Method and system for reconstructing tunnel engineering geology by using computer vision technology based on drilling data

Abstract

The invention discloses a method and system for reconstructing tunnel engineering geology by using a computer vision technology based on drilling data, and the method comprises the steps: carrying out the counting and numbering of all stratums exposed in a drill hole, and obtaining a drill hole database; according to the drilling database, determining a reconstruction task by utilizing a maximum information principle; wherein the reconstruction task comprises a task type and a reconstruction position; according to the reconstruction task, performing stratigraphic distribution reconstruction of the geological cross section by using a cross section reconstruction model based on drilling data to obtain geological distribution information of the key cross section; according to the geological distribution information, utilizing a middle section reconstruction model based on adjacent sections to reconstruct a middle section to obtain a geological section sequence; and processing the geological section sequence to obtain a three-dimensional geological model. The method has remarkable advantages in the aspects of processing high-dimensional data, extracting spatial features, reducing calculation complexity and generating image output, calculation time cost can be greatly reduced, and a three-dimensional geologic model of point cloud or voxels can be output.

Description

Technical Field

[0001] The present invention belongs to the technical field of three-dimensional geological reconstruction, and in particular relates to a method for reconstructing tunnel engineering geology based on drilling data by using computer vision technology. Background Art

[0002] In tunnel engineering, geological conditions are the key factor to ensure safe construction. Traditional two-dimensional geological cross-sections can no longer meet the requirements of modern engineering for the accuracy of geological information. Therefore, it is particularly important to establish a three-dimensional geological model.

[0003] At present, geological reconstruction mainly relies on two methods: model interpretation based on geophysical exploration technology and model interpretation based on drilling technology. Geophysical methods obtain stratigraphic information through physical detection technologies such as seismic waves and electromagnetic waves, while drilling technology obtains geological data through direct drilling. Although drilling technology can provide accurate geological information, due to cost constraints, drilling points are usually sparse, which requires effective geological modeling methods to infer stratigraphic distribution. Existing modeling methods include statistical kriging, probabilistic Markov random fields and Markov chains, and multi-point statistical methods. These methods have limitations in dealing with complex geological structures and simulating unknown stratigraphic changes.

[0004] Existing geological modeling methods, such as Kriging, are subject to the selection of variograms, are highly subjective, and are difficult to accurately simulate complex geological structures. Probabilistic methods rely on the accurate determination of transition probability matrices and stratigraphic directions, and although multi-point statistical methods use training images to simulate geological spatial structures, they cannot handle local stratigraphic changes that do not appear in the training images, and have high computational costs. Although machine learning methods perform well in feature mining and prediction, the computational cost is still high when combined with traditional methods, and pure machine learning methods are limited by the input and output formats of neural networks, requiring additional post-processing steps to generate intuitive geological models. Summary of the invention

[0005] In order to solve the above technical problems, the present invention proposes a method and system for reconstructing tunnel engineering geology based on drilling data using computer vision technology, and develops a new method that can reduce computing costs and directly output intuitive geological models, which is crucial to improving the accuracy and efficiency of geological reconstruction.

[0006] To achieve the above objectives, in a first aspect, the present invention provides a method for reconstructing tunnel engineering geology based on drilling data using computer vision technology, comprising:

[0007] Count and number all the strata exposed in the borehole to obtain a borehole database;

[0008] Determine a reconstruction task according to the drilling database using the maximum information principle; wherein the reconstruction task includes a task type and a reconstruction location;

[0009] According to the reconstruction task, the stratigraphic distribution of the geological section is reconstructed using a section reconstruction model based on the drilling data to obtain geological distribution information of the key section; according to the geological distribution information, the intermediate section is reconstructed using an intermediate section reconstruction model based on adjacent sections to obtain a geological section sequence;

[0010] The geological section sequence is processed to obtain a three-dimensional geological model.

[0011] Preferably, the drilling database contains x-coordinates, y-coordinates, formation top elevations, formation bottom elevations and formation numbers.

[0012] Preferably, reconstructing the stratum distribution of the geological section using the section reconstruction model based on the drilling data includes:

[0013] By using the image restoration method, the known borehole information is regarded as the real data of the image, and the unknown geological information between the boreholes is regarded as the damaged part of the image, so as to obtain the initial model of the geological section.

[0014] According to the initial model, the cross-section is reconstructed by using an image restoration neural network model based on an adversarial neural network to obtain geological distribution information of the key cross-section.

[0015] Preferably, the image restoration neural network model includes a generation network and a discrimination network.

[0016] Preferably, reconstructing the intermediate section using the intermediate section reconstruction model based on adjacent sections comprises:

[0017] According to the geological distribution information, a stratigraphic distribution characteristic data set required for model training is obtained;

[0018] According to the stratum distribution feature data set, the modified variational autoencoder structure model is used for learning to obtain an interpolation model;

[0019] According to the interpolation model, two reconstructed continuous adjacent sections are input to obtain the reconstructed section information at the middle position, thereby completing the geological reconstruction of the middle section.

[0020] In a second aspect, the present invention further provides a system for reconstructing tunnel engineering geology based on drilling data using computer vision technology, comprising:

[0021] A data acquisition module is used to count and number all the strata exposed in the borehole to obtain a borehole database;

[0022] A reconstruction task confirmation module, used to determine the reconstruction task according to the drilling database using the maximum information principle; wherein the reconstruction task includes a task type and a reconstruction location;

[0023] A cross-section reconstruction module is used to reconstruct the stratigraphic distribution of the geological section according to the reconstruction task using a cross-section reconstruction model based on the drilling data to obtain geological distribution information of the key section; and to reconstruct the intermediate section according to the geological distribution information using an intermediate section reconstruction model based on adjacent sections to obtain a geological section sequence;

[0024] The model building module processes the geological section sequence to obtain a three-dimensional geological model.

[0025] In a third aspect, the present invention further discloses a computer device, comprising a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method described in the first aspect.

[0026] In a fourth aspect, the present invention further discloses a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method described in the first aspect when executed by a processor.

[0027] In a fifth aspect, the present invention further discloses a computer program product, including a computer program, which implements the steps of the method described in the first aspect when executed by a processor.

[0028] Compared with the prior art, the present invention has the following advantages and technical effects:

[0029] The present invention provides a method for reconstructing tunnel engineering geology based on borehole data using computer vision technology. First, all strata exposed in the borehole are counted and numbered to obtain a borehole database; secondly, a reconstruction task is determined based on the borehole database using the maximum information principle; wherein the reconstruction task includes a task type and a reconstruction position; then, based on the reconstruction task, a section reconstruction model based on the borehole data is used to reconstruct the stratum distribution of the geological section to obtain geological distribution information of the key section; based on the geological distribution information, an intermediate section is reconstructed using an intermediate section reconstruction model based on adjacent sections to obtain a geological section sequence; finally, the geological section sequence is processed to obtain a three-dimensional geological model.

[0030] The present invention is based on the framework process of intelligent reconstruction of geological information based on the idea of ​​image restoration and video interpolation. First, the image restoration idea is used to reconstruct the borehole section, and the reconstructed section is regarded as a frame image in the video. The video interpolation idea is used to reconstruct the intermediate section to form a continuous parallel geological reconstruction section that meets the required accuracy. The present invention has significant advantages in processing high-dimensional data, extracting spatial features, reducing computational complexity and generating image outputs. It can greatly reduce the computational time cost and output a three-dimensional geological model of point clouds or voxels. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0032] Figure 1 A flow chart of a method for reconstructing tunnel engineering geology according to an embodiment of the present invention;

[0033] Figure 2 A schematic diagram of a reconstruction framework based on drilling data according to an embodiment of the present invention;

[0034] Figure 3 A schematic diagram of a cross-sectional reconstruction neural network model according to an embodiment of the present invention;

[0035] Figure 4 It is a schematic diagram of a model reconstructed from an intermediate section based on adjacent sections according to an embodiment of the present invention. DETAILED DESCRIPTION

[0036] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0037] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0038] First, the technical terms involved in the following embodiments are explained.

[0039] Computer vision technology: Computer vision refers to constructing a clear and meaningful description of the objective objects in the image, calculating the characteristics of the three-dimensional world from one or more digital images, and making useful decisions about the objective objects and scenes based on the perceived images. Its specific tasks include: image classification, object detection, semantic segmentation, image restoration, etc. The specific tasks involved in this patent are image restoration and video interpolation.

[0040] Borehole data: Borehole is a geological survey method used in the exploration phase of tunnel engineering to obtain geological information of the construction area. Soil samples are obtained through drilling to obtain detailed distribution of strata within the drilling range. The borehole data used in this patent is processed image data.

[0041] Geological reconstruction: is the process of using geological, geophysical, drilling and other earth science data to create three-dimensional models of geological structures.

[0042] The first law of geography: Everything is related to everything else, but things that are close are more related than things that are far away. In the field of geology, it can be seen that geographic information is spatially autocorrelated, and the closer the objects are, the more related their information is.

[0043] For the geological reconstruction problem of tunnel engineering, reconstruction tasks can be divided into two types. The first type is to reconstruct the key section based on the borehole data (called the repair task), and the second type is to reconstruct the intermediate section based on the adjacent sections (called the interpolation task). Based on the two tasks, the local geological reconstruction method mainly includes a reconstruction framework and two reconstruction models. The two reconstruction models are the section reconstruction model based on the borehole data (repair model) and the intermediate section reconstruction model based on the adjacent sections (interpolation model).

[0044] Embodiment 1

[0045] like Figure 1 As shown, this embodiment provides a method for reconstructing tunnel engineering geology based on drilling data using computer vision technology, including:

[0046] Step 1: Obtain the borehole database. First, count and number all the strata exposed in the borehole, and then process each stratum into a borehole database containing x-coordinates, y-coordinates, stratum top elevation, stratum bottom elevation, and stratum number.

[0047] Step 2: Reconstruction decision. In this framework process, we reconstruct the key cross section in stage 1 based on the exploration hole information. Then, in stages 2 and 3, the reconstruction results of the key cross section in stage 1 are used to reconstruct the intermediate cross section. Therefore, the reconstruction accuracy of the key cross section will affect the overall reconstruction result. For stage 1, the more known information in the reconstructed cross section (the more exploration holes), the less reconstruction information is required, and the more accurate the reconstruction result. Therefore, the maximum information criterion should be met when selecting the reconstruction cross section in stage 1, as shown in the following formula.

[0048]

[0049] Where x and y represent the direction of the cross section to be generated, DP x Indicates the number of holes to be drilled on the x-direction cross section, DP yIndicates the number of holes to be drilled on the y-direction cross section. x and DP y If the values ​​of are equal, the direction is determined randomly.

[0050] Step 3: Section reconstruction. After the reconstruction direction is determined, the section reconstruction model based on the drilling data is used to reconstruct the section based on the exploration hole layout. After the reconstruction is completed, enter stage 2, using the reconstructed continuous parallel key sections, input the adjacent sections into the interpolation model to reconstruct the intermediate sections. In stage 3, repeat the process to reconstruct until the required reconstruction accuracy is achieved.

[0051] As an additional implementation method, the step specifically includes:

[0052] (a) Cross-section reconstruction model based on drilling data;

[0053] During the geological survey of tunnel engineering, the specific information of multiple boreholes will be obtained. These boreholes are distributed on the cross sections in certain directions. These cross sections are analogized to images. The known borehole information is the real data of the image, and the unknown geological information between the boreholes can be regarded as the damaged part of the image. At this time, a cross-section reconstruction problem is transformed into an image restoration problem in the field of computer vision. Based on this, an improved image restoration neural network model based on adversarial neural network is proposed. The model consists of a generative network and a discriminative network. In the generative network, the required reconstruction data and mask are input, and the restoration image of the cross section is output. In the discriminative network, the restoration image and the real image output by the generative network are input to discriminate whether the restoration image and the real image are real or fake, so as to prompt the generative network to generate more real restoration results that are difficult to be recognized by the discriminative network.

[0054] (ii) Reconstruction of the model based on the intermediate sections of adjacent sections (interpolation model);

[0055] Based on the first law of geography, the properties of objects that are close in space are related. Therefore, it can be considered that when the distance is close, the stratigraphic distribution characteristics on the two parallel sections are similar to those in the middle section, and the stratigraphic information of the middle section can be inferred by learning the stratigraphic distribution characteristics of the front and rear sections. The interpolation model learns the stratigraphic distribution information of continuous and parallel reconstructed sections generated by the cross-sectional reconstruction model based on drilling data, and predicts the stratigraphic distribution information at the middle position of two adjacent sections based on the learning results. The model is modified based on the variational autoencoder structure, with the input being two reconstructed continuous adjacent sections and the output being the reconstructed section information at their middle position.

[0056] This embodiment innovatively introduces the idea of ​​image restoration and interpolation to solve the task of geological reconstruction.

[0057] Step 4: Process the cross-sections of a large amount of continuous image data obtained through reconstruction to form the required voxel or point cloud three-dimensional geological model.

[0058] Beneficial effects of this embodiment:

[0059] This embodiment is aimed at the process of reconstructing geological three-dimensional models. The continuous two-dimensional sections of a three-dimensional geological model in a certain direction are regarded as continuous two-dimensional images. In accordance with the first law of geography, geological modeling is performed based on computer vision technology of image restoration and video interpolation. The powerful deep feature mining, learning and reasoning capabilities of neural networks are used to learn the known stratum distribution characteristics and reconstruct the sections. At the same time, this embodiment relies on the inherent characteristics of convolutional neural networks, and its output result is visualized image information. Compared with other existing technologies, this embodiment can greatly reduce the computing time cost and directly output visualized geological modeling results.

[0060] Embodiment 2

[0061] Based on the same inventive concept, this embodiment also provides a system for reconstructing tunnel engineering geology based on drilling data using computer vision technology, including:

[0062] A data acquisition module is used to count and number all the strata exposed in the borehole to obtain a borehole database;

[0063] A reconstruction task confirmation module, used to determine the reconstruction task according to the drilling database using the maximum information principle; wherein the reconstruction task includes a task type and a reconstruction location;

[0064] A cross-section reconstruction module is used to reconstruct the stratigraphic distribution of the geological section according to the reconstruction task using a cross-section reconstruction model based on the drilling data to obtain geological distribution information of the key section; and to reconstruct the intermediate section according to the geological distribution information using an intermediate section reconstruction model based on adjacent sections to obtain a geological section sequence;

[0065] The model building module processes the geological section sequence to obtain a three-dimensional geological model.

[0066] The system for reconstructing tunnel engineering geology based on drilling data using computer vision technology provided in this embodiment has all the advantages of the method for reconstructing tunnel engineering geology based on drilling data using computer vision technology provided in the first embodiment.

[0067] Embodiment 3

[0068] This embodiment further discloses a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the method described in the first embodiment.

[0069] Embodiment 4

[0070] This embodiment further discloses a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described in the first embodiment are implemented.

[0071] Embodiment 5

[0072] This embodiment also discloses a computer program product, including a computer program, which implements the steps of the method described in the first embodiment when executed by a processor.

[0073] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for reconstructing tunnel engineering geology based on drilling data using computer vision technology, characterized in that: The following steps are involved: Count and number all the strata exposed in the borehole to obtain a borehole database; Determine a reconstruction task according to the drilling database using the maximum information principle; wherein the reconstruction task includes a task type and a reconstruction location; According to the reconstruction task, the stratigraphic distribution of the geological section is reconstructed using a section reconstruction model based on the drilling data to obtain geological distribution information of the key section; according to the geological distribution information, the intermediate section is reconstructed using an intermediate section reconstruction model based on adjacent sections to obtain a geological section sequence; The geological section sequence is processed to obtain a three-dimensional geological model.

2. The method according to claim 1, characterized in that: The drilling database includes an x-coordinate, a y-coordinate, a formation top elevation, a formation bottom elevation and a formation number.

3. The method according to claim 1, characterized in that: Reconstructing the stratigraphic distribution of geological sections using a cross-section reconstruction model based on borehole data includes: By using the image restoration method, the known borehole information is regarded as the real data of the image, and the unknown geological information between the boreholes is regarded as the damaged part of the image, so as to obtain the initial model of the geological section. According to the initial model, the cross-section is reconstructed by using an image restoration neural network model based on an adversarial neural network to obtain geological distribution information of the key cross-section.

4. The method according to claim 3, characterized in that The image restoration neural network model includes a generation network and a discrimination network.

5. The method according to claim 1, characterized in that Reconstructing the intermediate section using the intermediate section reconstruction model based on adjacent sections includes: According to the geological distribution information, a stratigraphic distribution characteristic data set required for model training is obtained; According to the stratum distribution feature data set, the modified variational autoencoder structure model is used for learning to obtain an interpolation model; According to the interpolation model, two reconstructed continuous adjacent sections are input to obtain the reconstructed section information at the middle position, thereby completing the geological reconstruction of the middle section.

6. A system for reconstructing tunnel engineering geology based on drilling data using computer vision technology, characterized in that: include: A data acquisition module is used to count and number all the strata exposed in the borehole to obtain a borehole database; A reconstruction task confirmation module, used to determine the reconstruction task according to the drilling database using the maximum information principle; wherein the reconstruction task includes a task type and a reconstruction location; A cross-section reconstruction module is used to reconstruct the stratigraphic distribution of the geological section using a cross-section reconstruction model based on the drilling data according to the reconstruction task, so as to obtain geological distribution information of the key section; and to reconstruct the intermediate section using an intermediate section reconstruction model based on adjacent sections according to the geological distribution information, so as to obtain a geological section sequence; The model building module processes the geological section sequence to obtain a three-dimensional geological model.

7. A computer device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method of claims 1-5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to claims 1 to 5 are implemented.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to claims 1 to 5 are implemented.

Images