Subsea near surface modeling method, apparatus, device, and storage medium

Abstract

The application discloses a seabed shallow formation modeling method, comprising the following steps: collecting various detection data of a seabed shallow formation; analyzing the detection data by using a predetermined spatial data analysis algorithm to obtain spatial feature data; analyzing formation-related detection data in the detection data by using a predetermined statistical algorithm to obtain formation feature data; preliminarily constructing a three-dimensional model of the seabed shallow formation based on the spatial feature data and the formation feature data; constructing a plurality of machine learning models, training corresponding machine learning models by using the spatial feature data, and training corresponding machine learning models by using the formation feature data; and integrating all the trained machine learning models into the three-dimensional model, and the integrated three-dimensional model is a geological model of the seabed shallow formation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of seabed geological structure modeling, in particular to a seabed shallow stratum modeling method, device, equipment and storage medium. BACKGROUND

[0002] The field of seabed geological structure research is a field that is currently of great concern in society, and the mapping and geological exploration of seabed topography are of great significance for understanding the natural form of the seabed and seabed resources. Traditional mapping and geological exploration of seabed topography is usually carried out by means of sonar detection, drilling and other methods to obtain seabed geological structure information. However, these methods use single technical means, and for complex and variable marine environments, their efficiency and accuracy need to be improved. SUMMARY

[0003] The present application provides a seabed shallow stratum modeling method, device, equipment and storage medium, which efficiently and accurately models the seabed shallow stratum profile, thereby obtaining accurate seabed geological structure information.

[0004] In order to achieve the above-mentioned purpose, the present application provides a seabed shallow stratum modeling method, which comprises:

[0005] Collecting a plurality of detection data of the seabed shallow stratum;

[0006] Using a predetermined spatial data analysis algorithm to analyze the detection data to obtain spatial feature data; and

[0007] Using a predetermined calculation algorithm to analyze the detection data related to the stratum to obtain stratum feature data;

[0008] Preliminarily constructing a three-dimensional model of the seabed shallow stratum based on the spatial feature data and the stratum feature data;

[0009] Constructing a plurality of machine learning models, training the corresponding machine learning model using the spatial feature data, and training the corresponding machine learning model using the stratum feature data;

[0010] Integrating all the trained machine learning models into the three-dimensional model, and the integrated three-dimensional model is a geological model of the seabed shallow stratum.

[0011] Optionally, the step of using a predetermined spatial data analysis algorithm to analyze the detection data to obtain spatial feature data comprises:

[0012] Constructing a partial differential equation according to the detection data, and confirming a base function according to the spatial feature data;

[0013] transforming the partial differential equation into an approximate equation based on the basis function, obtaining an equation solution of the approximate equation, the equation solution being the spatial feature data.

[0014] Optionally, the basis function comprises at least one of a polynomial, a trigonometric function, a Gaussian function, and a spline function.

[0015] Optionally, after the step of analyzing the detection data using a predetermined spatial data analysis algorithm to obtain spatial feature data, the method further comprises:

[0016] integrating and transforming the spatial feature data to obtain a plurality of visualized information;

[0017] constructing a seabed geological information database based on the spatial feature data and the plurality of visualized information.

[0018] Optionally, the stratigraphic feature data comprises geological composition data, geological distribution data, and geological change data, and the step of analyzing the detection data related to strata using a predetermined calculation algorithm to obtain stratigraphic feature data comprises:

[0019] analyzing the detection data related to strata using a geological calculation algorithm to obtain the geological composition data and the geological distribution data;

[0020] obtaining a pre-trained prediction model, obtaining representative detection data from the detection data related to strata, inputting the representative detection data into the prediction model, and obtaining the geological change data output by the prediction model.

[0021] Optionally, after the step of integrating all trained machine learning models into the three-dimensional model to obtain a geological model of the seabed shallow strata, the method further comprises a step of optimizing the geological model based on a seabed geological physical behavior simulation method, wherein the step of optimizing the geological model based on the seabed geological physical behavior simulation method comprises:

[0022] determining a seabed geological physical behavior, determining a simulation method according to the seabed geological physical behavior, and obtaining geological data to be simulated from the geological model;

[0023] simulating the seabed geological physical behavior based on the simulation method and the geological data to be simulated;

[0024] obtaining simulation result data, and verifying the accuracy of the geological model based on the simulation result data;

[0025] if the accuracy is lower than a predetermined accuracy threshold, optimizing the geological model.

[0026] Optionally, the seabed geophysical behavior simulation method further comprises, after the step of optimizing the geological model, integrating the optimized geological model into a user interface.

[0027] The embodiment of the present application further provides a seabed shallow formation modeling device, and the device comprises:

[0028] a collection module, configured to collect a plurality of detection data of the seabed shallow formation;

[0029] an analysis module, configured to analyze the detection data by using a predetermined analysis algorithm to obtain geological characteristic data, wherein the geological characteristic data comprises spatial characteristic data and formation characteristic data;

[0030] a construction module, configured to preliminarily construct a three-dimensional model of the seabed shallow formation based on the spatial characteristic data and the formation characteristic data;

[0031] a training module, configured to construct a plurality of machine learning models, train a corresponding machine learning model by using the spatial characteristic data, and train a corresponding machine learning model by using the formation characteristic data;

[0032] an integration module, configured to integrate all the trained machine learning models into the three-dimensional model, and the integrated three-dimensional model is a geological model of the seabed shallow formation.

[0033] The embodiment of the present application further provides a seabed shallow formation modeling device, and the device comprises a memory and a processor, wherein the memory stores computer executable instructions which can be run on the processor, and the computer executable instructions are executed by the processor to implement the above method.

[0034] The embodiment of the present application further provides a computer readable storage medium, wherein the computer readable storage medium stores computer executable instructions, and the computer executable instructions can be executed by one or more processors to implement the above method.

[0035] The method and system of the embodiment of the present application have the following advantages:

[0036] In the embodiment of the present application, first, a plurality of detection data of the shallow seabed stratum is collected, the detection data is analyzed by a spatial data analysis algorithm and a calculation algorithm to obtain spatial feature data for describing the seabed topography and stratum structure and stratum feature data for describing the seabed stratum, then, a three-dimensional model is constructed using the spatial feature data and the stratum feature data, and a plurality of machine learning models are trained using the spatial feature data and the stratum feature data, and finally, the plurality of trained machine learning models are integrated into the three-dimensional model. The embodiment of the present application can quickly and accurately obtain the spatial feature data and the stratum feature data for describing the seabed geology by using the spatial data analysis algorithm and the calculation algorithm to analyze the plurality of detection data, and can realize efficient and accurate modeling of the shallow seabed stratum profile based on the spatial feature data and the stratum feature data, and can obtain accurate seabed geological structure information after modeling. In addition, by integrating the plurality of trained machine learning models into the three-dimensional model, the three-dimensional model can automatically predict and classify a plurality of seabed geological structures, and the performance of the three-dimensional model is improved. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 A schematic diagram of a hardware device environment for a seabed shallow stratum modeling method in an embodiment;

[0038] Figure 2 A flowchart of a seabed shallow stratum modeling method in an embodiment;

[0039] Figure 3 A schematic diagram of a seabed shallow stratum modeling device in an embodiment;

[0040] Figure 4 A schematic diagram of the hardware composition of a seabed shallow stratum modeling device in an embodiment. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application. In addition, the technical features involved in each embodiment of the present application described below can be combined with each other as long as they do not conflict with each other.

[0042] Figure 1 A schematic diagram of a hardware device environment in an embodiment of the present application, such as Figure 1As shown, the seabed shallow stratum modeling device 001, sonar detection devices A1-A3, geological radar B1-B2, and other seabed detection devices are shown. For example, the sonar detection device can be a multi-beam sonar device, a side-scan sonar device, and the like. The geological radar can be carried by an underwater unmanned submarine, and of course, other suitable seabed detection devices can be used for detection and data acquisition. The number of each type of seabed detection device can be one or more, which is not limited herein. After each seabed detection device acquires corresponding data, the data can be transmitted to the seabed shallow stratum modeling device 001. The seabed shallow stratum modeling device 001 has storage and computing capabilities, and can analyze and process data of each detection device, and perform three-dimensional modeling to obtain accurate seabed geological structure information.

[0043] As shown in Figure 2 The embodiment of the present application provides a seabed shallow stratum modeling method, which is applied to Figure 1 the hardware device environment shown, and includes the following steps:

[0044] S101, collecting a plurality of detection data of a seabed shallow stratum;

[0045] In this embodiment, a plurality of detection devices or methods can be used to detect the seabed shallow stratum to obtain a plurality of detection data. The detection devices include sonar detection devices, geological radars, and other seabed detection devices, such as multi-beam sonar devices, side-scan sonar devices, underwater unmanned submarine-mounted geological radars, and the like. The above detection devices or the combination of the above detection devices and other suitable seabed detection devices can be used to detect the seabed shallow stratum and obtain corresponding detection data. Of course, other suitable seabed detection devices can also be used for detection and to obtain corresponding detection data, which is not limited herein.

[0046] Among them, the multi-beam sonar device can collect multi-beam sonar data, through which the seabed topography can be plotted; the side-scan sonar device can collect side-scan sonar data, through which the seabed stratum structure can be detected; the underwater unmanned submarine-mounted geological radar can collect radar data, through which the seabed sediments can be analyzed.

[0047] S102, analyzing the detection data using a predetermined spatial data analysis algorithm to obtain spatial feature data;

[0048] Before analyzing the detection data, the detection data can be preprocessed, including removing noise and irrelevant data from the detection data using data cleaning techniques, filling in missing data, and the like.

[0049] In this embodiment, the spatial feature data describes the seafloor topography and stratigraphic structure, which is a kind of spatially dependent data, including topographic features, stratigraphic thickness, sediment distribution, etc. Of course, other spatially dependent data describing the seafloor topography and stratigraphic structure are also included, which is not limited here.

[0050] It can be understood that the above-mentioned spatial data analysis algorithm can be a tool integrated in certain geographic information analysis software, or a separate algorithm tool with a corresponding input / output interface for calling by other devices. The spatial data analysis algorithm can be the Galerkin method, and of course other suitable spatial data analysis algorithms such as other approximate solving methods similar to the Galerkin method can also be used, which is not limited here.

[0051] In an embodiment, the detection data can be analyzed by the spatial data analysis algorithm integrated in the geographic information system:

[0052] Firstly, the geographic information system is configured, for example, the appropriate spatial coordinate system, data hierarchy and visualization parameters are set to adapt to the characteristics of the seafloor detection data.

[0053] Then, the detection data is imported into the geographic information system, for example, the collected multi-beam sonar data, side scan sonar data and geological radar data are imported into the geographic information system. The imported detection data is preprocessed, including format conversion, coordinate correction, etc. to ensure the accuracy and consistency of the data.

[0054] Then, the detection data is analyzed by using the spatial data analysis tool in the geographic information system, in order to analyze the topographic features of the seafloor, estimate the stratigraphic thickness, analyze the sediment distribution, etc. In an embodiment, the Galerkin method is used as the spatial data analysis algorithm in the geographic information system to analyze the detection data.

[0055] The Galerkin method is a numerical analysis technique used to solve partial differential equations, which is a method of finite element analysis, which solves the problem on a continuous domain by converting it into a discrete system. The core idea is to transform a complex problem into an approximate problem that is easier to solve, that is, the solution of the original problem is approximated as a linear combination of a set of basis functions, and these basis functions are usually selected as functions compatible with the boundary conditions of the problem. Specifically, a set of basis functions (such as polynomials or other simple functions) is first selected, and the basis functions are defined throughout the analysis domain. Then, the solution of the original partial differential equation is expressed as a linear combination of these basis functions. By minimizing the residual (i.e. the difference between the approximate solution of the equation and the actual solution), a set of linear equations can be obtained, and the approximate solution can be obtained by solving the linear equations. In this embodiment, the Galerkin method is applied to the spatial data analysis of the seafloor detection data to process the sonar data and topographic features, which can be used to simulate and predict the seafloor geological structure.

[0056] In particular, step S102 comprises:

[0057] A11, constructing a partial differential equation according to the detection data, and confirming a base function according to the spatial feature data;

[0058] Wherein, a mathematical model suitable for the characteristics of the detection data is established, including defining a partial differential equation and boundary conditions. The base function is confirmed according to the spatial feature data (i.e. the target of using the Galerkin method, for example, simulating the propagation of sonar data in different strata or predicting the distribution of strata).

[0059] The base function suitable for the characteristics of the seabed geological data is selected, and the selection of the base function directly affects the accuracy and efficiency of the Galerkin method. The base function can be a polynomial, a trigonometric function, etc., and needs to be selected according to the characteristics of the seabed geological data and the specific needs of the problem to be solved, i.e. the characteristics of the detection data and the analysis target need to be considered. The following are examples of base function types and their applications in seabed geological data analysis:

[0060] 1. Polynomial, application scenario example: when the data shows a relatively simple linear or quadratic relationship, a polynomial base function can be selected. For example, when simulating the uniform sediment layer of the seabed stratum, a polynomial base function can be used to approximate this regular change.

[0061] 2. Trigonometric function, application scenario example: when dealing with periodic or highly fluctuating seabed geological data, a trigonometric function (such as sine and cosine functions) can be selected. Due to its periodicity and fluctuation characteristics, the trigonometric function is suitable for simulating and analyzing seabed geological features with significant periodic changes. For example, when analyzing the structure of seabed sediment layers affected by tides, the trigonometric function can well simulate this periodic change.

[0062] 3. Gaussian function (radial basis function), application scenario example: Gaussian function is suitable for simulating geological data with local characteristics. When analyzing the seabed sediment characteristics in a specific area, Gaussian function can help highlight the local changes and characteristics in that area.

[0063] 4. Spline function, application scenario example: when the data shows different trends in different intervals, a spline function can be selected, which can flexibly fit. For example, when simulating the topographic changes along a certain seabed path, the spline function can adapt to the different rates of change of the terrain at different locations.

[0064] It can be understood that one or more of the above functions can be selected as the base function, and of course, one or more of the above functions and other suitable functions can be selected as the base function of the present embodiment, which is not limited here.

[0065] A12, transform the partial differential equation into a set of approximate equations based on the basis functions, obtain the equation solution of the approximate equations, and the equation solution is the spatial feature data.

[0066] The partial differential equation is transformed into a set of approximate equations based on the basis functions by using the Galerkin method, and the set of approximate equations is solved by using a numerical method. For example, when simulating the periodic distribution of seafloor sediments with depth, trigonometric functions are used as basis functions for approximation and equation solving:

[0067] 1. Define the objective and establish a mathematical model: the objective is to predict and simulate the periodic distribution of sediments. A partial differential equation is established, which describes the variation of sediment distribution with depth and other factors (e.g. tides).

[0068] 2. Select trigonometric functions as basis functions: select sine and cosine functions as basis functions to capture the periodicity and volatility of the data. Determine the parameters of the basis functions, such as frequency and amplitude, to better fit the actual data.

[0069] 3. Galerkin approximation: the original partial differential equation is transformed into a set of approximate equations by using the trigonometric function basis function, and the parameters and boundary conditions in the approximation process can be adjusted to ensure the accuracy of the approximation.

[0070] 4. Solve the approximate equation: solve the approximate equation by using a numerical method, and the equation solution is the spatial feature data, i.e. the prediction or simulation result of the sediment distribution. The error in the solving process can be analyzed, and other basis functions or parameters can be selected as needed to improve the result.

[0071] 5. Further analysis, application of prediction or simulation results: analyze the simulation results obtained by using the trigonometric function basis function, compare them with the actual collected geological sample data or known geological information, and apply the obtained sediment periodic distribution results to the analysis of seafloor geology, such as evaluating the impact of sediments on the seafloor environment, or guiding the development of seafloor resources.

[0072] The above Galerkin method can effectively simulate and predict seafloor geological data with significant periodic characteristics by using trigonometric functions as basis functions, providing a powerful tool for in-depth understanding of seafloor stratigraphic structure.

[0073] In other embodiments, the Galerkin method can also be optimized to achieve the purpose of optimizing spatial feature data, for example, the prediction or simulation result of the sediment distribution obtained above can be compared with actual geological data or existing geological models, and parameters and / or other basis functions can be adjusted based on the comparison results, and the result is optimized by re-solving based on the adjusted parameters and / or selected basis functions.

[0074] In other embodiments, the spatial feature data described above can be further processed into visualization data and a seabed geological information database, including the following steps:

[0075] S201, integrating and converting the spatial feature data to obtain various visualization information;

[0076] S202, constructing a seabed geological information database based on the spatial feature data and the various visualization information.

[0077] In this embodiment, the spatial feature data is integrated together to obtain all spatial information of the seabed geology, such as topographic features, stratum thickness, sediment distribution, etc. Thus, a seabed geological map can be further drawn to intuitively display the seabed geology. Specifically, the integrated spatial feature data can be converted, for example, directly using the visualization function of a geographic information system to obtain various visualization information, including three-dimensional topographic maps, stratum profile maps, and heat maps, etc.

[0078] In addition, the spatial feature data and the various visualization information can be saved to a database to construct a unified and multi-dimensional seabed geological information database. In order to ensure the wide application and sharing of the data in the database, the spatial feature data and the various visualization information can be output in a standard format, such as GeoJSON or Shapefile, etc. A data sharing mechanism is set up to allow other systems or applications to access and use the data in the database.

[0079] S103, analyzing the stratum-related detection data in the detection data using a predetermined statistical algorithm to obtain the stratum feature data.

[0080] The stratum feature data describes the characteristics of the seabed stratum, and includes geological composition data, geological distribution data, and geological change data, etc. The geological composition data refers to various geological compositions of the stratum, including rock types and mineral types, etc. The geological distribution data refers to the spatial distribution of various geological compositions of the stratum. The geological change data refers to the change of the geological compositions, including the deposition rate of the geological compositions and the change of the mineral compositions, etc.

[0081] The step S103 includes:

[0082] A21, analyzing the stratum-related detection data using a geological statistical algorithm to obtain the geological composition data and the geological distribution data;

[0083] A22, obtaining a pre-trained prediction model, obtaining representative survey data from the stratum-related survey data, inputting the representative survey data into the prediction model, and obtaining the geological change data output by the prediction model.

[0084] In an embodiment, different geological components can be identified using geostatistical methods such as principal component analysis, cluster analysis, etc. From this, the proportion and types of various minerals and sediments can be determined. Discriminant analysis and other methods can also be used to classify data based on geochemical characteristics to distinguish between different types of sedimentary and metamorphic rocks, etc. Geostatistical methods such as Kriging interpolation can also be used to estimate the geological components and properties of unexplored areas. Spatial autocorrelation analysis methods such as Moran's I can also be used to analyze the spatial distribution patterns and correlations of different geological components.

[0085] For the geological change data, a statistical model can be established to predict the geological change data of the seabed. Establishing the statistical model includes the following steps:

[0086] 1. Determine the prediction model (e.g., linear regression or logistic regression model), which can use random forests or support vector machines to improve prediction accuracy. This embodiment uses support vector machines to predict geological change data.

[0087] 2. Prepare the sample data set for the support vector machine. From the stratum-related survey data, select data that has a significant impact on geological features or is representative, so as to effectively represent the characteristics of the stratum, such as rock type, mineral content, stratum thickness, etc.

[0088] 3. Data preprocessing, including normalizing data to adapt to the requirements of the support vector machine, such as scaling all data to a uniform range.

[0089] 4. Use the selected representative survey data to build a support vector machine model, including selecting a suitable kernel function (e.g., linear, polynomial, or radial basis function) and adjusting model parameters (e.g., C parameter and kernel parameter).

[0090] 5. Divide the training set from the sample data set and label it, such as training samples of mineral composition and content, corresponding labels of mineral composition changes. Train the support vector machine using the training set.

[0091] 6. Evaluate the trained support vector machine model using methods such as cross-validation, confusion matrix, and ROC curve to determine its performance, such as accuracy. Optimize the model based on the evaluation results, including adjusting the kernel function type, optimizing the model parameters, increasing the training data, and balancing the training set, etc.

[0092] It can be understood that before analyzing the data, the survey data related to the stratum can be pre-processed, including data standardization, missing value processing, and outlier detection, etc. Exploratory data analysis can also be implemented to understand the basic characteristics of the data, such as central tendency, dispersion, and distribution pattern, etc., to provide a reference for subsequent modeling.

[0093] The above-mentioned optimized support vector machine model is put into practical application, and new or unlabeled data is input to predict the stratum feature data. Through the above steps, the support vector machine model can provide accurate prediction for the stratum sample data, effectively identify and predict the geological composition, distribution pattern and geological change, thereby providing important data support for the seabed stratum profile modeling.

[0094] S104, a three-dimensional model of the seabed shallow stratum is preliminarily constructed based on the spatial feature data and the stratum feature data;

[0095] The three-dimensional model constructed based on the spatial feature data and the stratum feature data can well show the seabed geological structure and material distribution.

[0096] In this embodiment, a three-dimensional geological modeling software can be used to construct the three-dimensional model:

[0097] Firstly, a three-dimensional geological modeling software with powerful data processing capability, flexible modeling tool, efficient rendering performance and suitable for seabed geological data processing is determined. The modeling software is configured, including setting parameters suitable for seabed geological modeling, such as resolution, stratum interface processing method and rendering option, etc.

[0098] Then, the spatial feature data and the stratum feature data are imported into the three-dimensional geological modeling software. The imported data is properly processed, for example, the topographic feature data is meshed to make it suitable for three-dimensional modeling. Alternatively, the spatial feature data can be integrated and converted in advance to obtain a visualization graph, and then the visualization graph is input into the three-dimensional geological modeling software, for example, the topographic features are integrated and converted to obtain a topographic map, and the topographic features and the topographic map are input into the three-dimensional geological modeling software.

[0099] Next, the three-dimensional model of the seabed is preliminarily constructed, including modeling of the topography, division of the stratum interface and mapping of the geological composition, etc., to reproduce the vertical and horizontal distribution of the stratum in the model, ensuring that the model can accurately reflect the actual seabed geological structure.

[0100] Finally, the constructed three-dimensional model is evaluated to verify its accuracy and reality, including comparison with the sample data collected on site or collation with known seabed geological information, etc. According to the evaluation results, the three-dimensional model is adjusted and optimized as necessary, for example, the stratum boundary is corrected, the distribution of the geological composition is adjusted, etc.

[0101] In addition, the three-dimensional visualization tool of the modeling software can be used to generate a three-dimensional view and animation of the three-dimensional model to visually display the seabed geological structure.

[0102] S105, constructing a plurality of machine learning models, training a corresponding machine learning model using the spatial feature data, and training a corresponding machine learning model using the formation feature data;

[0103] In order to better predict or classify the geological feature data, so as to integrate the machine learning model into the three-dimensional model, so that the three-dimensional model has the function of predicting and classifying the geological structure, the embodiment can use the machine learning algorithm to achieve the purpose of predicting or classifying the geological feature data, including the following steps:

[0104] 1. For a plurality of machine learning models, determine the target of each machine learning model, for example, predict formation characteristics, classify geological components, or identify specific geological structures, etc. Determine the performance indicators of machine learning, such as accuracy, recall rate and F1 score, to evaluate the effectiveness of the model.

[0105] 2. Obtain the spatial feature data and formation feature data described above, such as topographic feature data, geological component data, etc. Perform feature engineering, including selecting key features, generating new features, and normalizing data to improve model learning efficiency and prediction accuracy.

[0106] 3. Select a machine learning model, select a suitable machine learning algorithm according to the characteristics and target of the spatial feature data or formation feature data, such as support vector machine, random forest or deep learning network. This embodiment takes the selection of support vector machine as an example: select a suitable support vector machine model according to the characteristics of the seabed geological feature data, such as its dimension, complexity and expected output. Support vector machine includes support vector machine for classification and support vector machine for regression. Select a suitable kernel function, including linear kernel, polynomial kernel, radial basis (RBF) kernel, etc. The choice of kernel function depends on the characteristics of the geological feature data and the modeling target.

[0107] 4. Geological feature data preprocessing, including normalizing data, handling missing values and removing noise, etc. to meet the requirements of machine learning models. Select geological feature data to ensure that the data input into the machine learning model can effectively express the characteristics of the geological feature data and improve the training efficiency and prediction accuracy of the model.

[0108] 5. Train the machine learning models, train the corresponding machine learning model using spatial feature data, and train the corresponding machine learning model using formation feature data. In this step, model parameters such as C parameter (penalty factor of error term) and kernel function parameters need to be set.

[0109] 6. Implement cross-validation to evaluate the generalization ability of each machine learning model, and adjust the model parameters according to the verification results: according to the results of cross-validation and the performance indicators (such as accuracy, recall rate) of the model, the model is optimized, including adjusting the selection of kernel function, changing the parameter setting or introducing more training data. In addition, for particularly complex data sets, integrated methods or deep learning techniques can also be used to improve the performance of the machine learning model.

[0110] 7. Evaluate the performance of the machine learning model on the training set and the test set, and verify the prediction or classification results of the model, which can be achieved by comparing with the geological sample data collected in the field or comparing with the existing geological knowledge.

[0111] 8. Use methods such as confusion matrix, ROC curve, etc. to analyze the classification or prediction effect of the model. According to the analysis results, the model is further optimized, and the optimization operation includes adjusting the algorithm parameters, increasing the training data or improving the feature engineering.

[0112] It is worth noting that the execution order of the above steps S104 of constructing a three-dimensional model and S105 of training a plurality of machine learning models is not limited, and the step S104 can be executed first and then the step S105, or the step S105 can be executed first and then the step S104, which is not limited here.

[0113] S106, integrate all trained machine learning models into the three-dimensional model, and the integrated three-dimensional model is the geological model of the seafloor shallow formation.

[0114] The plurality of trained machine learning models are integrated into the three-dimensional model in the embodiment, and it can be understood that the trained machine learning models are the optimized machine learning models described above, and the output results of the machine learning models can be integrated into the three-dimensional model, i.e. the seafloor geological structure is automatically predicted and classified, so that the three-dimensional model has the corresponding seafloor geological structure prediction or classification function. Alternatively, the input and output interfaces of each machine learning model can be designed to effectively apply the output results of the machine learning model to the construction and optimization of the three-dimensional model.

[0115] It can be understood that the data in all the above embodiments can be appropriately preprocessed before being input into the model, such as standardization, normalization, etc., to adapt to the input requirements of the model.

[0116] The method of the embodiment of the application has the following advantages:

[0117] In the embodiment of the present application, first, a plurality of detection data of the shallow seabed stratum is collected, the detection data is analyzed by a spatial data analysis algorithm and a calculation algorithm to obtain spatial feature data for describing the seabed topography and stratum structure and stratum feature data for describing the seabed stratum, then, a three-dimensional model is constructed using the spatial feature data and the stratum feature data, and a plurality of machine learning models are trained using the spatial feature data and the stratum feature data, and finally, the plurality of trained machine learning models are integrated into the three-dimensional model. The embodiment of the present application can quickly and accurately obtain the spatial feature data and the stratum feature data for describing the seabed geology by using the spatial data analysis algorithm and the calculation algorithm to analyze the plurality of detection data, and can realize efficient and accurate modeling of the seabed shallow stratum profile based on the spatial feature data and the stratum feature data, and can obtain accurate seabed geological structure information after modeling. In addition, by integrating the plurality of trained machine learning models into the three-dimensional model, the three-dimensional model can automatically predict and classify a plurality of seabed geological structures, and the performance of the three-dimensional model is improved.

[0118] In an embodiment, on the basis of the above-mentioned embodiment, after the step S106, it further includes a step of optimizing the geological model based on a seabed geophysical behavior simulation method, wherein the step of optimizing the geological model based on the seabed geophysical behavior simulation method includes the following steps:

[0119] B11, determining a seabed geophysical behavior, and determining a simulation method and obtaining geological data to be simulated from the geological model according to the seabed geophysical behavior;

[0120] In the embodiment, first, a seabed geophysical behavior suitable for the seabed geological structure is determined, the seabed geophysical behavior includes but is not limited to seismic wave propagation, rock mechanics movement or fluid dynamics movement, and a corresponding simulation method is determined according to the seabed geophysical behavior, for example, the simulation method of seismic wave propagation is a seismic wave propagation simulation method.

[0121] Geological data to be simulated is extracted from the three-dimensional model, for example, the geological data to be simulated includes geological composition of the stratum, topographic features, etc., to ensure the authenticity and accuracy of the data and improve the authenticity of the simulation.

[0122] B12, simulating the seabed geophysical behavior based on the simulation method and the geological data to be simulated;

[0123] In an embodiment, simulation software can be selected to simulate the seabed geophysical behavior described above, which has high-fidelity simulation capabilities and can handle complex geological structures and physical processes. A corresponding calling interface can be set up to call the selected simulation software. After selecting the simulation software, corresponding simulation parameters such as wave speed, density, and fluid pressure are set to ensure that the simulation environment matches the actual environment. The seabed geophysical behavior is simulated, for example, based on the stratigraphic feature data of a certain area of the seabed, the propagation process of seismic waves in different strata is simulated to evaluate the response characteristics of the strata.

[0124] B13, obtaining simulation result data, verifying the accuracy of the geological model based on the simulation result data;

[0125] The simulation result data is obtained from the simulation software, for example, the propagation speed and attenuation speed of the strata when the seismic wave propagates in the strata, the simulation result data is analyzed and compared with the actual geological data or known geological information to verify the accuracy of the simulation, and the accuracy of the geological model is calculated or evaluated.

[0126] B14, if the accuracy is lower than a predetermined accuracy threshold, optimizing the geological model.

[0127] If the accuracy of the geological model is lower than the predetermined accuracy threshold, for example, the accuracy threshold is 0.95, the geological model is optimized or adjusted, for example, the physical properties or structural arrangement of the strata are corrected, the related geological feature data is modified, etc. The optimized geological model is integrated with the simulation result data, and the geological model is updated to ensure that it can accurately reflect the simulation results and the latest geological structure.

[0128] In this embodiment, by simulating the seabed geophysical behavior of the data in the geological model, not only a verification method is provided for the geological model, but also the understanding of the seabed geological structure and the seabed geophysical behavior is deepened, and the accuracy of the geological model is further improved.

[0129] In an embodiment, after the step of optimizing the geological model based on the seabed geophysical behavior simulation method, the optimized geological model is integrated into a user interface.

[0130] In this embodiment, the optimized geological model is integrated into a friendly user interface, and the user can understand the more realistic seabed geological structure through visual graphics and operations. The geological model can also provide geological information support for marine engineering, environmental protection, resource exploration, and geological disaster prevention. In addition, the geological model can be continuously updated and maintained to ensure its accuracy and practicality.

[0131] Exemplarily, when oil and gas resources are explored, oil and gas resources are explored in a sea area, a seabed stratum is modeled in detail by using a geological model, and a position and a scale of an oil and gas reservoir can be accurately predicted. When a seabed geological disaster is warned, a seabed stratum structure can be deeply understood by using a geological model, potential seabed landslides or earthquake activities are predicted and warned, and key information is provided for disaster prevention. When a marine environment is protected, a seabed environment is evaluated by using a geological model, and a scientific basis is provided for demarcation of a marine biological protection zone and marine environment protection.

[0132] As shown in Figure 3 The embodiment of the present application provides a seabed shallow stratum modeling device, which comprises:

[0133] A collection module is configured to collect a plurality of detection data of a seabed shallow stratum.

[0134] An analysis module is configured to analyze the detection data by using a predetermined analysis algorithm to obtain geological feature data, wherein the geological feature data comprises spatial feature data and stratum feature data.

[0135] A construction module is configured to preliminarily construct a three-dimensional model of the seabed shallow stratum based on the spatial feature data and the stratum feature data.

[0136] A training module is configured to construct a plurality of machine learning models, train a corresponding machine learning model by using the spatial feature data, and train a corresponding machine learning model by using the stratum feature data.

[0137] An integration module is configured to integrate all the trained machine learning models into the three-dimensional model, and the integrated three-dimensional model is a geological model of the seabed shallow stratum.

[0138] The functions of each module of the seabed shallow stratum modeling device of the embodiment can be referred to the corresponding part of the seabed shallow stratum modeling method described above, and will not be described here.

[0139] As shown in Figure 4 The embodiment of the present application also provides a seabed shallow stratum modeling device 1000, which comprises a memory 1001 and a processor 1002, wherein the memory stores computer executable instructions, and the processor implements the following steps when running the computer executable instructions on the memory:

[0140] A plurality of detection data of a seabed shallow stratum is collected.

[0141] The detection data is analyzed by using a predetermined analysis algorithm to obtain geological feature data, wherein the geological feature data comprises spatial feature data and stratum feature data.

[0142] A three-dimensional model of the seabed shallow stratum is preliminarily constructed based on the spatial feature data and the stratum feature data.

[0143] a plurality of machine learning models are constructed, a corresponding machine learning model is trained using the spatial feature data, and a corresponding machine learning model is trained using the formation feature data;

[0144] All the trained machine learning models are integrated into the three-dimensional model, and the integrated three-dimensional model is the geological model of the seafloor shallow formation.

[0145] In practical applications, the monitoring device can also include other necessary elements, including but not limited to any number of input 1003 and output devices 1004, processors, controllers, memories, etc., and all systems that can implement the big data management method of the embodiments of the present application are within the scope of protection of the present application.

[0146] The memory includes, but is not limited to, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), or compact disc read-only memory (CD-ROM). The memory is used for related instructions and data.

[0147] The input device is used to input data and / or signals, and the output device is used to output data and / or signals. The output device and the input device can be independent devices, or they can be a whole device.

[0148] The processor can include one or more processors, such as one or more central processing units (CPUs). In the case of a CPU, the CPU can be a single-core CPU or a multi-core CPU. The processor can also include one or more special-purpose processors, such as GPUs, FPGAs, etc., for accelerated processing.

[0149] The memory is used to store the program code and data of the network device.

[0150] The processor is used to call the program code and data in the memory, and execute the steps in the above method embodiments. For details, please refer to the description in the method embodiments, which will not be repeated here.

[0151] In several embodiments provided in the present application, it should be understood that the disclosed system and method can be implemented in other manners. For example, the division of the units is only a logical function division, and there can be another division manner for actual implementation. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. The displayed or discussed coupling, or direct coupling or communication connection between units can be indirect coupling or communication connection through some interface, and can be electrical, mechanical or in other forms.

[0152] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they can be located in one place or distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0153] In the above embodiments, all or part of the embodiments can be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, the computer program instructions generate the processes or functions according to the embodiments of the present application. The computer can be a general-purpose computer, a special-purpose computer, a computer network or other programmable system. The computer instructions can be stored in a computer-readable storage medium or transmitted by the computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through a wired (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.) manner. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. integrated with one or more available media. The available medium can be read-only memory (ROM), or random access memory (RAM), or magnetic medium, such as floppy disk, hard disk, magnetic tape, optical medium, such as digital versatile disc (DVD), or semiconductor medium, such as solid state disk (SSD), etc.

[0154] The above merely provides the specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of various equivalent modifications or replacements within the technical range disclosed by the present application, and these modifications or replacements should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for modeling shallow seafloor strata, characterized in that, The method includes: Collect various types of detection data from the shallow seabed; The probe data is analyzed using a predetermined spatial data analysis algorithm to obtain spatial feature data. This spatial feature data describes the seafloor topography and stratigraphic structure, including topographic features, stratigraphic thickness, and sediment distribution data. By using a predetermined statistical algorithm to analyze the stratigraphic-related detection data in the detection data, stratigraphic characteristic data is obtained. The stratigraphic characteristic data includes geological composition data, geological distribution data, and geological change data. The geological composition data includes various geological components of the stratigraphy, the geological distribution data includes the spatial distribution of various geological components of the stratigraphy, and the geological change data includes the changes in geological components. Based on the spatial feature data and the stratigraphic feature data, a preliminary three-dimensional model of the shallow seabed strata is constructed. Construct multiple machine learning models and determine the objective of each machine learning model; train the corresponding machine learning model using the spatial feature data; and train the corresponding machine learning model using the stratigraphic feature data. All trained machine learning models are integrated into the three-dimensional model, and the integrated three-dimensional model is the geological model of the shallow seabed. The step of analyzing the stratigraphic-related detection data in the detection data using a predetermined statistical algorithm to obtain stratigraphic characteristic data includes: analyzing the stratigraphic-related detection data using a geostatistical algorithm to obtain the geological composition data and the geological distribution data; obtaining a pre-trained prediction model; obtaining representative detection data from the stratigraphic-related detection data; inputting the representative detection data into the prediction model; and obtaining the geological change data output by the prediction model. The step of analyzing the probe data using a predetermined spatial data analysis algorithm to obtain spatial feature data includes: Partial differential equations are constructed based on the probe data, and basis functions are confirmed based on the spatial feature data; Based on the basis functions, the partial differential equation is transformed into an approximate equation, and the solution of the approximate equation is obtained, wherein the solution is the spatial feature data; Following the step of integrating all trained machine learning models into the 3D model, resulting in an integrated 3D model that serves as a geological model of the shallow seabed, the method further includes optimizing the geological model based on a seabed geophysical behavior simulation method. This optimization includes: Determine the seabed geophysical behavior, and based on the seabed geophysical behavior, determine the simulation method and obtain the geological data to be simulated from the geological model; Based on the simulation method and the geological data to be simulated, the seabed geophysical behavior is simulated; Obtain simulation result data, and verify the accuracy of the geological model based on the simulation result data; If the accuracy rate is lower than a predetermined accuracy threshold, the geological model is optimized.

2. The method according to claim 1, characterized in that, The basis functions include at least one of polynomial, trigonometric, Gaussian, and spline functions.

3. The method according to claim 1 or 2, characterized in that, After the step of analyzing the probe data using a predetermined spatial data analysis algorithm to obtain spatial feature data, the method further includes: The spatial feature data is integrated and transformed to obtain various types of visual information; A seabed geological information database is constructed based on the aforementioned spatial feature data and various visualization information.

4. The method according to claim 1, characterized in that, After the step of optimizing the geological model based on the seabed geophysical behavior simulation method, the method further includes: integrating the optimized geological model into the user interface.

5. A device for modeling shallow seabed strata, characterized in that, The apparatus for performing the seafloor shallow seismic modeling method according to any one of claims 1-4, the apparatus comprising: The acquisition module is used to collect various types of detection data from the shallow seabed. The analysis module is used to analyze the exploration data using a predetermined analysis algorithm to obtain geological feature data. The geological feature data includes spatial feature data and stratigraphic feature data. The spatial feature data includes topographic features, stratigraphic thickness, and sediment distribution data. The stratigraphic feature data includes geological composition data, geological distribution data, and geological change data. The geological composition data includes various geological components of the strata. The geological distribution data includes the spatial distribution of various geological components of the strata. The geological change data includes the changes in geological composition. The construction module is used to initially construct a three-dimensional model of the shallow seabed strata based on the spatial feature data and the stratigraphic feature data; The training module is used to construct multiple machine learning models and determine the target of each machine learning model, train the corresponding machine learning model using the spatial feature data, and train the corresponding machine learning model using the stratigraphic feature data. An integration module is used to integrate all trained machine learning models into the three-dimensional model, and the integrated three-dimensional model is a geological model of the shallow seabed. Following the step of integrating all trained machine learning models into the 3D model, resulting in an integrated 3D model that serves as a geological model of the shallow seabed, the method further includes optimizing the geological model based on a seabed geophysical behavior simulation method. This optimization includes: Determine the seabed geophysical behavior, and based on the seabed geophysical behavior, determine the simulation method and obtain the geological data to be simulated from the geological model; Based on the simulation method and the geological data to be simulated, the seabed geophysical behavior is simulated; Obtain simulation result data, and verify the accuracy of the geological model based on the simulation result data; If the accuracy rate is lower than a predetermined accuracy threshold, the geological model is optimized.

6. A device for modeling shallow seabed strata, characterized in that, The device includes a memory and a processor, the memory storing computer-executable instructions that can run on the processor, the computer-executable instructions being executed by the processor to implement the method as described in any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions that can be executed by one or more processors to implement the method as described in any one of claims 1-4.

Images