A method and system for constructing a three-dimensional model of urban engineering geology

Abstract

The present invention discloses a method and system for constructing a three-dimensional model of urban engineering geology, which relates to the technical field of constructing a three-dimensional model of urban engineering geology, and includes: constructing a stratigraphic sequence table of engineering geology; using combined geophysical prospecting to collect data on the factors in the stratigraphic sequence table of engineering geology; constructing and training an integrated learning neural network model according to the collected data; based on the prediction results, layer by layer identifying and classifying the geological properties of each underground layer; dividing the space into three-dimensional grids, and specifying corresponding geological properties within each grid; and using visualization software to display. The present invention establishes a stratigraphic sequence system of engineering geology, laying a foundation for the standardization work of subsequent geotechnical engineering investigation data; proposes a combined method of urban engineering geophysical prospecting suitable for assisting in the division of engineering geology strata, providing technical support for the application of geophysical prospecting methods in the field of urban engineering geology; and establishes a three-dimensional model of urban engineering geology, providing basic data support for the construction of smart cities.

Description

Technical Field

[0001] The present invention relates to the technical field of constructing three-dimensional models of urban engineering geology, and more specifically, to a method and system for constructing three-dimensional models of urban engineering geology. Background Technique

[0002] The research on three-dimensional geoscience visualization is mainly applied to the numerical simulation of oilfield development and mine exploitation. Later, many scholars have conducted a lot of research on its theoretical methods, data structures, and software development in different application fields. The modeling technology and modeling methods have been continuously developed and gradually matured, mainly including the ordinary borehole modeling method, the borehole modeling method based on horizon calibration, the method based on reticulated topological profiles and cross-fold profiles, the three-dimensional geological multi-field coupling modeling method, the multi-source interactive complex geological body modeling method, and the integrated modeling method of geological structure and in-situ stress simulation, etc.

[0003] In the process of urban layout optimization and adjustment, urban development space selection, agricultural zoning, and municipal construction planning, as well as the development and utilization of underground space and the construction of sponge cities, geological resources are needed as support and optimization in every aspect. Establishing a three-dimensional model of engineering geology has extremely important practical significance for the construction of a smart city, and is also an important support for constructing the basic framework of a smart city and an important link in building a smart city.

[0004] However, the existing three-dimensional models of engineering geology focus on model construction, and lack sufficient consideration of the data sources and classifications for model construction. There is no existing method for constructing a three-dimensional model of engineering geology centered around data selection and determination.

[0005] Therefore, how to propose a method and system for constructing a three-dimensional model of urban engineering geology, based on engineering exploration borehole data, and dividing strata according to geological age, geological origin, and lithofacies conditions, and constructing a three-dimensional model of urban engineering geology based on the engineering geology stratigraphic sequence table to improve the accuracy and authenticity of the model is an urgent problem for those skilled in the art to solve. Summary of the Invention

[0006] In view of this, the present invention provides a method and system for constructing a three-dimensional model of urban engineering geology, based on engineering exploration borehole data, dividing strata according to geological age, geological origin, and lithofacies conditions, and constructing a three-dimensional model of urban engineering geology based on the engineering geology stratigraphic sequence table to improve the accuracy and authenticity of the model. To achieve the above purpose, the present invention adopts the following technical solutions:

[0007] A method for constructing a three-dimensional model of urban engineering geology includes:

[0008] Based on engineering exploration borehole data, dividing strata according to geological age, geological origin, and lithofacies conditions, and constructing an engineering geology stratigraphic sequence table;

[0009] Use a composition exploration to collect data on the factors in the engineering geological stratigraphic sequence table;

[0010] Construct and train an integrated learning neural network model based on the collected data;

[0011] Based on the prediction results of the integrated learning neural network model, identify and classify the geological properties of each underground layer layer by layer;

[0012] Divide the space into three-dimensional grids, and specify the corresponding geological properties within each grid;

[0013] Use visualization software to display the constructed three-dimensional model of urban engineering geology.

[0014] Optionally, the construction of the engineering geological stratigraphic sequence table includes:

[0015] Division range: Divide respectively on the plane and in depth, and divide into detailed layers and general layers in depth;

[0016] Stratification and coding: Use a two-level coding unit of "main layer + sub-layer" to code the strata, and determine and number the engineering geological main layers according to the formation age, sedimentary environment, sedimentary sequence and rock and soil types of the strata;

[0017] Each genetic layer and main layer have a top-down sequential relationship, and the sub-layers within each main layer are coded according to the sedimentary priority and particle size.

[0018] Optionally, the composition exploration includes: surface microgravity method, microtremor method, high-density resistivity method, equivalent reverse magnetic flux transient electromagnetic method, transient surface wave method, H / V spectral ratio method and ground penetrating radar method.

[0019] Optionally, it also includes preprocessing the collected geophysical exploration data, and the preprocessing includes: cleaning, denoising and normalizing the collected geophysical exploration data;

[0020] Use the Z-score method to identify outliers in sequence, and delete the identified outliers;

[0021] Check whether there are duplicate data, delete duplicate records, and retain unique data;

[0022] Apply median filtering to the signal containing noise and retain the edge information;

[0023] Scale the data to the range of [0, 1] through min-max normalization.

[0024] Optionally, the construction and training of the integrated learning neural network model based on the collected data includes:

[0025] Divide the collected data into a training set, a validation set and a test set;

[0026] Select a fully connected neural network, a convolutional neural network, a recurrent neural network, a long short-term memory network, a graph neural network, and a deep belief network to form an ensemble learning neural network model;

[0027] Train and validate the ensemble learning neural network model using a training set and a validation set;

[0028] Evaluate the performance of the ensemble learning neural network model on a test set;

[0029] Use cross-validation to tune the hyperparameters of the model and the ensemble strategy.

[0030] Optionally, it further includes: Voting separately based on the prediction results of the fully connected neural network, convolutional neural network, recurrent neural network, long short-term memory network, graph neural network, and deep belief network to form the ensemble learning neural network model, and selecting the category with the most votes through the voting method as the output of the ensemble learning neural network model. Among them, the prediction result of the recurrent neural network is used as the control group of the prediction result of the ensemble learning neural network model, and the prediction result of the ensemble learning neural network model is determined to be true within a preset range not exceeding the prediction result of the recurrent neural network.

[0031] Optionally, the step of layer-by-layer identifying and classifying the geological properties of each underground layer includes:

[0032] Determine the thickness, depth, and interlayer transition conditions of each layer according to the layering standard;

[0033] Determine the corresponding geological properties for each layer through identification and classification;

[0034] Input the geological properties layer by layer in finite element analysis software to construct a three-dimensional model;

[0035] Define the boundary conditions and interactions of each layer;

[0036] Compare the model calculation results with the field observation data to verify the accuracy of the model.

[0037] Optionally, the step of dividing the space into three-dimensional grids and specifying the corresponding geological properties for each grid includes:

[0038] Establish an initial solid model, perform mesh generation and initial geological property assignment on the initial solid model to obtain an initial grid model;

[0039] Perform solidification processing on the grid cells in the initial grid model to determine a preset solid model;

[0040] Perform a Boolean operation on a preset entity model, obtain a fitted entity and a non-fitted entity according to the result of the Boolean operation, re-partition the fitted entity and perform a second geological attribute assignment to obtain a re-fitted sub-entity, perform a volume mesh division on the re-fitted sub-entity to obtain a re-divided volume mesh element; reconstruct the volume mesh element of the non-fitted entity to obtain a reconstructed volume mesh element;

[0041] Combine the re-divided volume mesh element and the reconstructed volume mesh element, determine it as the target mesh model, and specify the corresponding geological attribute for each grid.

[0042] Optionally, the use of visualization software to display the constructed urban engineering geological three-dimensional model includes:

[0043] Import geological attribute data into GOCAD visualization software, create a three-dimensional geological model according to geological attributes and structures, and use the modeling tools of the software to draw formation and fault features;

[0044] Utilize the visualization function of GOCAD software to set lighting, color, and material parameters;

[0045] Create a three-dimensional view, use a cross-sectional view to observe data at different depths, and generate an animation to display the geological structure;

[0046] Export the visualization result as an image, video, or interactive model;

[0047] Display and interact through the Web GIS platform.

[0048] Optionally, an urban engineering geological three-dimensional model construction system includes:

[0049] Sorting module: used to build an engineering geological stratigraphic sequence table based on engineering exploration borehole data and perform stratigraphic division according to geological age, geological origin, and lithofacies conditions;

[0050] Collection module: used to collect data on the factors in the engineering geological stratigraphic sequence table by using combined geophysical prospecting;

[0051] Model construction module: used to build and train an integrated learning neural network model according to the collected data;

[0052] Classification module: used to layer-by-layer identify and classify the geological attributes of each underground layer based on the prediction results of the integrated learning neural network model;

[0053] Three-dimensional reconstruction module: used to divide the space into three-dimensional grids and specify the corresponding geological attribute for each grid;

[0054] Visualization module: used to display the constructed urban engineering geological three-dimensional model using visualization software.

[0055] As can be seen from the above technical solutions, compared with the prior art, the present invention discloses a method and system for constructing a three-dimensional model of urban engineering geology, which has the following beneficial effects:

[0056] The present invention proposes a method for constructing a three-dimensional model of urban engineering geology, including: based on engineering exploration drilling data, stratigraphic division is carried out according to geological age, geological origin and lithofacies conditions to construct an engineering geology stratigraphic sequence table; using a combination of geophysical prospecting to collect data on the factors in the engineering geology stratigraphic sequence table; constructing and training an integrated learning neural network model according to the collected data; based on the prediction results of the integrated learning neural network model, layer by layer identifying and classifying the geological attributes of each underground layer; dividing the space into three-dimensional grids, and specifying corresponding geological attributes within each grid; using visualization software to display the constructed three-dimensional model of urban engineering geology. The present invention establishes an engineering geology stratigraphic sequence system, laying a foundation for the standardization work of subsequent geotechnical engineering exploration data; proposes a combined method of urban engineering geophysical prospecting suitable for assisting in engineering geology stratigraphic division, providing technical support for the application of geophysical prospecting methods in the field of urban engineering geology; establishes a three-dimensional model of urban engineering geology, providing basic data support for the construction of smart cities; based on engineering exploration drilling data, stratigraphic division is carried out according to geological age, geological origin and lithofacies conditions, and a three-dimensional model of urban engineering geology is constructed based on the engineering geology stratigraphic sequence table, improving the accuracy and authenticity of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to the provided drawings.

[0058] Figure 1 It is a schematic flow chart of a method for constructing a three-dimensional model of urban engineering geology provided by the present invention.

[0059] Figure 2 It is a structural framework diagram of a system for constructing a three-dimensional model of urban engineering geology provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0060] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0061] An embodiment of the present invention discloses a method for constructing a three-dimensional model of urban engineering geology, as Figure 1 shown, including:

[0062] Based on the engineering investigation borehole data, and stratigraphic division is carried out according to geological age, geological origin and lithofacies conditions to construct an engineering geological stratigraphic sequence table;

[0063] Using combined geophysical prospecting to collect data on the factors in the engineering geological stratigraphic sequence table;

[0064] Construct and train an integrated learning neural network model according to the collected data;

[0065] Based on the prediction results of the integrated learning neural network model, layer by layer identify and classify the geological properties of each underground layer;

[0066] Divide the space into three-dimensional grids, and specify the corresponding geological properties within each grid;

[0067] Use visualization software to display the constructed three-dimensional model of urban engineering geology.

[0068] Furthermore, the construction of the engineering geological stratigraphic sequence table includes:

[0069] Division range: Divide respectively on the plane and in depth, and divide detailed stratification and general stratification in depth;

[0070] Stratification and coding: Use a two-level coding unit of "main layer + sub-layer" to code the strata, and determine and number the engineering geological main layers according to the stratigraphic sedimentary age, sedimentary environment, sedimentary sequence and rock and soil types;

[0071] Each genetic layer and main layer have a top-down sequential relationship, and the sub-layers within each main layer are coded according to sedimentary priority and particle size.

[0072] In the specific implementation manner, the construction of the engineering geological stratigraphic sequence table specifically includes:

[0073] Division principle:

[0074] The engineering geological stratigraphic division is based on the engineering investigation borehole data collected in the region, considering the stratigraphic division and coding habits formed by local investigation and design units over the years. From the perspective of geotechnical engineering, in order to clearly reflect the unfavorable strata, main bearing strata and aquifer strata for the project, the method of mainly considering the geotechnical engineering characteristics of the strata and supplemented by their geological age, geological origin and lithofacies conditions is used for stratigraphic division. The main considerations are as follows:

[0075] (1) The classification and naming of lithology and soil, the standardization of borehole data, and the drawing of profiles shall be carried out strictly in accordance with the provisions of relevant codes such as the Code for Geotechnical Engineering Investigation (GB50021 - 20019 (2009 Edition)) and the Standard for Compiling Geotechnical Engineering Investigation Reports (CECS99:88). 100% manual data checking work has been carried out on the formation of tables such as borehole information and stratification information and the links such as inputting into the system.

[0076] (2) The division of stratigraphic ages shall be determined comprehensively based on the research results of regional stratigraphy.

[0077] (3) The geological origins of soil layers include residual, slope wash, diluvial, alluvial, siltation, aeolian, and artificial accumulation, etc.

[0078] (4) There are three basic types of lithofacies conditions: continental facies, marine facies, and marine - continental transitional facies.

[0079] (5) The main layer of the layer group should exist generally within the region. It can be determined through regional stratigraphic comparison and should follow the stratigraphic sequence law of older below and younger above.

[0080] (6) For strata with obvious differences in lithology or engineering properties within the same main layer, they can be divided into sub - layers, and the sub - layers are divided according to the order of sedimentation.

[0081] (7) For simplicity, convenience in operation, and respect for habits, a two - level coding unit of "main layer + sub - layer" is used to code the strata.

[0082] Division method:

[0083] (1) Division scope: Horizontally; Vertically, for the shallow rock and soil layers within 50m, detailed stratification is carried out, and for the rock and soil layers between 50 - 100m, only general stratification is done.

[0084] (2) Stratification and coding: According to the sedimentary age of the strata, the soil layers within 100m depth are divided into the Holocene Series (Q4), Upper Pleistocene Series (Q3), Middle Pleistocene Series (Q2), and Lower Pleistocene Series (Q1). Determine the engineering geological main layers and number them according to the sedimentary environment, sedimentary sequence, and rock and soil types, divided into a total of 22 main layers numbered 1, 2, ……, 22. Among them, layers 1 - 8 are the Holocene Series (Q4), layers 9 - 14 are the Upper Pleistocene Series (Q3), layers 15 - 20 are the Middle Pleistocene Series (Q2), and layers 21 - 22 are the Lower Pleistocene Series (Q1);

[0085] (3) Each genetic layer and main layer have a sequential relationship from top to bottom. The sub - layers within each main layer are coded according to the order of sedimentation and particle size, in the form of "main layer number + sub - layer number", represented by circled numbers plus digital subscripts, such as ①₁, ②₂, ③₃, etc., and a total of 78 stratigraphic sequences are divided.

[0086] Stratigraphic sequence division result:

[0087] According to the aforementioned division principles and methods, the shallow strata within 100 m in the study area are divided into 22 engineering geological stratum groups and 78 sequences, with the standards being: geological age, genesis, rock and soil type, lithology name, general bottom depth of the stratum and stratum thickness.

[0088] Specifically, based on the engineering investigation borehole data and according to the geological age, geological genesis and lithofacies conditions, the stratum division is carried out, and the construction of the engineering geological stratum sequence table includes:

[0089] Collect the engineering geological investigation borehole data within the study area;

[0090] Standardize the borehole data, and unify the borehole coordinates, elevations and lithology names of the strata;

[0091] Draw the stratum profile, study the regional stratum structure through regional stratum comparison, and determine the stratum sequence system and unified coding work;

[0092] Statistically analyze the physical and mechanical parameters of the strata, including density, water content, specific gravity, etc., and construct the engineering geological stratum sequence table.

[0093] Specifically, (1) Lower Pleistocene Gu'an Formation (Q1): The bottom plate depth is about 510 m, and the thickness is about 156 m. It is mainly composed of brownish-red, reddish-brown with rusty clay, and there is little silty soil and silty clay in grayish-yellow and grayish-green. It is intercalated with medium sand and fine sand layers;

[0094] (2) Middle Pleistocene Yangliuqing Formation (Q2): The bottom plate depth is about 354 m, and the thickness is about 206 m. The thickness of the lower segment is about 96 m. The lithology is mainly silty soil in grayish-yellow, grayish-green, brownish-yellow with a small amount of rust spots, and there is little silty clay and clay. The sand layers are generally distributed with many layers and large thickness, mainly fine and medium sand. The thickness of the upper segment is about 110 m. The lithology is mainly silty soil and silty clay in yellowish-gray, grayish-green, yellowish-green, gray, yellowish-brown with rust stains. The sand layers are thicker and more, generally mainly fine and medium sand, and there is little coarse-medium sand;

[0095] (3) Upper Pleistocene Ouzhuang Formation (Q3): The bottom plate depth is about 148 m, and the thickness is about 118 m. The lithology is mainly silty soil and silty clay in grayish-yellow, yellowish-gray or greenish-gray, yellowish-green gray. The sand layers are less, and mostly contain silty soil, with fine particles, mainly fine-silty sand and medium-fine sand;

[0096] (4) Holocene (Q4): The thickness is about 30 m. The lithology is mainly silty soil and silty clay in gray, grayish-green, blackish-gray and yellowish-gray, followed by clay. The sand layers only appear locally, and mostly are fine-silty sand, with uneven thickness and mostly containing silty soil.

[0097] Furthermore, the composition exploration methods include: surface microgravity method, microtremor method, high-density resistivity method, equivalent counter magnetic flux transient electromagnetic method, transient surface wave method, H / V spectral ratio method, and ground penetrating radar method.

[0098] Based on the analysis of existing data, 3 profiles were selected to carry out geophysical exploration profile measurement tests such as surface microgravity method, microtremor method, high-density resistivity method, equivalent counter magnetic flux transient electromagnetic method, transient surface wave method, H / V spectral ratio method, and ground penetrating radar method, to determine the urban engineering geophysical exploration combination methods suitable for assisting in engineering geological stratigraphic division. The urban engineering geophysical exploration combination methods were used to carry out profile measurement work in the drilling blank area to assist in engineering geological stratigraphic division and draw stratigraphic profiles.

[0099] Furthermore, it also includes preprocessing the collected geophysical exploration data, and the preprocessing includes: cleaning, denoising, and normalizing the collected geophysical exploration data;

[0100] The Z-score method was used to identify outliers in sequence, and the identified outliers were deleted;

[0101] Check whether there are duplicate data, delete the duplicate records, and retain the unique data;

[0102] Apply median filtering to the signal containing noise and retain the edge information;

[0103] Through min-max normalization, the data was scaled to the range of [0, 1].

[0104] Furthermore, the construction and training of the integrated learning neural network model based on the collected data includes:

[0105] The collected data was divided into a training set, a validation set, and a test set;

[0106] Select fully connected neural network, convolutional neural network, recurrent neural network, long short-term memory network, graph neural network, and deep belief network to form an integrated learning neural network model;

[0107] The integrated learning neural network model was trained and validated through the training set and the validation set;

[0108] The performance of the integrated learning neural network model was evaluated through the test set;

[0109] The cross-validation method was used to optimize the hyperparameters of the model and the integration strategy.

[0110] Further, it also includes: voting based on the prediction results of an ensemble learning neural network model formed by a fully connected neural network, a convolutional neural network, a recurrent neural network, a long short-term memory network, a graph neural network, and a deep belief network respectively, and selecting the category with the most votes through the voting method as the output of the ensemble learning neural network model. Among them, the prediction result of the recurrent neural network is used as the control group of the prediction result of the ensemble learning neural network model, and the prediction result of the ensemble learning neural network model is determined to be true within the preset range not exceeding the prediction result of the recurrent neural network.

[0111] Further, the step of layer-by-layer identifying and classifying the geological properties of each underground layer includes:

[0112] Determining the thickness, depth, and interlayer transition conditions of each layer according to the stratification standard;

[0113] Determining the corresponding geological properties for each layer through identification and classification;

[0114] Inputting the geological properties layer by layer in finite element analysis software to construct a three-dimensional model;

[0115] Defining the boundary conditions and interactions of each layer;

[0116] Comparing the model calculation results with the field observation data to verify the accuracy of the model.

[0117] Using three-dimensional geological modeling software, based on borehole data and constraint profiles, establishing a three-dimensional engineering geological model of the urban area by the layer-by-layer volume-forming method to achieve goals such as arbitrary sectioning and generating borehole histograms at arbitrary points.

[0118] Further, the step of dividing the space into three-dimensional grids and specifying the corresponding geological properties in each grid includes:

[0119] Establishing an initial solid model, performing mesh division and first geological property assignment on the initial solid model to obtain an initial grid model;

[0120] Performing solidification processing on the grid cells in the initial grid model to determine a preset solid model;

[0121] Performing Boolean operations on the preset solid model, obtaining a fitting solid and a non-fitting solid according to the Boolean operation results, performing rezoning and second geological property assignment on the fitting solid to obtain a re-fitting solid, performing volume mesh division on the re-fitting solid to obtain re-divided volume mesh cells; reconstructing the volume mesh cells of the non-fitting solid to obtain reconstructed volume mesh cells;

[0122] Combining the re-divided volume mesh cells and the reconstructed volume mesh cells to determine the target grid model, and corresponding geological properties in each grid.

[0123] Specifically, an initial entity model is established, and the initial entity model is meshed and assigned geological attributes for the first time to obtain an initial three-dimensional mesh;

[0124] The meshes in the initial three-dimensional mesh are solidified to obtain an entity combination corresponding one-to-one with the meshes, and based on the entity combination, a preset entity model is determined;

[0125] Boolean operations are performed on the preset entity model;

[0126] The fitted entity is re-partitioned and assigned geological attributes for the second time to obtain a re-fitted entity; the fitted entity belongs to the area affected by the Boolean operation; the fitted entity is the entity whose geological attributes are reset to zero after performing the Boolean operation on the preset entity model;

[0127] The re-fitted entity is divided into volume meshes to obtain re-divided volume mesh elements;

[0128] The volume mesh elements corresponding to the non-fitted entities are reconstructed to obtain reconstructed volume mesh elements; the non-fitted entities belong to the area not affected by the Boolean operation; the non-fitted entities are the entities whose geological attributes remain unchanged after performing the Boolean operation on the preset entity model;

[0129] The non-fitted entities are re-partitioned and assigned geological attributes for the third time to obtain re-non-fitted entities; the volume mesh elements corresponding to the re-non-fitted entities are reconstructed to obtain re-reconstructed volume mesh elements; the re-non-fitted entities belong to the area not affected by the Boolean operation; the re-non-fitted entities are the entities whose geological attributes remain unchanged after performing the Boolean operation on the preset entity model;

[0130] The combination of the re-divided volume mesh elements, the reconstructed volume mesh elements and the re-reconstructed volume mesh elements is determined as the target mesh model, and the corresponding geological attributes are assigned to each mesh.

[0131] Specifically, for re-dividing volume mesh elements: based on the preliminary analysis results, the area to be refined is selected. A division algorithm is used to generate new refined mesh elements, and the subdivision quantity is increased, such as refining to a specific mesh size target;

[0132] For reconstructing volume mesh elements: according to the actual engineering requirements or physical principles, a specific geometric shape is reconstructed to ensure that the reconstructed mesh conforms to the actual situation. A CAD software or a mesh generation tool is used to import the reconstructed geometric shape and convert it into a mesh;

[0133] For re-reconstructing volume mesh elements: on the reconstructed mesh, a specific area is further refined to further improve the resolution of the result. Ensure that the accuracy of the re-reconstructed mesh is compatible with the accuracy of other mesh elements (reconstructed and refined);

[0134] Merge the re-divided, reconstructed, and re-reconstructed grid cells to form a complete grid model. Ensure good connections between different types of grid cells. Adjust the orientation, shape, and connectivity of the grid according to the merged structure to ensure reasonable expression of the physical properties of the model.

[0135] In the specific implementation, re-partition the fitted entity and perform a second geological property assignment to obtain a re-fitted entity, including: if the fitted entity is an entity within the interlayer of the Boolean operation affected area, obtain the geological properties of the entity adjacent to the fitted entity; if the geological properties of the adjacent entity are not 0, assign the geological properties of the adjacent entity to the fitted entity.

[0136] If the fitted entity is an entity outside the interlayer of the Boolean operation affected area, obtain the control point coordinates of the fitted entity on the Boolean operation affected area to obtain the centroid of the fitted entity; search for the closest grid cell of the fitted entity in the initial grid model based on the criterion of the minimum distance from the centroid of the adjacent initial grid cell, and assign the material properties of the closest grid cell to the fitted entity. Specifically, the process of re-partitioning the non-fitted entity and performing a third geological property assignment to obtain a re-non-fitted entity is the same.

[0137] Furthermore, the use of visualization software to display the constructed 3D urban engineering geological model includes:

[0138] Import geological property data into GOCAD visualization software, create a 3D geological model based on geological properties and structures, and use the modeling tools of the software to draw formation and fault features.

[0139] Utilize the visualization function of GOCAD software to set lighting, color, and material parameters.

[0140] Create a 3D view, use a cross-sectional view to observe data at different depths, and generate an animation to display the geological structure.

[0141] Export the visualization results as images, videos, or interactive models.

[0142] Display and interact through the Web GIS platform.

[0143] In the specific implementation, it also includes: sequentially input the sorted borehole data into the Lizheng exploration system according to the stratification results to generate a Lizheng exploration database. Use the Lizheng 3D geological modeling software to directly retrieve the Lizheng exploration database, and adopt the layer-by-layer volume modeling method to establish a 3D urban engineering geological model, which adapts to the stratigraphic sedimentation law and conforms to the professional habits of engineering geology, and directly constructs 3D geological bodies layer by layer from top to bottom.

[0144] In a specific embodiment, an urban engineering geology three-dimensional model construction system, as Figure 2 shown, includes:

[0145] Sorting module: used to divide strata based on engineering investigation borehole data and according to geological age, geological origin, and lithofacies conditions, and construct an engineering geology stratigraphic sequence table;

[0146] Collection module: used to collect data on the factors in the engineering geology stratigraphic sequence table by using a combination of geophysical prospecting;

[0147] Model construction module: used to construct and train an integrated learning neural network model based on the collected data;

[0148] Classification module: used to layer-by-layer identify and classify the geological attributes of each underground layer based on the prediction results of the integrated learning neural network model;

[0149] Three-dimensional reconstruction module: used to divide the space into three-dimensional grids and specify corresponding geological attributes within each grid;

[0150] Visualization module: used to display the constructed urban engineering geology three-dimensional model using visualization software.

[0151] In this specification, each embodiment is described in a progressive manner. The key point of each embodiment is to illustrate the differences from other embodiments. The same or similar parts among the embodiments can be referred to each other. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0152] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for constructing a three-dimensional model of urban engineering geology, characterized in that, Including: Based on the engineering exploration borehole data, stratigraphic division is carried out according to geological age, geological origin and lithofacies conditions, and an engineering geological stratigraphic sequence table is constructed; Using a combination of geophysical prospecting to collect data on the factors in the engineering geological stratigraphic sequence table; Construct and train an integrated learning neural network model based on the collected data; Based on the prediction results of the integrated learning neural network model, identify and classify the geological properties of each underground layer layer by layer; Divide the space into three-dimensional grids, and specify the corresponding geological properties within each grid; The dividing the space into three-dimensional grids and specifying the corresponding geological properties within each grid includes: Establish an initial entity model, perform grid meshing and first geological property assignment on the initial entity model to obtain an initial grid model; Perform entity processing on the grid cells in the initial grid model to determine a preset entity model; Perform a Boolean operation on the preset entity model, obtain a fitted entity and a non-fitted entity according to the Boolean operation result, perform rezoning and second geological property assignment on the fitted entity to obtain a re-fitted sub-entity, perform volume grid meshing on the re-fitted sub-entity to obtain re-divided volume grid cells; reconstruct the volume grid cells of the non-fitted entity to obtain reconstructed volume grid cells; Determine the combination of the re-divided volume grid cells and the reconstructed volume grid cells as the target grid model, and specify the corresponding geological properties within each grid; Establish an initial entity model, perform grid meshing and first geological property assignment on the initial entity model to obtain an initial three-dimensional grid, including: Perform entity processing on the grids in the initial three-dimensional grid to obtain an entity combination corresponding to each grid one by one, and determine a preset entity model based on the entity combination; Perform a Boolean operation on the preset entity model; Perform rezoning and second geological property assignment on the fitted entity to obtain a re-fitted entity; the fitted entity belongs to the area affected by the Boolean operation; the fitted entity is the entity whose geological property becomes zero after performing the Boolean operation on the preset entity model; Perform volume grid meshing on the re-fitted entity to obtain re-divided volume grid cells; Reconstruct the volume grid cells corresponding to the non-fitted entity to obtain reconstructed volume grid cells; the non-fitted entity belongs to the area not affected by the Boolean operation; the non-fitted entity is the entity whose geological property remains unchanged after performing the Boolean operation on the preset entity model; Perform rezoning and third geological property assignment on the non-fitted entity to obtain a re-non-fitted entity; reconstruct the volume grid cells corresponding to the re-non-fitted entity to obtain re-reconstructed volume grid cells; the re-non-fitted entity belongs to the area not affected by the Boolean operation; the re-non-fitted entity is the entity whose geological property remains unchanged after performing the Boolean operation on the preset entity model; Determine the combination of the re-divided volume grid cells, the reconstructed volume grid cells and the re-reconstructed volume grid cells as the target grid model, and correspond the geological properties within each grid; Re-partition the fitted entity and perform a second geological property assignment to obtain a re-fitted entity, including: if the fitted entity is an entity within the interlayer of the Boolean operation affected area, obtain the geological properties of the entity adjacent to the fitted entity; if the geological properties of the adjacent entity are not 0, assign the geological properties of the adjacent entity to the fitted entity; if the fitted entity is an entity outside the interlayer of the Boolean operation affected area, obtain the control point coordinates of the fitted entity on the Boolean operation affected area to obtain the centroid of the fitted entity; search for the closest grid cell of the fitted entity in the initial grid model according to the criterion of the minimum distance from the centroid of the adjacent initial grid cell, and assign the material properties of the closest grid cell to the fitted entity. The process of re-partitioning and performing a third geological property assignment on non-fitted entities is the same as that for obtaining re-non-fitted entities; Use visualization software to display the constructed 3D urban engineering geology model.

2. The method for constructing a three-dimensional model of urban engineering geology according to claim 1, characterized in that, The constructed engineering geology stratigraphic sequence table includes: Division range: Divide in the plane and depth respectively, and divide into detailed layers and general layers in depth; Layer division and coding: Code the strata using a two-level coding unit of "main layer + sub-layer", and determine and number the engineering geology main layers according to the stratigraphic deposition age, deposition environment, deposition sequence and rock and soil types; Each genetic layer and main layer have a top-down sequential relationship, and the sub-layers within each main layer are coded according to the deposition sequence and particle size.

3. A method for constructing a three-dimensional model of urban engineering geology according to claim 1, characterized in that, The combined geophysical exploration includes: surface microgravity method, microtremor method, high-density resistivity method, equivalent anti-flux transient electromagnetic method, transient surface wave method, H / V spectral ratio method and ground penetrating radar method.

4. A method for constructing a three-dimensional model of urban engineering geology according to claim 1, characterized in that, It also includes preprocessing the collected geophysical exploration data, and the preprocessing includes: cleaning, denoising and normalizing the collected geophysical exploration data; Use the Z-score method to identify outliers in sequence, and delete the identified outliers; Check for duplicate data, delete duplicate records, and retain unique data; Apply median filtering to the signal containing noise and retain the edge information; Scale the data to the range [0, 1] through min-max normalization.

5. A method for constructing a three-dimensional model of urban engineering geology according to claim 1, characterized in that, The construction and training of the integrated learning neural network model based on the collected data include: Divide the dataset into a training set, a validation set and a test set with the collected data; Select a fully connected neural network, a convolutional neural network, a recurrent neural network, a long short-term memory network, a graph neural network and a deep belief network to form an integrated learning neural network model; Train and validate the integrated learning neural network model through the training set and the validation set; Evaluate the performance of the integrated learning neural network model through the test set; Use the cross-validation method to optimize the hyperparameters of the model and the integration strategy.

6. The method for constructing a three-dimensional model of urban engineering geology according to claim 5, characterized in that, It also includes: Votes are cast based on the prediction results of the integrated learning neural network model constructed according to the fully connected neural network, convolutional neural network, recurrent neural network, long short-term memory network, graph neural network, and deep belief network. The category with the most votes is selected through the voting method as the output of the integrated learning neural network model. Among them, the prediction result of the recurrent neural network is used as the control group of the prediction result of the integrated learning neural network model, and the prediction result of the integrated learning neural network model is determined to be true within the preset range not exceeding the prediction result of the recurrent neural network.

7. A method for constructing a three-dimensional model of urban engineering geology according to claim 1, characterized in that, The geological attributes of each underground layer identified and classified layer by layer include: Determine the thickness, depth, and interlayer transition conditions of each layer according to the layering standard; Determine the corresponding geological attributes for each layer through identification and classification; Input the geological attributes layer by layer in the finite element analysis software to construct a three-dimensional model; Define the boundary conditions and interactions of each layer; Compare the model calculation results with the field observation data to verify the accuracy of the model.

8. A method for constructing a three-dimensional model of urban engineering geology according to claim 1, characterized in that, The use of visualization software to display the constructed three-dimensional model of urban engineering geology includes: Import the geological attribute data into the GOCAD visualization software, create a three-dimensional geological model according to the geological attributes and structures, and use the modeling tools of the software to draw the formation and fault features; Utilize the visualization function of the GOCAD software to set the lighting, color, and material parameters; Create a three-dimensional view, use the cross-sectional view to observe the data at different depths, and generate an animation to display the geological structure; Export the visualization results as images, videos, or interactive models; Display and interact through the Web GIS platform.

9. A three-dimensional model construction system for urban engineering geology, characterized in that, Include: Sorting module: used to build an engineering geological stratigraphic sequence table based on the engineering exploration borehole data and stratify the strata according to the geological age, geological origin, and lithofacies conditions; Collection module: used to collect data on the factors in the engineering geological stratigraphic sequence table using combined geophysical prospecting; Model construction module: used to construct and train an integrated learning neural network model based on the collected data; Classification module: used to identify and classify the geological attributes of each underground layer layer by layer based on the prediction results of the integrated learning neural network model; Three-dimensional reconstruction module: used to divide the space into three-dimensional grids and specify the corresponding geological attributes within each grid; Visualization module: used to display the constructed three-dimensional model of urban engineering geology using visualization software.

Images