Three-dimensional geological modeling optimization method based on GIS

Abstract

The invention discloses a GIS-based three-dimensional geological modeling optimization method. The method comprises the steps of data integration processing, information clustering extraction, drilling plane visualization, drilling model construction, geologic body model construction and three-dimensional geological modeling optimization. The invention relates to the technical field of three-dimensional geologic modeling, in particular to a GIS-based three-dimensional geologic modeling optimization method, which adopts a three-dimensional geologic modeling overall thought combining information clustering extraction, spatial interpolation optimization and geologic body model construction optimization, and obtains a three-dimensional geologic model through information clustering extraction, spatial interpolation optimization and geologic body model construction optimization. According to the method, the multi-source and multi-scale extraction capability of information in a data source is optimized, the precision and performance of drilling data modeling and geologic body data modeling are improved through spatial interpolation optimization and geologic body model construction optimization, the GIS-based three-dimensional modeling performance is optimized on the whole, and the usability and universality of the method are improved; and constructing a geologic body model by adopting a multi-coupling cascade mixed density network combining a random algorithm and geostatistical calculation.

Description

Technical Field

[0001] The invention relates to the technical field of three-dimensional geological modeling, and in particular to a three-dimensional geological modeling optimization method based on GIS. Background Art

[0002] The GIS-based 3D geological modeling optimization method combines geographic information system (GIS) and 3D modeling technology to build accurate and reliable geological 3D models. By collecting geological exploration data, remote sensing images and drilling data, and using the powerful spatial data analysis and processing functions of GIS, a 3D visualization model of the geological body is generated. This method improves modeling accuracy and efficiency through optimization algorithms, can accurately reflect the spatial distribution and structural characteristics of geological bodies, and provide support for mineral resource exploration, groundwater management and disaster prediction. It not only improves the accuracy and automation of modeling, but also supports real-time monitoring and decision analysis through dynamic updates and interactive analysis, and is widely used in geological research and engineering practice.

[0003] However, in the existing 3D geological modeling methods, there are technical problems that the existing 3D model construction methods are difficult to fully and accurately reflect the complexity and diversity of geological bodies in the address information extraction, spatial interpolation process and geological body modeling process; in the existing modeling drilling and stratigraphic information extraction methods, there is a traditional stratigraphic information extraction that uses a standard K-means clustering algorithm to intelligently extract information, but when faced with data sources of multiple distances and scales, this method is very susceptible to interference from the data source due to the complexity of the stratigraphic information, which leads to large errors and fluctuations in the extracted information, making it difficult to provide high-quality data for subsequent multiple modeling needs; in the existing borehole model spatial interpolation and visualization process, there is a problem that the spatial interpolation method is not suitable for boreholes. During model interpolation, since the borehole model is a prerequisite for modeling the geological body model, the borehole model also has high requirements for accuracy when used as a prerequisite. Therefore, the traditional spatial interpolation method also has technical problems that need to be improved in this regard. Among the existing geological body model construction optimization methods, there is a traditional method that uses the classic Kriging interpolation method to perform direct interpolation statistical analysis and then interpolate. The existing automation ideas mostly focus on the generation of original data sources, or use adversarial generative networks to generate samples in terms of parameters. Since the multi-source and multi-scale geological modeling task itself is a complex multi-variable modeling task, the existing methods have insufficient technical problems in dealing with complex spatial data distribution, uncertainty, and the correlation between different strata and geological parameters. Summary of the invention

[0004] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides a 3D geological modeling optimization method based on GIS. This scheme creatively adopts the overall idea of ​​3D geological modeling that combines information clustering extraction, spatial interpolation optimization and geological body model construction optimization. Through information clustering extraction, the multi-source and multi-scale extraction capabilities of information in the data source are optimized, and through spatial interpolation optimization and geological body model construction optimization, the accuracy and performance of borehole data modeling and geological body data modeling are improved respectively, which optimizes the GIS-based 3D modeling performance as a whole and improves the usability and versatility of the method. This scheme creatively adopts a clustering method improved by dynamic nearest neighbor labeling to perform information clustering extraction. By combining an adaptive weighted and optimized clustering algorithm, the quantity and quality of information extracted from modeling boreholes and stratum information are improved, so as to provide good data support for subsequent method improvements. This scheme The scheme creatively adopts the adaptive weighted Gaussian spatial interpolation method to optimize spatial interpolation and visualize the borehole plane. By continuing the weighted analysis of distance attributes and local density attributes extracted by clustering information, and optimizing the consistency of the data processing process, the accuracy of the spatial interpolation of the borehole model is improved, and a good model and parameter basis is provided for the subsequent geological body modeling. This scheme creatively adopts a multi-coupled cascade mixed density network that combines random algorithms and geostatistical calculations to construct a geological body model. By combining geostatistics as a reference data sample and using the cascade coupling training of the mixed density network, it can perform high-precision probability distribution modeling of geological parameters (such as porosity, lithology distribution, stratum thickness, etc.), capture complex distribution forms such as multi-peak distribution or long-tail characteristics, and at the same time, the data distribution of different strata or regions is decoupled layer by layer through the cascade structure, gradually approaching the real distribution. Each layer of the module further models the output of the previous layer, so that the multi-scale and multi-variable characteristics under complex geological environments can be better captured. Overall, it meets the geological modeling needs of multi-stage and multi-parameter complex modeling of multi-scale and multi-source data, and provides good practical experience for the implementation path of three-dimensional geological modeling.

[0005] The technical solution adopted by the present invention is as follows: The present invention provides a GIS-based three-dimensional geological modeling optimization method, which includes the following steps:

[0006] Step S1: data integration processing;

[0007] Step S2: information clustering extraction;

[0008] Step S3: Drilling plane visualization;

[0009] Step S4: drilling model construction;

[0010] Step S5: constructing a geological body model;

[0011] Step S6: Optimization of three-dimensional geological modeling.

[0012] Further, in step S1, the data integration process is used to collect, integrate and preprocess the original data set required for three-dimensional geological modeling optimization, specifically, to obtain the original data of the surface matrix survey point from the geographic information system through data collection, and to obtain the original data set of geological modeling through data integration and preprocessing;

[0013] The original data of the surface matrix survey points specifically include drilling engineering data, basic drilling information data, drilling stratification data and standard formation data;

[0014] The data integration and preprocessing specifically include data outlier detection, data missing value detection and data integrity detection, and data integration and preprocessing are performed by formatting the borehole depth and lithology description data;

[0015] The original data set for geological modeling specifically includes integrated drilling engineering data, integrated drilling basic information data, integrated drilling stratification data and integrated standard bottom layer data.

[0016] Further, in step S2, the information clustering extraction is used to extract hierarchical information from the drilling information in the original data set, specifically, based on the original data set of geological modeling, a clustering method improved by dynamic nearest neighbor labeling is used to perform information clustering extraction to obtain geological hierarchical information data, specifically including the following steps:

[0017] Step S21: data cleaning and optimization, specifically, using a local anomaly factor algorithm to identify outliers and perform data cleaning and optimization on the original data set of geological modeling to obtain cleaned and optimized data;

[0018] Step S22: cluster distance weighted optimization, specifically, by calculating the cluster weighted distance, performing weighted distance optimization according to the sample distance calculated in the cleaning optimization data, and obtaining the weighted distance weight;

[0019] Step S23: local density enhancement, specifically, performing dynamic K nearest neighbor value calculation by calculating local density information to obtain an adaptive K nearest neighbor value, and the calculation formula is:

[0020] ;

[0021] In the formula, K dy is the adaptive K nearest neighbor value, used as the clustering parameter for clustering information extraction, min(·) is the minimum value function, K max is the maximum K nearest neighbor value, density(·) is the local density calculation function, x i is the i-th input data sample, used to represent the data in the original data set of geological modeling, is the adaptive adjustment coefficient;

[0022] Step S24: Dynamic neighbor label enhancement, specifically, performing majority voting dynamic neighbor label enhancement through the cluster distance weighted optimization and the local density enhancement to obtain dynamic nearest neighbor improved label data, and the calculation formula is:

[0023] ;

[0024] Where label(·) is the dynamic nearest neighbor labeling improvement function, x i is the i-th input data sample, used to represent the data in the original data set of geological modeling, argmax is the maximum value function, L is the label index, is the total number of neighbor nodes after adaptive optimization, where x j is the jth neighbor node sample, j is the neighbor node sample index, K dy is the adaptive K nearest neighbor value, used as the clustering parameter for clustering information extraction, Wdist(·) is the weighted distance weight calculation function, Is an indicator function used to determine the neighbor node sample x j The corresponding label L j Is it equal to the label index L?

[0025] Step S25: multi-level label optimization, specifically adopting a multi-level K-means clustering method to extract hierarchical clustering information, and improving label data according to the dynamic nearest neighbor, performing iterative label enhancement, and obtaining multi-level optimized label data;

[0026] Step S26: Information clustering extraction, specifically, performing information clustering extraction of a clustering method improved by dynamic nearest neighbor labeling through the dynamic neighbor label enhancement and the multi-level label optimization to obtain geological stratification information data.

[0027] Further, in step S3, the borehole plane visualization is used to perform spatial interpolation optimization in combination with the borehole information and perform visualization settings, specifically, based on the geological stratification information data, an adaptive weighted Gaussian spatial interpolation method is used to perform spatial interpolation optimization and borehole plane visualization to obtain borehole plane visualization data, specifically including the following steps:

[0028] Step S31: standardizing the drilling data, specifically, performing Z-score standardization processing on the geological stratification information data to obtain standardized drilling data;

[0029] Step S32: adaptive weighted operator calculation, specifically, based on the standardized drilling data, by calculating the distance similarity in the drilling data, an adaptive weighted factor is obtained, and the calculation formula is:

[0030] ;

[0031] Where W(·) is the adaptive weighting factor calculation function, X a is the first drilling data point, used to represent the data point in the standardized drilling data, X b is the second borehole data point, used to represent the neighborhood data point of the first borehole data point, exp(·) is the natural base function, d(·) is the similarity calculation function, used to represent the distance similarity calculation function and attribute, specifically calculating the Euclidean distance, is the data point density parameter, used as an adaptive weighting parameter;

[0032] Step S33: Adaptive Gaussian interpolation calculation, specifically, by combining the adaptive weighting factor and setting an adaptive weighted Gaussian function, performing spatial interpolation calculation to obtain adaptive interpolation data, the calculation formula is:

[0033] ;

[0034] Where Z(·) is the geological attribute value calculation function, which is used to represent the adaptive interpolation data corresponding to the interpolation point, P is the interpolation point index, is the total number of interpolation points, X b is the second drilling data point, used to represent the neighborhood data point of the first drilling data point, W(P,X b ) is the corresponding interpolation point P and the second drilling data point X b The adaptive weighting factor, Z(X b ) is the second drilling data point X b The geological attribute values;

[0035] Step S34: spatial interpolation optimization, specifically, smoothing and optimizing the adaptive interpolation data through Gaussian filtering, combining the adaptive weighted operator described above, and introducing a composite adaptive weighted operator to obtain smooth optimized interpolation data, the calculation formula is:

[0036] ;

[0037] In the formula, Z s (·) is the Gaussian filter smoothing function, P is the interpolation point index, is the total number of interpolation points, X b is the second drilling data point, is the self-adaptive weighting factor introduced into the composite adaptive weighting operator, Z(X b ) is the second drilling data point X b The geological attribute value, W(P,X b) is the corresponding interpolation point P and the second drilling data point X b The adaptive weighting factor of is the composite weighted local density adjustment parameter, density(·) is the local density calculation function;

[0038] Step S35: drilling plane visualization, specifically, performing spatial interpolation according to the smoothed optimized interpolation data to obtain a spatial interpolation reference result, and combining the spatial interpolation reference result with the geological stratification information data for visualization to obtain geological stratification visualization plane data;

[0039] Step S36: Interactive visual optimization, specifically using WebGIS technology to perform interactive visualization of drilling information, representing drilling information through point layers, and visualizing depth, stratum and spatial position through attribute list information to obtain drilling plane visualization data.

[0040] Further, in step S4, the drilling model is constructed to perform drilling model initialization construction and geological simulation in combination with drilling information, specifically, according to the drilling plane visualization data, drilling model parameters are set, and drilling model initialization generation is performed to obtain initial drilling model data, specifically including the following steps:

[0041] Step S41: initializing drilling model parameters, specifically setting a drilling model construction initialization parameter set according to the initial drilling model data;

[0042] The drilling model constructs an initialization parameter set, specifically including model name, drilling radius and texture unit size parameters;

[0043] Step S42: standard formation editing, specifically, performing initial editing of the drilling model by setting the initialization parameter set of the drilling model, and obtaining formation information data by customizing the formation information;

[0044] The stratigraphic information data specifically includes stratigraphic codes, rock and soil names, stratigraphic sequences, and stratigraphic legend styles;

[0045] Step S43: generating a drilling model, specifically dividing the drilling model construction initialization parameter set into different stratigraphic units according to the stratigraphic information data, and generating a three-dimensional model for each stratigraphic unit to obtain an initial shape set of the stratigraphic three-dimensional model, and performing model optimization simulation according to the drilling model construction initialization parameter set and the texture unit size and stratigraphic legend style in the stratigraphic information data to generate initial drilling model data.

[0046] Further, in step S5, the geological body model is constructed for comprehensively constructing the geological body model in combination with the initialized borehole model, specifically, based on the geological stratification information data, the borehole plane visualization data and the initial borehole model data, a multi-coupled cascade mixed density network combining a random algorithm and geostatistical calculation is used to construct the geological body model to obtain geological body modeling data, specifically including the following steps:

[0047] Step S51: initializing geological body model parameters, specifically setting geological body model parameters according to the geological stratification information data, the drilling plane visualization data and the initial drilling model data to obtain a geological body model parameter set;

[0048] The geological body model parameter set specifically includes the geological body model name, horizontal grid spacing, sample influence radius, stratum variation coefficient and geological body texture unit size;

[0049] Step S52: geological body stratum editing, specifically, performing initial editing of the geological body stratum model by setting the geological body model parameter set, and obtaining geological body stratum information data by customizing stratum information;

[0050] The geological body stratigraphic information data specifically includes the geological body stratigraphic code, geological body rock and soil name, geological body stratigraphic sequence and geological body legend style;

[0051] Step S53: geostatistical analysis, specifically, using a kriging interpolation method to perform geostatistical analysis based on the geological body stratigraphic information data and the geological body model parameter set to obtain statistical attribute reference values ​​of geological body interpolation points;

[0052] Step S54: constructing a random simulation algorithm, specifically, generating spatial statistical features using the Monte Carlo method based on the statistical attribute reference values ​​of the geological body interpolation points, and obtaining statistical feature data of the random simulated geological body interpolation points through expected value and room difference calculation;

[0053] Step S55: constructing a multi-coupled cascade density hybrid network, specifically training a plurality of standard hybrid density networks and performing cascade combination to construct the multi-coupled cascade density hybrid network, performing multivariate joint analysis, and obtaining the output of geological body spatial parameters;

[0054] The multi-coupled cascade density hybrid network specifically includes an input layer, a feature extraction layer, a hybrid layer and an output layer;

[0055] The calculation formula for the construction process of the multi-coupled cascade density hybrid network is:

[0056] ;

[0057] In the formula, Model GC It is a multi-coupled cascade density hybrid network, which is used to represent the model for generating spatial parameters of geological body modeling and calculate the output of geological body spatial parameters. The whole is used to represent the sum of the outputs of the first c layers of density mixture network, where is the total number of coupled layers in the density hybrid network, is the coupling index of the density hybrid network, which is used to indicate the input source of the c-th layer hybrid density network, where c is the hierarchical index of the hybrid density network. It is The output of the output layer of the layer density hybrid network is used to represent the spatial parameter output of the geological body and is used as the input data of the hybrid density network of the c+1th layer, CN c+1 It is the calculated output of the feature extraction layer of the c+1th layer mixed density network, which is specifically used to represent the standard convolutional layer feature output using a nonlinear activation function. c+1 It is the calculation output of the mixed layer of the c+1th layer mixed density network, which is specifically used to represent the calculation output using the standard mixed density model. c+1 is the output of the output layer of the c+1th layer of the mixed density network;

[0058] Step S56: generating geological body modeling spatial parameters, specifically, training the geological body modeling spatial parameter generation model through the multi-coupled cascade density hybrid network based on the statistical feature data of the randomly simulated geological body interpolation points, and obtaining the geological body modeling spatial parameter generation model Model GC , and use the geological body modeling space parameters to generate the model Model GC , generate the spatial parameters of geological body modeling and obtain the geological attribute parameter set;

[0059] Step S57: constructing a geological body model, specifically, constructing a geological body model scene based on the geological attribute parameter set and combining image rendering technology to obtain geological body modeling data;

[0060] Furthermore, in step S6, the three-dimensional geological modeling optimization is used to store, manage and comprehensively optimize the three-dimensional model of the geological body. Specifically, the management and optimization of the geological body model are performed based on the geological body modeling data, and the drilling model data and the geological body model data are grouped and stored, a model directory and management system are established, and the three-dimensional geological modeling optimization is performed to obtain a three-dimensional geological modeling optimization database.

[0061] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0062] (1) In view of the technical problem that the existing 3D geological modeling methods are difficult to fully and accurately reflect the complexity and diversity of geological bodies in the process of address information extraction, spatial interpolation and geological body modeling, this scheme creatively adopts the overall idea of ​​3D geological modeling that combines information clustering extraction, spatial interpolation optimization and geological body model construction optimization. Through information clustering extraction, the multi-source and multi-scale extraction capabilities of information in the data source are optimized, and through spatial interpolation optimization and geological body model construction optimization, the accuracy and performance of borehole data modeling and geological body data modeling are improved respectively. Overall, the GIS-based 3D modeling performance is optimized, and the usability and versatility of the method are improved;

[0063] (2) In the existing modeling borehole and stratigraphic information extraction methods, there is a traditional stratigraphic information extraction method that uses a standard K-means clustering algorithm to intelligently extract information. However, when facing data sources of different distances and scales, this method is very susceptible to interference from data sources due to the complexity of stratigraphic information, which leads to large errors and fluctuations in the extracted information, making it difficult to provide high-quality data for subsequent modeling needs. This solution creatively uses a clustering method improved by dynamic nearest neighbor labeling to perform information clustering extraction. By combining an adaptive weighted and optimized clustering algorithm, the quantity and quality of information extracted from modeling borehole and stratigraphic information are improved, which can provide good data support for subsequent method improvements.

[0064] (3) In the existing borehole model spatial interpolation and visualization process, there is a problem that the spatial interpolation method is not applicable to the borehole model interpolation. Since the borehole model is a prerequisite for the geological body model, the borehole model also has high accuracy requirements when it is used as a prerequisite. Therefore, the traditional spatial interpolation method also has technical problems that need to be improved in this regard. This scheme creatively adopts an adaptive weighted Gaussian spatial interpolation method to perform spatial interpolation optimization and borehole plane visualization. By continuing the weighted analysis of distance attributes and local density attributes extracted by clustering information, and optimizing the consistency of the data processing process, the accuracy of the borehole model spatial interpolation is improved, and a good model and parameter basis is provided for subsequent geological body modeling.

[0065] (4) In the existing optimization methods for building geological models, there is a traditional method that uses the classic Kriging interpolation method to perform direct interpolation statistical analysis and then interpolate. The existing automation ideas mostly focus on the generation of original data sources, or use adversarial generative networks to generate samples in terms of parameters. Since the multi-source and multi-scale geological modeling task itself is a complex multi-variable modeling task, the existing methods have insufficient technical problems in dealing with complex spatial data distribution, uncertainty, and the correlation between different strata and geological parameters. This scheme creatively uses a multi-coupled cascade mixed density network that combines random algorithms and geostatistical calculations to build a geological model. By combining geological statistics as reference data samples and using cascade coupled training of mixed density networks, it is possible to perform high-precision probability distribution modeling of geological parameters (such as porosity, lithology distribution, stratum thickness, etc.) and capture complex distribution forms such as multi-peak distribution or long-tail characteristics. At the same time, the data distribution of different strata or regions is decoupled layer by layer through the cascade structure, gradually approaching the real distribution. Each layer of modules further models the output of the previous layer, so that the multi-scale and multi-variable characteristics in complex geological environments can be better captured. Overall, it meets the geological modeling needs of multi-stage and multi-parameter complex modeling of multi-scale, multi-source data, and provides good practical experience for the implementation path of three-dimensional geological modeling. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 A schematic diagram of a flow chart of a GIS-based three-dimensional geological modeling optimization method provided by the present invention;

[0067] Figure 2 This is a schematic diagram of the process of information clustering extraction in step S2;

[0068] Figure 3 A schematic diagram of the process of visualizing the drilling plane in step S3;

[0069] Figure 4 A schematic diagram of the process of constructing the drilling model in step S4;

[0070] Figure 5 Schematic diagram of the process of building a geological model in step S5.

[0071] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0072] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0073] In the description of the present invention, it should be understood that terms such as “upper”, “lower”, “front”, “back”, “left”, “right”, “top”, “bottom”, “inside” and “outside” indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as limiting the present invention.

[0074] Example 1, see Figure 1 The present invention provides a GIS-based three-dimensional geological modeling optimization method, which includes the following steps:

[0075] Step S1: data integration processing;

[0076] Step S2: information clustering extraction;

[0077] Step S3: Drilling plane visualization;

[0078] Step S4: drilling model construction;

[0079] Step S5: constructing a geological body model;

[0080] Step S6: Optimization of three-dimensional geological modeling.

[0081] By executing the above operations, in view of the technical problem that the existing three-dimensional geological modeling methods are difficult to fully and accurately reflect the complexity and diversity of geological bodies in the address information extraction, spatial interpolation process and geological body modeling process, this scheme creatively adopts the overall idea of ​​three-dimensional geological modeling that combines information clustering extraction, spatial interpolation optimization and geological body model construction optimization. Through information clustering extraction, the multi-source and multi-scale extraction capabilities of information in the data source are optimized, and through spatial interpolation optimization and geological body model construction optimization, the accuracy and performance of drilling data modeling and geological body data modeling are improved respectively, which optimizes the GIS-based three-dimensional modeling performance as a whole and improves the usability and versatility of the method.

[0082] Example 2, see Figure 1In step S1, the data integration process is used to collect, integrate and preprocess the original data set required for three-dimensional geological modeling optimization, specifically, from the geographic information system, through data collection, to obtain the original data of the surface matrix survey point, and through data integration and preprocessing, to obtain the original data set for geological modeling;

[0083] The original data of the surface matrix survey points specifically include drilling engineering data, basic drilling information data, drilling stratification data and standard formation data;

[0084] The data integration and preprocessing specifically include data outlier detection, data missing value detection and data integrity detection, and data integration and preprocessing are performed by formatting the borehole depth and lithology description data;

[0085] The original data set for geological modeling specifically includes integrated drilling engineering data, integrated drilling basic information data, integrated drilling layer data and integrated standard bottom layer data;

[0086] The integrated drilling engineering data specifically includes the engineering name, engineering code and coordinate system data;

[0087] The integrated basic information data of the drilling hole specifically includes hole code, hole mouth elevation, hole depth and geographic coordinate data;

[0088] The integrated drilling layer data specifically includes hole code, stratum code, layer bottom depth data, layer top depth data, layer thickness data, rock and soil name and engineering code;

[0089] The integrated standard bottom layer data specifically include stratum code, rock and soil name, stratum sequence, stratum color and engineering code.

[0090] Example 3, see Figure 1 and Figure 2 This embodiment is based on the above embodiment. In step S2, the information clustering extraction is used to extract hierarchical information from the drilling information in the original data set. Specifically, based on the original data set of geological modeling, a clustering method improved by dynamic nearest neighbor labeling is used to perform information clustering extraction to obtain geological hierarchical information data. Specifically, the following steps are included:

[0091] Step S21: data cleaning and optimization, specifically, using a local anomaly factor algorithm to identify outliers and perform data cleaning and optimization on the original data set of geological modeling to obtain cleaned and optimized data. The calculation formula is:

[0092] ;

[0093] Where LOF(·) is the local outlier factor algorithm function, sum(·) is the summation function, and xi is the i-th input data sample, used to represent the data in the original data set of geological modeling, i is the data sample index, dist(·) is the sample distance calculation function, specifically the Euclidean distance, x j is the jth neighbor node sample, j is the neighbor node sample index, reachability(·) is the reachability calculation function, which is used to represent the reachability between the input data sample and the neighbor node sample, N(x i ,k) is the input data sample x i The total number of K neighbor nodes, K is the K nearest neighbor value, which is used to represent the number of neighbor nodes of the input data sample and serves as a clustering parameter;

[0094] Step S22: cluster distance weighted optimization, specifically, by calculating the cluster weighted distance, according to the sample distance calculated in the cleaning optimization data, weighted distance optimization is performed to obtain the weighted distance weight, and the calculation formula is:

[0095] ;

[0096] Where Wdist(·) is the weighted distance calculation function, dist(·) is the sample distance calculation function, specifically the Euclidean distance, x j is the jth neighbor node sample, j is the neighbor node sample index, is the regularization constant;

[0097] Step S23: local density enhancement, specifically, performing dynamic K nearest neighbor value calculation by calculating local density information to obtain an adaptive K nearest neighbor value, and the calculation formula is:

[0098] ;

[0099] In the formula, K dy is the adaptive K nearest neighbor value, used as the clustering parameter for clustering information extraction, min(·) is the minimum value function, K max is the maximum K nearest neighbor value, density(·) is the local density calculation function, x i is the i-th input data sample, used to represent the data in the original data set of geological modeling, is the adaptive adjustment coefficient;

[0100] Step S24: Dynamic neighbor label enhancement, specifically, performing majority voting dynamic neighbor label enhancement through the cluster distance weighted optimization and the local density enhancement to obtain dynamic nearest neighbor improved label data, and the calculation formula is:

[0101] ;

[0102] Where label(·) is the dynamic nearest neighbor labeling improvement function, x i is the i-th input data sample, used to represent the data in the original data set of geological modeling, argmax is the maximum value function, L is the label index, is the total number of neighbor nodes after adaptive optimization, where x j is the jth neighbor node sample, j is the neighbor node sample index, K dy is the adaptive K nearest neighbor value, used as the clustering parameter for clustering information extraction, Wdist(·) is the weighted distance weight calculation function, Is an indicator function used to determine the neighbor node sample x j The corresponding label L j Is it equal to the label index L?

[0103] Step S25: multi-level label optimization, specifically adopting a multi-level K-means clustering method to extract hierarchical clustering information, and improving label data according to the dynamic nearest neighbor, performing iterative label enhancement, and obtaining multi-level optimized label data;

[0104] Step S26: Information clustering extraction, specifically, performing information clustering extraction of a clustering method improved by dynamic nearest neighbor labeling through the dynamic neighbor label enhancement and the multi-level label optimization to obtain geological stratification information data.

[0105] By performing the above operations, in view of the fact that in the existing modeling drilling and formation information extraction methods, the traditional formation information extraction adopts the standard K-means clustering algorithm to intelligently extract information, but this method is very susceptible to interference from data sources at multiple distances and scales due to the complexity of the formation information, which leads to large errors and fluctuations in the extracted information, making it difficult to provide high-quality data for subsequent various modeling needs. This solution creatively adopts a clustering method improved by dynamic nearest neighbor labeling to perform information clustering extraction. By combining the adaptive weighted and optimized clustering algorithm, the quantity and quality of information extracted from modeling drilling and formation information are improved, which can provide good data support for subsequent method improvements.

[0106] Example 4, see Figure 1 and Figure 3 This embodiment is based on the above embodiment. In step S3, the borehole plane visualization is used to combine the borehole information for spatial interpolation optimization and visualization setting. Specifically, based on the geological layering information data, an adaptive weighted Gaussian spatial interpolation method is used to perform spatial interpolation optimization and borehole plane visualization to obtain borehole plane visualization data. Specifically, the following steps are included:

[0107] Step S31: standardizing the drilling data, specifically, performing Z-score standardization processing on the geological stratification information data to obtain standardized drilling data;

[0108] Step S32: adaptive weighted operator calculation, specifically, based on the standardized drilling data, by calculating the distance similarity in the drilling data, an adaptive weighted factor is obtained, and the calculation formula is:

[0109] ;

[0110] Where W(·) is the adaptive weighting factor calculation function, X a is the first drilling data point, used to represent the data point in the standardized drilling data, X b is the second borehole data point, used to represent the neighborhood data point of the first borehole data point, exp(·) is the natural base function, d(·) is the similarity calculation function, used to represent the distance similarity calculation function and attribute, specifically calculating the Euclidean distance, is the data point density parameter, used as an adaptive weighting parameter;

[0111] Step S33: Adaptive Gaussian interpolation calculation, specifically, by combining the adaptive weighting factor and setting an adaptive weighted Gaussian function, performing spatial interpolation calculation to obtain adaptive interpolation data, the calculation formula is:

[0112] ;

[0113] Where Z(·) is the geological attribute value calculation function, which is used to represent the adaptive interpolation data corresponding to the interpolation point, P is the interpolation point index, is the total number of interpolation points, X b is the second drilling data point, used to represent the neighborhood data point of the first drilling data point, W(P,X b ) is the corresponding interpolation point P and the second drilling data point X b The adaptive weighting factor, Z(X b ) is the second drilling data point X b The geological attribute values;

[0114] Step S34: spatial interpolation optimization, specifically, smoothing and optimizing the adaptive interpolation data through Gaussian filtering, combining the adaptive weighted operator described above, and introducing a composite adaptive weighted operator to obtain smooth optimized interpolation data, the calculation formula is:

[0115] ;

[0116] In the formula, Z s (·) is the Gaussian filter smoothing function, P is the interpolation point index, is the total number of interpolation points, X b is the second drilling data point, is the self-adaptive weighting factor introduced into the composite adaptive weighting operator, Z(X b ) is the second drilling data point X b The geological attribute value, W(P,X b ) is the corresponding interpolation point P and the second drilling data point X b The adaptive weighting factor of is the composite weighted local density adjustment parameter, density(·) is the local density calculation function;

[0117] Step S35: drilling plane visualization, specifically, performing spatial interpolation according to the smoothed optimized interpolation data to obtain a spatial interpolation reference result, and combining the spatial interpolation reference result with the geological stratification information data for visualization to obtain geological stratification visualization plane data;

[0118] The geological layer visualization plane data specifically includes geological layer information and spatial distribution information;

[0119] Step S36: Interactive visual optimization, specifically using WebGIS technology to perform interactive visualization of drilling information, representing drilling information through point layers, and visualizing depth, stratum and spatial position through attribute list information to obtain drilling plane visualization data.

[0120] By executing the above operations, in the existing borehole model spatial interpolation and visualization process, there is a problem that the spatial interpolation method is not applicable to the borehole model interpolation. Since the borehole model is a prerequisite for the modeling of the geological body model, the borehole model also has high requirements for accuracy when used as a prerequisite. Therefore, the traditional spatial interpolation method also has technical problems that need to be improved in this regard. This solution creatively adopts an adaptive weighted Gaussian spatial interpolation method to perform spatial interpolation optimization and borehole plane visualization. By continuing the weighted analysis of distance attributes and local density attributes extracted by clustering information, and optimizing the consistency of the data processing process, the accuracy of the borehole model spatial interpolation is improved, and a good model and parameter basis is provided for subsequent geological body modeling.

[0121] Example 5, see Figure 1 and Figure 4 This embodiment is based on the above embodiment. In step S4, the drilling model is constructed to perform drilling model initialization construction and geological simulation in combination with drilling information. Specifically, according to the drilling plane visualization data, drilling model parameters are set and drilling model initialization generation is performed to obtain initial drilling model data. Specifically, the following steps are included:

[0122] Step S41: initializing drilling model parameters, specifically setting a drilling model construction initialization parameter set according to the initial drilling model data;

[0123] The drilling model constructs an initialization parameter set, specifically including model name, drilling radius and texture unit size parameters;

[0124] Step S42: standard formation editing, specifically, performing initial editing of the drilling model by setting the initialization parameter set of the drilling model, and obtaining formation information data by customizing the formation information;

[0125] The stratigraphic information data specifically includes stratigraphic codes, rock and soil names, stratigraphic sequences, and stratigraphic legend styles;

[0126] Step S43: generating a drilling model, specifically dividing the drilling model construction initialization parameter set into different stratigraphic units according to the stratigraphic information data, and generating a three-dimensional model for each stratigraphic unit to obtain an initial shape set of the stratigraphic three-dimensional model, and performing model optimization simulation according to the drilling model construction initialization parameter set and the texture unit size and stratigraphic legend style in the stratigraphic information data to generate initial drilling model data.

[0127] Example 5, see Figure 1 and Figure 5 This embodiment is based on the above embodiment. In step S5, the geological body model is constructed to comprehensively construct the geological body model in combination with the initialized borehole model. Specifically, the geological body model is constructed based on the geological stratification information data, the borehole plane visualization data and the initial borehole model data, using a multi-coupled cascade mixed density network that combines a random algorithm and geostatistical calculations to obtain geological body modeling data. Specifically, the following steps are included:

[0128] Step S51: initializing geological body model parameters, specifically setting geological body model parameters according to the geological stratification information data, the drilling plane visualization data and the initial drilling model data to obtain a geological body model parameter set;

[0129] The geological body model parameter set specifically includes the geological body model name, horizontal grid spacing, sample influence radius, stratum variation coefficient and geological body texture unit size;

[0130] The geological model name is used to represent the identifier of the geological model;

[0131] The horizontal grid spacing is used to control the meshing accuracy of the model, as the horizontal size of each grid unit;

[0132] The sample influence radius is used to define the range of influence of geological variables in space;

[0133] The formation change coefficient is used to describe the steepness of the formation change. The closer the value is to 1, the flatter the formation is, and the closer the value is to 0, the more dramatic the formation change is.

[0134] The geological body texture unit size is used to define the size of the texture unit within the geological body and the detail level of the geological body;

[0135] Step S52: geological body stratum editing, specifically, performing initial editing of the geological body stratum model by setting the geological body model parameter set, and obtaining geological body stratum information data by customizing stratum information;

[0136] The geological body stratigraphic information data specifically includes the geological body stratigraphic code, geological body rock and soil name, geological body stratigraphic sequence and geological body legend style;

[0137] The geological body stratum code is used to represent the unique code of each geological body stratum;

[0138] The geological body rock and soil name is used to indicate the rock and soil type in each geological body stratum;

[0139] The geological body stratigraphic sequence is used to indicate the order of geological body stratigraphic layers;

[0140] The geological body legend style is used to represent the display style of the geological body strata in the three-dimensional model;

[0141] Step S53: geostatistical analysis, specifically, using the Kriging interpolation method, based on the geological body stratigraphic information data and the geological body model parameter set, to perform geostatistical analysis to obtain the statistical attribute reference value of the geological body interpolation point, the calculation formula is:

[0142] ;

[0143] Where y(·) is a semi-differential calculation function, which is used to represent the degree of variation of geological variables with distance, h is the geostatistical distance variable, and n(h) is the total number of all data points within the geostatistical distance h. is the index of data points within the geostatistical distance h, It is a geological variable value calculation function, which is specifically used to represent the Kriging interpolation attribute value. It is The original geological data corresponding to the data points are used to represent the geological body stratigraphic information data and the geological parameters in the geological body model parameter set. is the reference value of the statistical attributes of the geological body interpolation points, are the data points to be interpolated by Kriging, It is Kriging interpolation weights corresponding to data points;

[0144] Step S54: constructing a random simulation algorithm, specifically, generating spatial statistical features using the Monte Carlo method based on the statistical attribute reference values ​​of the geological body interpolation points, and obtaining statistical feature data of the random simulated geological body interpolation points through expected value and room difference calculation;

[0145] Step S55: constructing a multi-coupled cascade density hybrid network, specifically training a plurality of standard hybrid density networks and performing cascade combination to construct the multi-coupled cascade density hybrid network, performing multivariate joint analysis, and obtaining the output of geological body spatial parameters;

[0146] The multi-coupled cascade density hybrid network specifically includes an input layer, a feature extraction layer, a hybrid layer and an output layer;

[0147] The calculation formula for the construction process of the multi-coupled cascade density hybrid network is:

[0148] ;

[0149] In the formula, Model GC It is a multi-coupled cascade density hybrid network, which is used to represent the model for generating spatial parameters of geological body modeling and calculate the output of geological body spatial parameters. The whole is used to represent the sum of the outputs of the first c layers of density mixture network, where is the total number of coupled layers in the density hybrid network, is the coupling index of the density hybrid network, which is used to indicate the input source of the c-th layer hybrid density network, where c is the hierarchical index of the hybrid density network. It is The output of the output layer of the layer density hybrid network is used to represent the spatial parameter output of the geological body and is used as the input data of the hybrid density network of the c+1th layer, CN c+1 It is the calculated output of the feature extraction layer of the c+1th layer mixed density network, which is specifically used to represent the standard convolutional layer feature output using a nonlinear activation function. c+1 It is the calculation output of the mixed layer of the c+1th layer mixed density network, which is specifically used to represent the calculation output using the standard mixed density model. c+1 is the output of the output layer of the c+1th layer of the mixed density network;

[0150] Step S56: generating geological body modeling spatial parameters, specifically, training the geological body modeling spatial parameter generation model through the multi-coupled cascade density hybrid network based on the statistical feature data of the randomly simulated geological body interpolation points, and obtaining the geological body modeling spatial parameter generation model Model GC , and use the geological body modeling space parameters to generate the model Model GC , generate the spatial parameters of geological body modeling and obtain the geological attribute parameter set;

[0151] Step S57: constructing a geological body model, specifically, constructing a geological body model scene based on the geological attribute parameter set and combining image rendering technology to obtain geological body modeling data;

[0152] The image rendering technology is specifically implemented using an API interface provided by WebGL (Web Graphics Library) graphics rendering technology.

[0153] By performing the above operations, in the existing geological body model construction optimization methods, there is a traditional method that uses the classic Kriging interpolation method to perform direct interpolation statistical analysis and then interpolation ideas, while the existing automation ideas mostly focus on the generation of original data sources, or use adversarial generative networks to generate samples in terms of parameters. Since the multi-source and multi-scale geological modeling task itself is a complex multi-variable modeling task, the existing methods have insufficient technical problems in dealing with complex spatial data distribution, uncertainty, and the correlation between different strata and geological parameters. This scheme creatively uses a multi-coupled cascade mixed density network that combines random algorithms and geostatistical calculations to construct a geological body model. By combining geological statistics as a reference data sample and using the cascade coupling training of the mixed density network, it can perform high-precision probability distribution modeling of geological parameters (such as porosity, lithology distribution, stratum thickness, etc.), capture complex distribution forms such as multi-peak distribution or long-tail characteristics, and at the same time, through the cascade structure, the data distribution of different strata or regions is decoupled layer by layer, gradually approaching the real distribution. Each layer of modules further models the output of the previous layer, so that the multi-scale and multi-variable characteristics in complex geological environments can be better captured. Overall, it meets the geological modeling needs of multi-stage and multi-parameter complex modeling of multi-scale, multi-source data, and provides good practical experience for the implementation path of three-dimensional geological modeling.

[0154] Example 6, see Figure 1 and Figure 5This embodiment is based on the above embodiment. In step S6, the three-dimensional geological modeling optimization is used to store and manage the three-dimensional model of the geological body and perform comprehensive optimization. Specifically, the management and optimization of the geological body model are performed based on the geological body modeling data, and the drilling model data and the geological body model data are grouped and stored, and a model directory and management system are established to perform three-dimensional geological modeling optimization and obtain a three-dimensional geological modeling optimization database.

[0155] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0156] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

[0157] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. A GIS-based three-dimensional geological modeling optimization method, characterized by: The method comprises the following steps: Step S1: Data integration and processing to obtain the original data set for geological modeling; Step S2: information clustering extraction, specifically, based on the original data set of geological modeling, using a clustering method improved by dynamic nearest neighbor labeling to perform information clustering extraction to obtain geological stratification information data; Step S3: drilling plane visualization, specifically, based on the geological stratification information data, an adaptive weighted Gaussian spatial interpolation method is used to perform spatial interpolation optimization and drilling plane visualization to obtain drilling plane visualization data, including the following steps: step S31: drilling data standardization; step S32: adaptive weighted operator calculation; step S33: adaptive Gaussian interpolation calculation; step S34: spatial interpolation optimization; step S35: drilling plane visualization; step S36: interactive visual optimization; Step S4: constructing a drilling model, specifically setting drilling model parameters according to the drilling plane visualization data, and initializing and generating the drilling model to obtain initial drilling model data; Step S5: geological body model construction, which is used to comprehensively construct the geological body model in combination with the initialized drilling model. Specifically, based on the geological stratification information data, the drilling plane visualization data and the initial drilling model data, a multi-coupled cascade mixed density network combining a random algorithm and geostatistical calculation is used to construct the geological body model to obtain geological body modeling data, including the following steps: step S51: geological body model parameter initialization; step S52: geological body stratum editing; step S53: geostatistical analysis; step S54: random simulation algorithm construction; step S55: multi-coupled cascade density mixed network construction; step S56: geological body modeling space parameter generation; step S57: geological body model construction; Step S6: Optimize the three-dimensional geological modeling to obtain a three-dimensional geological modeling optimization database.

2. The GIS-based three-dimensional geological modeling optimization method according to claim 1, characterized in that: In step S1, the original data of the surface matrix survey point specifically includes drilling engineering data, drilling basic information data, drilling layering data and standard formation data; The data integration and preprocessing specifically include data outlier detection, data missing value detection and data integrity detection, and data integration and preprocessing are performed by formatting the borehole depth and lithology description data; The original data set for geological modeling specifically includes integrated drilling engineering data, integrated drilling basic information data, integrated drilling stratification data and integrated standard bottom layer data.

3. The GIS-based three-dimensional geological modeling optimization method according to claim 2, characterized in that: In step S2, the information clustering extraction is used to extract hierarchical information from the drilling information in the original data set. Specifically, based on the original data set of geological modeling, a clustering method improved by dynamic nearest neighbor labeling is used to perform information clustering extraction to obtain geological hierarchical information data, which specifically includes the following steps: Step S21: data cleaning and optimization, specifically, using a local anomaly factor algorithm to identify outliers and perform data cleaning and optimization on the original data set of geological modeling to obtain cleaned and optimized data; Step S22: cluster distance weighted optimization, specifically, by calculating the cluster weighted distance, performing weighted distance optimization according to the sample distance calculated in the cleaning optimization data, and obtaining the weighted distance weight; Step S23: local density enhancement, specifically, performing dynamic K nearest neighbor value calculation by calculating local density information to obtain an adaptive K nearest neighbor value, and the calculation formula is: ； In the formula, K dy is the adaptive K nearest neighbor value, used as the clustering parameter for clustering information extraction, min(·) is the minimum value function, K max is the maximum K nearest neighbor value, density(·) is the local density calculation function, x i is the i-th input data sample, used to represent the data in the original data set of geological modeling, is the adaptive adjustment coefficient; Step S24: Dynamic neighbor label enhancement, specifically, performing majority voting dynamic neighbor label enhancement through the cluster distance weighted optimization and the local density enhancement to obtain dynamic nearest neighbor improved label data, and the calculation formula is: ； Where label(·) is the dynamic nearest neighbor labeling improvement function, x i is the i-th input data sample, used to represent the data in the original data set of geological modeling, argmax is the maximum value function, L is the label index, is the total number of neighbor nodes after adaptive optimization, where x j is the jth neighbor node sample, j is the neighbor node sample index, K dy is the adaptive K nearest neighbor value, used as the clustering parameter for clustering information extraction, Wdist(·) is the weighted distance weight calculation function, Is an indicator function used to determine the neighbor node sample x j The corresponding label L j Is it equal to the label index L? Step S25: multi-level label optimization, specifically adopting a multi-level K-means clustering method to extract hierarchical clustering information, and improving label data according to the dynamic nearest neighbor, performing iterative label enhancement, and obtaining multi-level optimized label data; Step S26: Information clustering extraction, specifically, performing information clustering extraction of a clustering method improved by dynamic nearest neighbor labeling through the dynamic neighbor label enhancement and the multi-level label optimization to obtain geological stratification information data.

4. The GIS-based three-dimensional geological modeling optimization method according to claim 3, characterized in that: In step S3, the borehole plane visualization is used to perform spatial interpolation optimization in combination with the borehole information and perform visualization settings. Specifically, based on the geological stratification information data, an adaptive weighted Gaussian spatial interpolation method is used to perform spatial interpolation optimization and borehole plane visualization to obtain borehole plane visualization data, which specifically includes the following steps: Step S31: standardizing the drilling data, specifically, performing Z-score standardization processing on the geological stratification information data to obtain standardized drilling data; Step S32: adaptive weighted operator calculation, specifically, based on the standardized drilling data, by calculating the distance similarity in the drilling data, an adaptive weighted factor is obtained, and the calculation formula is: ； Where W(·) is the adaptive weighting factor calculation function, X a is the first drilling data point, used to represent the data point in the standardized drilling data, X b is the second borehole data point, used to represent the neighborhood data point of the first borehole data point, exp(·) is the natural base function, d(·) is the similarity calculation function, used to represent the distance similarity calculation function and attribute, specifically calculating the Euclidean distance, is the data point density parameter, used as an adaptive weighting parameter; Step S33: Adaptive Gaussian interpolation calculation, specifically, by combining the adaptive weighting factor and setting an adaptive weighted Gaussian function, performing spatial interpolation calculation to obtain adaptive interpolation data, the calculation formula is: ； Where Z(·) is the geological attribute value calculation function, which is used to represent the adaptive interpolation data corresponding to the interpolation point, P is the interpolation point index, is the total number of interpolation points, X b is the second drilling data point, used to represent the neighborhood data point of the first drilling data point, W(P,X b ) is the corresponding interpolation point P and the second drilling data point X b The adaptive weighting factor, Z(X b ) is the second drilling data point X b The geological attribute values; Step S34: spatial interpolation optimization, specifically, smoothing and optimizing the adaptive interpolation data through Gaussian filtering, combining the adaptive weighted operator described above, and introducing a composite adaptive weighted operator to obtain smooth optimized interpolation data, the calculation formula is: ； In the formula, Z s (·) is the Gaussian filter smoothing function, P is the interpolation point index, is the total number of interpolation points, X b is the second borehole data point, is the self-adaptive weighting factor introduced into the composite adaptive weighting operator, Z(X b ) is the second drilling data point X b The geological attribute value, W(P,X b ) is the corresponding interpolation point P and the second drilling data point X b The adaptive weighting factor of is the composite weighted local density adjustment parameter, density(·) is the local density calculation function; Step S35: drilling plane visualization, specifically, performing spatial interpolation according to the smoothed optimized interpolation data to obtain a spatial interpolation reference result, and combining the spatial interpolation reference result with the geological stratification information data for visualization to obtain geological stratification visualization plane data; Step S36: Interactive visual optimization, specifically using WebGIS technology to perform interactive visualization of drilling information, representing drilling information through point layers, and visualizing depth, stratum and spatial position through attribute list information to obtain drilling plane visualization data.

5. The GIS-based three-dimensional geological modeling optimization method according to claim 4, characterized in that: In step S4, the drilling model is constructed to perform drilling model initialization construction and geological simulation in combination with drilling information, specifically, according to the drilling plane visualization data, drilling model parameters are set, and drilling model initialization generation is performed to obtain initial drilling model data, specifically including the following steps: Step S41: initializing drilling model parameters, specifically setting a drilling model construction initialization parameter set according to the initial drilling model data; The drilling model constructs an initialization parameter set, specifically including model name, drilling radius and texture unit size parameters; Step S42: standard formation editing, specifically, performing initial editing of the drilling model by setting the initialization parameter set of the drilling model, and obtaining formation information data by customizing the formation information; The stratigraphic information data specifically includes stratigraphic codes, rock and soil names, stratigraphic sequences, and stratigraphic legend styles; Step S43: generating a drilling model, specifically dividing the drilling model construction initialization parameter set into different stratigraphic units according to the stratigraphic information data, and generating a three-dimensional model for each stratigraphic unit to obtain an initial shape set of the stratigraphic three-dimensional model, and performing model optimization simulation according to the drilling model construction initialization parameter set and the texture unit size and stratigraphic legend style in the stratigraphic information data to generate initial drilling model data.

6. The GIS-based three-dimensional geological modeling optimization method according to claim 5, characterized in that: In step S5, the geological body model is constructed for comprehensive construction of the geological body model in combination with the initialized borehole model. Specifically, the geological body model is constructed based on the geological stratification information data, the borehole plane visualization data and the initial borehole model data, using a multi-coupled cascade mixed density network combining a random algorithm and geostatistical calculation to obtain geological body modeling data, which specifically includes the following steps: Step S51: initializing geological body model parameters, specifically setting geological body model parameters according to the geological stratification information data, the drilling plane visualization data and the initial drilling model data to obtain a geological body model parameter set; The geological body model parameter set specifically includes the geological body model name, horizontal grid spacing, sample influence radius, stratum variation coefficient and geological body texture unit size; Step S52: geological body stratum editing, specifically, performing initial editing of the geological body stratum model by setting the geological body model parameter set, and obtaining geological body stratum information data by customizing stratum information; The geological body stratigraphic information data specifically includes the geological body stratigraphic code, geological body rock and soil name, geological body stratigraphic sequence and geological body legend style; Step S53: geostatistical analysis, specifically, using a kriging interpolation method to perform geostatistical analysis based on the geological body stratigraphic information data and the geological body model parameter set to obtain statistical attribute reference values ​​of geological body interpolation points; Step S54: constructing a random simulation algorithm, specifically, generating spatial statistical features using the Monte Carlo method based on the statistical attribute reference values ​​of the geological body interpolation points, and obtaining statistical feature data of the random simulated geological body interpolation points through expected value and room difference calculation; Step S55: constructing a multi-coupled cascade density hybrid network, specifically training a plurality of standard hybrid density networks and performing cascade combination to construct the multi-coupled cascade density hybrid network, performing multivariate joint analysis, and obtaining the output of geological body spatial parameters; The multi-coupled cascade density hybrid network specifically includes an input layer, a feature extraction layer, a hybrid layer and an output layer; The calculation formula for the construction process of the multi-coupled cascade density hybrid network is: ； In the formula, Model GC It is a multi-coupled cascade density hybrid network, which is used to represent the model for generating spatial parameters of geological body modeling and calculate the output of geological body spatial parameters. The whole is used to represent the sum of the outputs of the first c layers of density mixture network, where is the total number of coupled layers in the density hybrid network, is the coupling index of the density hybrid network, which is used to indicate the input source of the c-th layer hybrid density network, where c is the hierarchical index of the hybrid density network. It is The output of the output layer of the layer density hybrid network is used to represent the spatial parameter output of the geological body and is used as the input data of the hybrid density network of the c+1th layer, CN c+1 It is the calculated output of the feature extraction layer of the c+1th layer mixed density network, which is specifically used to represent the standard convolutional layer feature output using a nonlinear activation function. c+1 It is the calculation output of the mixed layer of the c+1th layer mixed density network, which is specifically used to represent the calculation output using the standard mixed density model. c+1 is the output of the output layer of the c+1th layer of the mixed density network; Step S56: generating geological body modeling spatial parameters, specifically, training the geological body modeling spatial parameter generation model through the multi-coupled cascade density hybrid network based on the statistical feature data of the randomly simulated geological body interpolation points, and obtaining the geological body modeling spatial parameter generation model Model GC , and use the geological body modeling space parameters to generate the model Model GC , generate the spatial parameters of geological body modeling and obtain the geological attribute parameter set; Step S57: constructing a geological body model, specifically, constructing a geological body model scene based on the geological attribute parameter set in combination with image rendering technology to obtain geological body modeling data.

7. The GIS-based three-dimensional geological modeling optimization method according to claim 6, characterized in that: In step S6, the three-dimensional geological modeling optimization is used to store, manage and comprehensively optimize the three-dimensional model of the geological body. Specifically, the management and optimization of the geological body model are performed based on the geological body modeling data, and the drilling model data and the geological body model data are grouped and stored, and a model directory and management system are established to perform three-dimensional geological modeling optimization and obtain a three-dimensional geological modeling optimization database.

Images