three-dimensional seismic data includes propagation characteristic data of seismic waves in underground media; the three-dimensional seismic data is preprocessed to obtain standard three-dimensional seismic data; three-dimensional seismic inversion technology is used to infer underground rock attribute information from the standard three-dimensional seismic data based on the propagation characteristic data; multi-attribute analysis is performed on the underground rock attribute information to obtain underground rock formation identification characteristics, wherein the underground rock formation identification characteristics include rock formation distribution characteristics and rock formation hierarchy; a rock formation characteristic identification model is constructed, and the underground rock formation identification characteristics are analyzed and optimized through the rock formation characteristic identification model to determine the underground rock formation characteristic identification results; three-dimensional modeling is performed based on the underground rock formation characteristic identification results to generate a three-dimensional model of the underground rock formation.

[0007] A second aspect of the present application provides an underground rock formation identification system based on three-dimensional seismic data, the system comprising:

[0008] Data acquisition module: using seismic exploration equipment to acquire three-dimensional seismic data of the target area, the three-dimensional seismic data contains propagation characteristic data of seismic waves in underground media; preprocessing module: preprocessing the three-dimensional seismic data to obtain standard three-dimensional seismic data; data inversion module: applying three-dimensional seismic inversion technology to infer underground rock attribute information from the standard three-dimensional seismic data based on the propagation characteristic data; multi-attribute analysis module: performing multi-attribute analysis on the underground rock attribute information to obtain underground rock formation identification characteristics, the underground rock formation identification characteristics include rock formation distribution characteristics and rock formation hierarchy; three-dimensional modeling module: building a rock formation feature identification model, analyzing and optimizing the underground rock formation identification characteristics through the rock formation feature identification model, determining the underground rock formation feature identification results, and performing three-dimensional modeling based on the underground rock formation feature identification results to generate a three-dimensional model of the underground rock formation.

[0009] According to a third aspect of the present application, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0010] Seismic exploration equipment is used to collect and acquire three-dimensional seismic data of a target area, wherein the three-dimensional seismic data includes propagation characteristic data of seismic waves in underground media; the three-dimensional seismic data is preprocessed to obtain standard three-dimensional seismic data; three-dimensional seismic inversion technology is used to infer underground rock attribute information from the standard three-dimensional seismic data based on the propagation characteristic data; multi-attribute analysis is performed on the underground rock attribute information to obtain underground rock formation identification characteristics, wherein the underground rock formation identification characteristics include rock formation distribution characteristics and rock formation hierarchy; a rock formation characteristic identification model is constructed, and the underground rock formation identification characteristics are analyzed and optimized through the rock formation characteristic identification model to determine the underground rock formation characteristic identification results; three-dimensional modeling is performed based on the underground rock formation characteristic identification results to generate a three-dimensional model of the underground rock formation.

[0011] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0012] First, the three-dimensional seismic data of the target area is collected and acquired by using seismic exploration equipment, and the three-dimensional seismic data contains the propagation characteristic data of seismic waves in the underground medium; then, the three-dimensional seismic data is preprocessed to obtain standard three-dimensional seismic data; further, the three-dimensional seismic inversion technology is used to infer the underground rock attribute information from the standard three-dimensional seismic data based on the propagation characteristic data, and then the underground rock attribute information is analyzed by multi-attribute analysis to obtain the underground rock layer identification characteristics, and the underground rock layer identification characteristics include rock layer distribution characteristics and rock layer hierarchy structure; finally, a rock layer feature identification model is built, and the underground rock layer identification characteristics are analyzed and optimized by the rock layer feature identification model to determine the underground rock layer feature identification results, and three-dimensional modeling is performed based on the underground rock layer feature identification results to generate a three-dimensional model of the underground rock layer. The technical problems of insufficient analysis ability of traditional rock layer identification methods for complex geological structures and low accuracy of inversion results are solved, and the technical effect of high-precision characterization of rock layer spatial distribution characteristics and improved reliability of complex geological structure identification is achieved by integrating three-dimensional seismic inversion and multi-attribute intelligent analysis technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0014] Figure 1 A schematic diagram of a method for identifying underground rock formations based on three-dimensional seismic data provided in an embodiment of the present application;

[0015] Figure 2 A schematic diagram of the structure of an underground rock formation identification system based on three-dimensional seismic data provided in an embodiment of the present application.

[0016] Explanation of the reference numerals: data acquisition module 11 , preprocessing module 12 , data inversion module 13 , multi-attribute analysis module 14 , three-dimensional modeling module 15 . DETAILED DESCRIPTION

[0017] The present application solves the technical problems of insufficient analysis capability of traditional rock formation identification methods for complex geological structures and low accuracy of inversion results by providing an underground rock formation identification method, system and equipment based on three-dimensional seismic data.

[0018] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0019] It should be noted that the terms "including" and "having" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or are inherent to these processes, methods, products or devices.

[0020] Embodiment 1, as Figure 1 As shown, the present application provides a method for identifying underground rock formations based on three-dimensional seismic data, wherein the method comprises:

[0021] The three-dimensional seismic data of the target area is acquired by using seismic exploration equipment, and the three-dimensional seismic data includes propagation characteristic data of seismic waves in underground media.

[0022] In one embodiment, the three-dimensional seismic data of the target area is collected using seismic exploration equipment, the core of which is to artificially excite seismic waves and record their propagation characteristics in underground rock formations. Specifically, the seismic exploration equipment arranges seismic sources and receivers on the surface or shallow layers. The seismic sources generate seismic waves. When the seismic waves encounter different rock formations during underground propagation, reflection, refraction and attenuation will occur. The receiver records the arrival time, amplitude, frequency and other characteristics of these seismic waves to form three-dimensional seismic data. These data contain the propagation characteristics of seismic waves in underground media, such as propagation speed, propagation path, reflection and refraction characteristics, and other information, which can reflect the physical properties and structural characteristics of underground rock formations. By analyzing these propagation characteristic data, the distribution, thickness, lithology and other information of underground rock formations can be inferred, providing basic data support for subsequent rock formation identification and modeling.

[0023] The three-dimensional seismic data is preprocessed to obtain standard three-dimensional seismic data.

[0024] In one embodiment, after obtaining the 3D seismic data of the target area, a series of preprocessing operations are performed on the 3D seismic data to ensure the quality and accuracy of the data. These processing steps include removing noise, correcting irregular signals, filling missing data, etc., so that it meets the requirements of analysis and inversion. Through these preprocessing steps, interference and errors can be eliminated, and clear, stable standard 3D seismic data can be obtained, providing a reliable data basis for subsequent analysis and modeling.

[0025] Furthermore, the obtaining of standard three-dimensional seismic data includes:

[0026] According to the application requirements of seismic data, a filter is selected to filter the three-dimensional seismic data to obtain usable three-dimensional seismic data; a signal regularization processing flow is obtained, and the signal regularization processing flow includes denoising processing, detrending processing and normalization; based on the signal regularization processing flow, the usable three-dimensional seismic data is de-signaled and regularized to obtain regularized three-dimensional seismic data; linear interpolation is used to fill the missing data of the regularized three-dimensional seismic data to obtain the standard three-dimensional seismic data.

[0027] Preferably, according to the application requirements of seismic data, the original three-dimensional seismic data is first subjected to spectrum analysis to determine the frequency range of noise (such as high-frequency noise or low-frequency interference). If the noise is mainly high-frequency components, a low-pass filter is selected to retain the low-frequency effective signal; if the noise is mainly low-frequency components, a high-pass filter is selected to retain the high-frequency effective signal; if the noise is concentrated in a specific frequency band, a band-stop filter is selected to remove the noise in the frequency band. Subsequently, according to the frequency range of the noise, the cutoff frequency or bandwidth of the filter is set, and the filter is applied to the three-dimensional seismic data to remove the noise and retain the effective signal. By comparing the data spectrum before and after filtering, it is ensured that the noise is effectively removed and the effective signal is not destroyed, and then the filtered data is saved to obtain the available three-dimensional seismic data as the basis for subsequent processing. Afterwards, a signal regularization process is executed, including denoising (removing interference signals in the data), detrending (eliminating long-term change trends in the data) and normalization (adjusting the data to a uniform scale range) to improve the stability and consistency of the data. For denoising, select a suitable denoising algorithm according to the characteristics of the data, such as wavelet transform, median filtering or adaptive filtering, and then set the denoising parameters. For example, select the wavelet basis function and the number of decomposition layers in the wavelet transform, and then perform denoising on the filtered 3D seismic data trace by trace to remove random noise and outliers. For detrending, identify the linear or nonlinear trend in the data through polynomial fitting or sliding average method, and then subtract the fitted trend term from the data to make the data more stable. For normalization, count the maximum and minimum values ​​of each data item, and scale the data of each data point to a uniform range through the maximum and minimum method to ensure that all data have consistent dimensions for subsequent analysis. After completing the signal regularization processing described above, regularized three-dimensional seismic data is obtained. In this regularized three-dimensional seismic data, there may be some missing or incomplete data points. These missing data are filled by using the linear interpolation method, wherein linear interpolation is to infer the value of the missing part through the numerical value of the existing data points, for example, by taking the difference through the mean of adjacent data points, so as to complete the regularized three-dimensional seismic data and obtain the final standard three-dimensional seismic data, providing high-quality data support for subsequent analysis and modeling.

[0028] The three-dimensional seismic inversion technology is used to infer the underground rock property information from the standard three-dimensional seismic data based on the propagation characteristic data.

[0029] In one embodiment, the application of 3D seismic inversion technology is to infer the physical property information of underground rocks by analyzing the propagation characteristics (such as wave velocity, attenuation and other parameters) in standard 3D seismic data. Specifically, firstly, a preliminary underground rock property model is constructed by using the processed standard 3D seismic data and combining the propagation law of seismic waves in the underground medium, and then the model is optimized by the inversion algorithm so that the calculation results of the model are more matched with the actual collected seismic data. After multiple iterations of optimization, accurate underground rock property information is finally obtained, including the density, wave velocity, porosity, etc. of the rock layer, providing accurate data for subsequent underground structure analysis.

[0030] Furthermore, the application of the 3D seismic inversion technology to infer the underground rock property information from the standard 3D seismic data based on the propagation characteristic data includes:

[0031] The three-dimensional seismic inversion technology is applied to perform inversion modeling based on the standard three-dimensional seismic data to establish an initial underground rock property model; the inversion algorithm is used to iteratively optimize the initial underground rock property model based on the propagation characteristic data to obtain a target underground rock property model; the target underground rock property model is used to perform inversion calculations on the standard three-dimensional seismic data to obtain the underground rock property information.

[0032] Preferably, according to standard three-dimensional seismic data and business needs, the rock attribute parameters that need to be inverted, such as wave impedance, density, longitudinal wave velocity (Vp), shear wave velocity (Vs), etc., are clearly defined, and then combined with geological data, logging data and other geophysical data to obtain the rock attribute range (such as the upper and lower limits of velocity and density) of the target area. Subsequently, the target area is divided into three-dimensional grids, each grid unit represents a small underground volume, and an initial rock attribute value (such as uniform distribution or interpolation result based on logging data) is assigned to each grid unit according to the determined rock attribute range, and then the gridded rock attribute parameters are integrated into an initial underground rock attribute model as the starting point of the inversion. Afterwards, the propagation characteristic data collected at the historical moment is used as the input of the inversion, and based on the initial underground rock attribute model, the seismic wave propagation equation (such as the acoustic wave equation or the elastic wave equation) is used for forward simulation to generate simulated seismic data. The forward simulation needs to be consistent with the observation system of the actual seismic data, including the source location, the receiver location and the time sampling rate. Then the simulated seismic data is compared with the historical propagation characteristic data, and the error between the two (such as residual, mean square error or correlation coefficient) is calculated. Further, an inversion algorithm (such as the least squares method, the conjugate gradient method or the global optimization algorithm) is used to adjust the rock attribute parameters of the initial model according to the error information. For example, in the least squares method, the model parameters are updated to reduce the error by solving the gradient direction of the error function. Repeat the process of forward simulation, error calculation and model parameter optimization to gradually reduce the difference between the simulated data and the actual data. When the error information meets the error threshold or reaches the upper limit of the number of iterations, the iteration is stopped and the optimized initial underground rock attribute model is saved as the target underground rock attribute model. Then, the optimized target underground rock attribute model and the standard three-dimensional seismic data are read, and the standard three-dimensional seismic data is inverted based on the target underground rock attribute model to extract more detailed rock attribute information, such as seismic record traces, wave impedance, density, velocity, porosity, etc. These attribute information will help the system terminal better understand the characteristics of the underground structure and provide basic data for subsequent rock formation identification and geological interpretation.

[0033] Furthermore, obtaining the target underground rock property model includes:

[0034] Based on the propagation characteristic data, the initial underground rock property model is forward simulated to obtain simulated seismic data; the simulated seismic data is error calculated to obtain seismic data error information; based on the inversion algorithm, the initial underground rock property model is iteratively optimized through the seismic data error information to obtain the target underground rock property model.

[0035] Optionally, the initial underground rock attribute model and historical propagation characteristic data (historical propagation characteristic data in the historical database with absolute errors within the tolerance range) are loaded, the forward simulation parameters (such as acoustic wave equation, source type, receiver position) are set, and the finite difference method is used for forward simulation to generate simulated seismic data, such as seismic record traces, wave impedance, etc. Subsequently, the simulated seismic data are aligned with the historical propagation characteristic data, the residual (the difference between the simulated data and the actual data) is calculated for each trace, and the mean square error (MSE) is calculated based on the residual as the error function. After that, the least square method is selected as the inversion algorithm, and the model parameters are iteratively optimized with the goal of minimizing the mean square error, that is, the rock attribute parameters of the initial model are adjusted according to the error information, and the forward simulation is re-performed and the new error is calculated until the mean square error is less than the preset threshold or the number of iterations reaches the upper limit. Finally, the optimized target underground rock attribute model is output. This process inverts the underground rock attribute information with high precision through forward simulation, error calculation and iterative optimization, providing a reliable basis for rock formation identification and geological interpretation.

[0036] The underground rock attribute information is subjected to multi-attribute analysis to obtain underground rock strata identification features, wherein the underground rock strata identification features include rock strata distribution features and rock strata hierarchical structure.

[0037] In one embodiment, after three-dimensional seismic inversion technology and model optimization, physical property information of underground rocks, such as wave velocity, density, porosity, etc. in different depth areas, is obtained. Subsequently, by clustering analysis of these underground rock property information, multiple physical property information is combined to obtain underground rock formation identification features, which are detailed descriptions of underground rock formations, including rock formation distribution characteristics and rock formation hierarchy. Among them, rock formation distribution characteristics can reveal information such as the boundaries, positions, and thicknesses of each rock formation, and help determine the spatial distribution of each underground rock formation; rock formation hierarchy can reveal the continuity of rock formations, interlayer transitions, and interface structures between rock formations, which helps to understand the interaction and relative positions between rock formations, and further reveals the complexity of underground structures. In summary, through the multi-attribute analysis of underground rock property information, key identification features of underground rock formations can be extracted, providing important geological data support for subsequent rock formation modeling, resource exploration, etc.

[0038] Furthermore, the obtaining of underground rock formation identification features includes:

[0039] Performing cluster analysis on the attribute data of the underground rock attribute information to obtain underground rock attribute clustering results; performing multi-attribute analysis based on the underground rock attribute clustering results to determine rock stratum distribution characteristics and rock stratum hierarchy; and composing the underground rock stratum identification characteristics based on the rock stratum distribution characteristics and rock stratum hierarchy.

[0040] Preferably, after obtaining the underground rock attribute information, multiple attribute data are combined into a multi-dimensional feature vector. For example, each data point can be represented as [Z, ρ, V p ][Z,ρ,Vp], where Z is the wave impedance, ρ is the density, and Vp is the longitudinal wave velocity. Subsequently, the sum of squared clustering errors (SSE) under different K values ​​is calculated, and the relationship curve between K value and SSE is plotted. The K value corresponding to the inflection point of the curve is selected as the optimal number of clusters, and K data points are randomly selected as the initial cluster centers. The distance between each data point and the cluster center is calculated using the Euclidean distance, and each data point is assigned to the cluster center with the nearest distance. The mean of all data points in each cluster is calculated as the new cluster center. Repeat the above process until the cluster center no longer changes or the change is less than the preset threshold, thereby obtaining the clustering results of underground rock properties. Afterwards, use 3D visualization tools (such as volume rendering or slice display) to display the clustering results, understand the spatial distribution of different rock categories, and divide the target area into different rock distribution areas; use edge detection algorithms (such as Sobel operator or Canny operator) to identify rock layer boundaries; count the thickness of each rock layer vertically (in the depth direction) to generate a rock layer thickness distribution map; obtain rock layer distribution characteristics by summarizing rock spatial distribution, rock layer boundaries, and rock layer thickness. At the same time, analyze the changes in rock attribute values ​​(such as wave impedance and velocity) vertically, identify rock layer interfaces, and mark the interfaces of different rock layers according to the mutation points of attribute values; divide the target area into different rock layer levels according to the vertical stratification results, and clarify the upper and lower relationships of each rock layer, such as mudstone, sandstone, and limestone from top to bottom; analyze the lateral continuity of rock layers, and identify geological structures such as faults and cracks; obtain the rock layer hierarchy by summarizing vertical stratification, upper and lower relationships, and rock layer continuity. Finally, the rock stratum distribution characteristics and rock stratum hierarchy are integrated into underground rock stratum identification features, which provides comprehensive data support for underground structure analysis and lays the foundation for the subsequent construction of underground rock stratum three-dimensional models.

[0041] A rock formation feature recognition model is constructed, and the underground rock formation recognition features are analyzed and optimized through the rock formation feature recognition model to determine the underground rock formation feature recognition results, and three-dimensional modeling is performed based on the underground rock formation feature recognition results to generate a three-dimensional model of the underground rock formation.

[0042] In one embodiment, the process of building a rock formation feature recognition model first starts with data preparation, loading underground rock formation recognition feature data obtained through multi-attribute analysis, including rock formation distribution characteristics (such as spatial distribution, thickness variation, boundary morphology) and rock formation hierarchy (such as vertical stratification, up-down relationship, continuity), and collecting historical underground rock formation feature data sets as training data. Subsequently, a suitable deep learning network model is selected, and the historical underground rock formation feature data sets are used as training data for model training. After the model training is completed, the performance is further improved by hyperparameter tuning and model integration, thereby obtaining a rock formation feature recognition model. Afterwards, the underground rock formation recognition feature data to be analyzed is input into the trained rock formation feature recognition model, the required features (such as wave impedance, density, velocity, thickness, hierarchical relationship, etc.) are extracted, the model prediction is executed and the recognition results are optimized, and the final underground rock formation feature recognition results are generated. Based on the recognition results, the topological relationship between the rock formations is analyzed, and the three-dimensional modeling software (such as Petrel, GOCAD) is imported to generate a three-dimensional model of the underground rock formation, providing important support for geological research and resource development.

[0043] Furthermore, the building of the rock formation feature recognition model includes:

[0044] Collect historical underground rock formation characteristic data sets, use deep learning network to train the historical underground rock formation characteristic data sets to obtain an initial feature recognition model; verify the initial feature recognition model to obtain model performance parameters; perform back propagation optimization on the initial feature recognition model based on the model performance parameters to build the rock formation characteristic recognition model.

[0045] Preferably, when building a rock formation feature recognition model, it is first necessary to collect a historical underground rock formation feature data set, which includes historical underground rock formation recognition features (historical rock formation distribution features, historical rock formation hierarchy) and historical underground rock formation features (rock formation categories, rock formation properties, and rock formation structures in each area). Subsequently, the historical underground rock formation feature data set is divided into a training set, a validation set, and a test set, and a suitable deep learning network is selected, such as a fully connected neural network (FCN) to build an initial feature recognition model structure, including an input layer, a hidden layer, and an output layer. The training set data is then input into the deep learning network, the output result of the network is calculated, and compared with the historical underground rock formation characteristics, the loss value is calculated through the cross entropy loss, the gradient is calculated according to the loss function, the network parameters (such as weights and biases) are updated, and the forward propagation and back propagation processes are repeated until the loss function converges or reaches a preset number of iterations, thereby obtaining a trained initial feature recognition model. After that, the initial feature recognition model is used to predict the validation set data to obtain the predicted underground rock formation feature recognition results. The performance parameters of the model, such as accuracy and recall, are calculated as model performance parameters, and then the deficiencies of the model (such as overfitting and underfitting) are analyzed based on these model performance parameters. If a certain model performance parameter does not meet the expected requirements, the network structure and hyperparameters will be adjusted, such as increasing or decreasing the number of hidden layers, adjusting the number of neurons, adjusting the learning rate, batch size, regularization parameters, etc., to optimize the model performance. Then, the model is retrained using the adjusted network structure and hyperparameters, and the optimized model is verified using the validation set data to calculate new performance parameters. This process is repeated until the model performance parameters meet the expected requirements, thereby building a rock formation feature recognition model for the identification and analysis of underground rock formation features, ensuring that underground rock formation features can be efficiently identified, and providing important support for geological interpretation and resource exploration.

[0046] Furthermore, generating the three-dimensional model of underground rock formations includes:

[0047] According to the geological application objectives, the modeling boundary range and modeling element information are determined; based on the modeling boundary range and modeling element information, the underground rock formation feature identification results are subjected to topological relationship analysis to obtain rock formation modeling feature data; based on the rock formation modeling feature data, three-dimensional space conversion is performed to generate the underground rock formation three-dimensional model.

[0048] Optionally, according to the geological application goals predetermined at the beginning, such as oil and gas exploration, mineral resource assessment or geological disaster prediction, combined with the geological background and research needs of the target area, the spatial scope of modeling (such as horizontal range and vertical depth) is determined. For example, the horizontal range can be the longitude and latitude range of a certain block, and the vertical range can be the strata from the surface to a certain depth. Then, according to the geological application goals, the element information required for modeling is extracted, such as rock layer interface, rock layer thickness, faults and cracks, rock layer properties, etc. Subsequently, the underground rock layer feature recognition results obtained by the rock layer feature recognition model are read, including the rock layer type, thickness, hierarchical structure and other information of each area, and then the underground rock layer feature recognition results are clipped according to the modeling boundary range to retain the data in the target area. By analyzing the contact relationship between different rock formations (such as conformable contact and unconformable contact), the interface relationship of the rock formations is obtained; by analyzing the cutting relationship between faults and rock formations, the strike, dip and fault throw of the faults are determined, thereby obtaining the relationship between faults and fractures; by analyzing the lateral and vertical continuity of rock formations, missing or repeated formations are identified, and the continuity relationship of rock formations is obtained; then, the multiple topological relationships obtained by the analysis are integrated to obtain rock formation modeling feature data, such as the spatial position and attributes of faults and fractures, rock formation interface coordinates, rock formation thickness distribution, rock formation topological relationships, etc., and then interpolation methods (such as Kriging interpolation and inverse distance weighted interpolation) are used to convert the rock formation modeling feature data into continuous three-dimensional spatial data. Then, a three-dimensional grid model is constructed according to the modeling boundary range and resolution. Each grid unit contains information such as rock layer type, thickness, and attributes. The three-dimensional interface of the rock layer is generated according to the rock layer interface coordinates and topological relationships in the converted three-dimensional spatial data. Faults and cracks are added according to their spatial positions and attributes to obtain the final three-dimensional model of the underground rock layer. The model is saved in a common format (such as STL, OBJ, GRDECL) to facilitate the use of three-dimensional visualization tools (such as Petrel, GOCAD, Paraview) to display the model, and to perform operations such as rotation, scaling, and slicing, providing important support for geological research and resource development.

[0049] Furthermore, the method comprises:

[0050] The three-dimensional model of the underground rock formation is verified and evaluated using actual geological exploration data to obtain model reliability parameters, and the three-dimensional model of the underground rock formation is optimized and corrected using the model reliability parameters.

[0051] Optionally, the process of using actual geological exploration data to verify and evaluate the three-dimensional model of underground rock formations first requires loading actual geological exploration data, including logging data, drilling data, seismic data, etc., which provide real information about underground rock formations. Subsequently, the actual data is compared with the three-dimensional model of underground rock formations, and the two are compared and analyzed at the same position and depth through spatial alignment and data matching. After that, the error between the model and the actual data is calculated. Commonly used error indicators include mean square error (MSE), mean absolute error (MAE) and correlation coefficient (R2), which are used to quantify the accuracy and reliability of the model. For example, the mean square error can reflect the deviation between the model prediction value and the actual value, and the correlation coefficient can measure the consistency between the model and the actual data. Based on the error analysis results, the reliability parameters of the model are evaluated, such as precision, recall rate and F1 score, which can fully reflect the performance of the model. After obtaining the model reliability parameters, the shortcomings of the model are optimized and corrected. For example, if the model has large errors in certain areas, the model can be optimized by adjusting the interpolation method or re-dividing the rock interface; if the geometry of the fault or fracture does not match the actual data, the strike, dip or distance of the fault can be corrected. In addition, the model can be iteratively optimized in combination with more actual data (such as new logging data or seismic data) to gradually improve the accuracy and reliability of the model. Finally, the optimized three-dimensional model of the underground rock formation is saved and re-verified to ensure that it meets the needs of geological applications. This process can significantly improve the reliability of the model through continuous comparison, evaluation and optimization, and provide more accurate support for geological interpretation and resource exploration.

[0052] In summary, the embodiments of the present application have at least the following technical effects:

[0053] First, the three-dimensional seismic data of the target area is collected and acquired by using seismic exploration equipment, and the three-dimensional seismic data contains the propagation characteristic data of seismic waves in the underground medium; then, the three-dimensional seismic data is preprocessed to obtain standard three-dimensional seismic data; further, the three-dimensional seismic inversion technology is used to infer the underground rock attribute information from the standard three-dimensional seismic data based on the propagation characteristic data, and then the underground rock attribute information is analyzed by multi-attribute analysis to obtain the underground rock layer identification characteristics, and the underground rock layer identification characteristics include rock layer distribution characteristics and rock layer hierarchy structure; finally, a rock layer feature identification model is built, and the underground rock layer identification characteristics are analyzed and optimized by the rock layer feature identification model to determine the underground rock layer feature identification results, and three-dimensional modeling is performed based on the underground rock layer feature identification results to generate a three-dimensional model of the underground rock layer. The technical problems of insufficient analysis ability of traditional rock layer identification methods for complex geological structures and low accuracy of inversion results are solved, and the technical effect of high-precision characterization of rock layer spatial distribution characteristics and improved reliability of complex geological structure identification is achieved by integrating three-dimensional seismic inversion and multi-attribute intelligent analysis technology.

[0054] Embodiment 2 is based on the same inventive concept as the underground rock layer identification method based on three-dimensional seismic data in the above embodiment. Figure 2 As shown, the present application provides an underground rock formation identification system based on three-dimensional seismic data, wherein the system includes:

[0055] Data acquisition module 11: using seismic exploration equipment to acquire three-dimensional seismic data of the target area, the three-dimensional seismic data includes propagation characteristic data of seismic waves in underground media; preprocessing module 12: preprocessing the three-dimensional seismic data to obtain standard three-dimensional seismic data; data inversion module 13: applying three-dimensional seismic inversion technology to infer underground rock attribute information from the standard three-dimensional seismic data based on the propagation characteristic data; multi-attribute analysis module 14: performing multi-attribute analysis on the underground rock attribute information to obtain underground rock formation identification characteristics, the underground rock formation identification characteristics include rock formation distribution characteristics and rock formation hierarchy; three-dimensional modeling module 15: building a rock formation feature identification model, analyzing and optimizing the underground rock formation identification characteristics through the rock formation feature identification model, determining the underground rock formation feature identification results, and performing three-dimensional modeling based on the underground rock formation feature identification results to generate a three-dimensional model of the underground rock formation.

[0056] Furthermore, the preprocessing module 12 is used to perform the following method:

[0057] According to the application requirements of seismic data, a filter is selected to filter the three-dimensional seismic data to obtain usable three-dimensional seismic data; a signal regularization processing flow is obtained, and the signal regularization processing flow includes denoising processing, detrending processing and normalization; based on the signal regularization processing flow, the usable three-dimensional seismic data is de-signaled and regularized to obtain regularized three-dimensional seismic data; linear interpolation is used to fill the missing data of the regularized three-dimensional seismic data to obtain the standard three-dimensional seismic data.

[0058] Furthermore, the data inversion module 13 is used to execute the following method:

[0059] The three-dimensional seismic inversion technology is applied to perform inversion modeling based on the standard three-dimensional seismic data to establish an initial underground rock property model; the inversion algorithm is used to iteratively optimize the initial underground rock property model based on the propagation characteristic data to obtain a target underground rock property model; the target underground rock property model is used to perform inversion calculations on the standard three-dimensional seismic data to obtain the underground rock property information.

[0060] Furthermore, the data inversion module 13 is used to execute the following method:

[0061] Based on the propagation characteristic data, the initial underground rock property model is forward simulated to obtain simulated seismic data; the simulated seismic data is error calculated to obtain seismic data error information; based on the inversion algorithm, the initial underground rock property model is iteratively optimized through the seismic data error information to obtain the target underground rock property model.

[0062] Furthermore, the multi-attribute parsing module 14 is used to perform the following method:

[0063] Performing cluster analysis on the attribute data of the underground rock attribute information to obtain underground rock attribute clustering results; performing multi-attribute analysis based on the underground rock attribute clustering results to determine rock stratum distribution characteristics and rock stratum hierarchy; and composing the underground rock stratum identification characteristics based on the rock stratum distribution characteristics and rock stratum hierarchy.

[0064] Furthermore, the three-dimensional modeling module 15 is used to perform the following method:

[0065] Collect historical underground rock formation characteristic data sets, use deep learning network to train the historical underground rock formation characteristic data sets to obtain an initial feature recognition model; verify the initial feature recognition model to obtain model performance parameters; perform back propagation optimization on the initial feature recognition model based on the model performance parameters to build the rock formation characteristic recognition model.

[0066] Furthermore, the three-dimensional modeling module 15 is used to perform the following method:

[0067] According to the geological application objectives, the modeling boundary range and modeling element information are determined; based on the modeling boundary range and modeling element information, the underground rock formation feature identification results are subjected to topological relationship analysis to obtain rock formation modeling feature data; based on the rock formation modeling feature data, three-dimensional space conversion is performed to generate the underground rock formation three-dimensional model.

[0068] Furthermore, the three-dimensional modeling module 15 is used to perform the following method:

[0069] The three-dimensional model of the underground rock formation is verified and evaluated using actual geological exploration data to obtain model reliability parameters, and the three-dimensional model of the underground rock formation is optimized and corrected using the model reliability parameters.

[0070] Embodiment 3, based on the same inventive concept as the underground rock formation identification method based on three-dimensional seismic data in the aforementioned embodiment 1, the present application provides an electronic device, which may be a server, comprising a processor, a memory and a network interface connected via a system bus, wherein the processor of the electronic device is used to provide computing and control capabilities, the memory of the electronic device comprises a non-volatile storage medium and an internal memory, the non-volatile storage medium stores an operating system, a computer program and a database, the internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium, the database of the electronic device is used to store data, the network interface of the electronic device is used to communicate with an external terminal via a network connection, and the computer program is executed by the processor to implement the underground rock formation identification method based on three-dimensional seismic data.

[0071] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0072] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0073] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A method for identifying underground rock formations based on three-dimensional seismic data, characterized in that: The method comprises: Using seismic exploration equipment to collect and acquire three-dimensional seismic data of the target area, wherein the three-dimensional seismic data includes propagation characteristic data of seismic waves in underground media; Preprocessing the three-dimensional seismic data to obtain standard three-dimensional seismic data; Applying 3D seismic inversion technology to infer underground rock attribute information from the standard 3D seismic data based on the propagation characteristic data; Performing multi-attribute analysis on the underground rock attribute information to obtain underground rock strata identification features, wherein the underground rock strata identification features include rock strata distribution features and rock strata hierarchical structure; A rock formation feature recognition model is constructed, and the underground rock formation recognition features are analyzed and optimized through the rock formation feature recognition model to determine the underground rock formation feature recognition results, and three-dimensional modeling is performed based on the underground rock formation feature recognition results to generate a three-dimensional model of the underground rock formation.

2. The underground rock layer identification method based on three-dimensional seismic data according to claim 1, characterized in that: The step of obtaining standard three-dimensional seismic data comprises: According to the application requirements of the seismic data, a filter is selected to filter the three-dimensional seismic data to obtain usable three-dimensional seismic data; Acquire a signal regularization processing flow, wherein the signal regularization processing flow includes denoising processing, detrending processing and normalization; Performing signal regularization processing on the available three-dimensional seismic data based on the signal regularization processing flow to obtain regularized three-dimensional seismic data; Linear interpolation is used to fill missing data in the regularized three-dimensional seismic data to obtain the standard three-dimensional seismic data.

3. The underground rock layer identification method based on three-dimensional seismic data according to claim 1, characterized in that: The method of applying the three-dimensional seismic inversion technology to infer underground rock property information from the standard three-dimensional seismic data based on the propagation characteristic data includes: Applying 3D seismic inversion technology to perform inversion modeling based on the standard 3D seismic data to establish an initial underground rock property model; Iteratively optimizing the initial underground rock property model based on the propagation characteristic data using an inversion algorithm to obtain a target underground rock property model; The target underground rock property model is used to perform inversion calculation on the standard three-dimensional seismic data to obtain the underground rock property information.

4. The underground rock layer identification method based on three-dimensional seismic data according to claim 3, characterized in that: The step of obtaining the target underground rock property model comprises: Performing forward simulation on the initial underground rock property model based on the propagation characteristic data to obtain simulated seismic data; Performing error calculation on the simulated seismic data to obtain seismic data error information; The initial underground rock property model is iteratively optimized through the seismic data error information based on an inversion algorithm to obtain the target underground rock property model.

5. The underground rock layer identification method based on three-dimensional seismic data according to claim 1, characterized in that: The obtaining of underground rock formation identification features includes: Performing cluster analysis on the attribute data of the underground rock attribute information to obtain underground rock attribute clustering results; Performing multi-attribute analysis based on the underground rock attribute clustering results to determine the rock layer distribution characteristics and rock layer hierarchy; Based on the rock stratum distribution characteristics and rock stratum hierarchy, the underground rock stratum identification characteristics are obtained.

6. The underground rock layer identification method based on three-dimensional seismic data according to claim 1, characterized in that: The step of building a rock formation feature recognition model includes: Collecting a historical underground rock formation feature data set, and using a deep learning network to train the historical underground rock formation feature data set to obtain an initial feature recognition model; Verifying the initial feature recognition model to obtain model performance parameters; The initial feature recognition model is back-propagated and optimized based on the model performance parameters to build the rock formation feature recognition model.

7. The underground rock layer identification method based on three-dimensional seismic data according to claim 1, characterized in that: The generating of the underground rock stratum three-dimensional model comprises: Determine the modeling boundary range and modeling element information according to the geological application objectives; Performing a topological relationship analysis on the underground rock formation feature recognition result based on the modeling boundary range and modeling element information to obtain rock formation modeling feature data; A three-dimensional space conversion is performed based on the rock formation modeling feature data to generate the three-dimensional model of the underground rock formation.

8. The underground rock layer identification method based on three-dimensional seismic data according to claim 7, characterized in that: The method comprises: The three-dimensional model of the underground rock formation is verified and evaluated using actual geological exploration data to obtain model reliability parameters, and the three-dimensional model of the underground rock formation is optimized and corrected using the model reliability parameters.

9. An underground rock formation identification system based on three-dimensional seismic data, characterized in that: The system for implementing the underground rock formation identification method based on three-dimensional seismic data according to any one of claims 1 to 8 comprises: Data acquisition module: using seismic exploration equipment to acquire three-dimensional seismic data of the target area, wherein the three-dimensional seismic data includes propagation characteristic data of seismic waves in underground media; Preprocessing module: preprocessing the three-dimensional seismic data to obtain standard three-dimensional seismic data; Data inversion module: applying 3D seismic inversion technology to infer underground rock attribute information from the standard 3D seismic data based on the propagation characteristic data; Multi-attribute analysis module: performs multi-attribute analysis on the underground rock attribute information to obtain underground rock strata identification features, wherein the underground rock strata identification features include rock strata distribution features and rock strata hierarchical structure; Three-dimensional modeling module: build a rock formation feature recognition model, analyze and optimize the underground rock formation recognition features through the rock formation feature recognition model, determine the underground rock formation feature recognition results, and perform three-dimensional modeling based on the underground rock formation feature recognition results to generate a three-dimensional model of the underground rock formation.

10. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the underground rock formation identification method based on three-dimensional seismic data described in any one of claims 1 to 8 are implemented.

Images