Geological survey prediction method and system based on multi-node correlation

Through the multi-node correlation geological survey method and combined with deep learning models, the problems of lack of comprehensiveness and inaccurate prediction in the existing technology are solved, and a more scientific and accurate geological assessment is achieved.

CN118628756BActive Publication Date: 2024-12-06DALIAN JIANXING CONSTR ENG CO LTD
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202411104036.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-13
Publication Date
2024-12-06
Estimated Expiration
2044-08-13

AI Technical Summary

Technical Problem

There is a lack of comprehensive analysis and prediction of geology in the prior art, a lack of prescient in evaluation, and multiple influencing factors are not effectively considered.

Method used

Using a geological survey and prediction method based on multi-node correlation, by defining the sampling point as a node, obtaining its own parameters, images and correlation parameters, performing preprocessing and feature extraction, constructing the relationship between feature quantity and surrounding rock level, and training deep learning models. Calculate the overall stability.

Benefits of technology

It improves the comprehensiveness of geological analysis and the scientificity of evaluation, reduces the computational workload, enhances the stability and accuracy of deep learning models, and achieves more accurate predictions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118628756B_ABST
    Figure CN118628756B_ABST
Patent Text Reader

Abstract

The present invention discloses a geological survey prediction method and system based on multi-node correlation, which relates to the technical field of geological prediction, and includes the following steps: extracting the characteristic quantities of the parameters of the object to be measured itself and the associated parameters of the object to be measured after the second preprocessing, and performing characteristic fusion, constructing the corresponding relationship between the first characteristic quantity, the second characteristic quantity and the third characteristic quantity and the surrounding rock grade, training a deep learning model for feedback of the surrounding rock grade, and testing the deep learning model, and calculating the comprehensive stability of the object to be measured. The present invention makes the analysis of geology more comprehensive by collecting various parameters, preprocessing the collected data, which is conducive to the subsequent calculation of the data and reduces the workload of calculation, fusing the characteristic quantities, which is conducive to the stability of the deep learning model, testing the deep learning model, improving the accuracy of the deep learning model, and enhancing the integrity and scientificity of the evaluation.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of geological prediction technology, and in particular to a geological survey prediction method and system based on multi-node correlation. Background Art

[0002] In recent years, the application of digital technology and artificial intelligence in geological prediction has gradually increased. Geological prediction technology has continued to improve in data acquisition, analysis and interpretation capabilities. Combined with the development of modern science and technology, it has become one of the important tools for resource exploration, environmental protection and disaster prevention.

[0003] At present, a method and device for predicting geological conditions of a horizontal well are disclosed in a Chinese invention patent with publication number CN 103498670 B. The method determines the position of a virtual well used for geological prediction of a horizontal well through the position of a vertical well that has been drilled in the work area where the horizontal well is located and the position of a horizontal section of the horizontal well. The reservoir profile of the virtual well is determined based on the reservoir profile of the vertical well that has been drilled. The reservoir profile of the horizontal section of the horizontal well is determined based on the reservoir profile of the virtual well. The geological prediction is performed based on the reservoir profile of the horizontal section of the horizontal well. However, in the related art, the geology is not analyzed based on multiple influencing factors, the analysis lacks comprehensiveness, the geology is not predicted, and the evaluation lacks foresight. Summary of the invention

[0004] The technical problem solved by the present invention is that in the related art, the geology is not analyzed according to multiple influencing factors, the analysis lacks comprehensiveness, the geology is not predicted, and the evaluation lacks foresight.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: a geological survey prediction method based on multi-node correlation, comprising the following steps:

[0006] Step S1, defining each sampling point as a node, and obtaining the parameters of the object to be tested, the image of the object to be tested, the associated image of the object to be tested, and the associated parameters of the object to be tested respectively;

[0007] Step S2, performing a first preprocessing on the self-image of the object to be measured and the associated image of the object to be measured, respectively extracting feature quantities of the self-image of the object to be measured and the associated image of the object to be measured after the first preprocessing, performing a second preprocessing on the self-parameters of the object to be measured and the associated parameters of the object to be measured, respectively extracting feature quantities of the self-parameters of the object to be measured and the associated parameters of the object to be measured after the second preprocessing;

[0008] Step S3, feature fusion of the first preprocessed image of the object to be measured, the feature quantity of the associated image of the object to be measured, the second preprocessed parameter of the object to be measured, and the feature quantity of the associated parameter of the object to be measured after the second preprocessing, to construct a corresponding relationship between the first feature quantity, the second feature quantity, and the third feature quantity and the surrounding rock grade;

[0009] Step S4, dividing the first characteristic quantity, the second characteristic quantity and the third characteristic quantity into a training set and a test set according to a certain ratio, training a deep learning model for feedback of the surrounding rock level, and testing the deep learning model;

[0010] Step S5, traverse each node, obtain the predicted surrounding rock level, and calculate the comprehensive stability of the object to be tested according to the predicted surrounding rock level.

[0011] As a preferred solution of the geological survey prediction method based on multi-node association described in the present invention, step S1 includes the following sub-steps:

[0012] Step S11, taking the first value as the interval distance, selecting a number of sampling points in the operation area, and defining each sampling point as a node;

[0013] Step S12, respectively obtaining the parameters of the object to be measured, the image of the object to be measured, the associated image of the object to be measured and the associated parameters of the object to be measured, wherein the image of the object to be measured includes a two-dimensional image of the core, a two-dimensional image of the tunnel face and a three-dimensional image of the tunnel face, the associated image of the object to be measured includes a two-dimensional image of the geophysical profile and a three-dimensional image of the geophysical profile, the parameters of the object to be measured include porosity, bulk density, density, water content, permeability, compressive strength, tensile strength, shear strength, Poisson's ratio and cohesion, and the associated parameters of the object to be measured include ground stress value, ground stress direction, underground water seepage, groundwater pressure, underground temperature, drilling speed, drilling pressure, drilling continuity and backflow volume.

[0014] As a preferred solution of the geological survey prediction method based on multi-node association described in the present invention, step S2 includes the following sub-steps:

[0015] Step S21, performing a second preprocessing on the parameters of the object to be tested and the parameters associated with the object to be tested, wherein the second preprocessing includes a normalization process and an outlier removal process, and the parameters of the object to be tested and the parameters associated with the object to be tested are classified within the range of 0 to 1;

[0016] Step S22, obtaining the self-image of the object to be measured and the associated image of the object to be measured, performing a first preprocessing on the self-image of the object to be measured and the associated image of the object to be measured, wherein the first preprocessing includes filtering processing and grayscale processing, extracting feature quantities of the self-image of the object to be measured after the first preprocessing, wherein the feature quantity of the two-dimensional image of the core is the rock type, the feature quantity of the two-dimensional image of the face is the rock cracks and rock microcracks, and the feature quantity of the three-dimensional image of the face is the rock fault;

[0017] Extracting the characteristic quantity of the associated image of the object to be measured after the first preprocessing, the characteristic quantity of the two-dimensional image of the geophysical exploration section is the underground temperature gradient and the geological type, and the characteristic quantity of the three-dimensional image of the geophysical exploration section is the geological scale;

[0018] Step S23, respectively extracting the feature quantities of the parameters of the object to be measured itself and the associated parameters of the object to be measured after the second preprocessing.

[0019] As a preferred solution of the geological survey prediction method based on multi-node correlation described in the present invention, wherein: the outlier elimination process calculates the first fluctuation interval of the parameters of the object to be measured and the second fluctuation interval of the parameters associated with the object to be measured respectively through the standard deviation formula, eliminates the data whose parameters of the object to be measured are not distributed in the first fluctuation interval, and eliminates the data whose parameters associated with the object to be measured are not distributed in the second fluctuation interval;

[0020] The calculation logic of the fluctuation range of the parameter of the object to be measured is as follows: calculating the average value of the parameter of the object to be measured, calculating the standard deviation of the parameter of the object to be measured, calculating the sum of the average value of the parameter of the object to be measured and the standard deviation of the parameter of the object to be measured, setting the sum of the average value of the parameter of the object to be measured and the standard deviation of the parameter of the object to be measured as a first value, calculating the difference between the average value of the parameter of the object to be measured and the standard deviation of the parameter of the object to be measured, setting the difference between the average value of the parameter of the object to be measured and the standard deviation of the parameter of the object to be measured as a second value, taking the first value as the upper limit and the second value as the lower limit, and constructing the first fluctuation range;

[0021] The calculation logic of the fluctuation range of the parameters associated with the object to be measured is: calculate the average value of the parameters associated with the object to be measured, calculate the standard deviation of the parameters associated with the object to be measured, calculate the sum of the average value of the parameters associated with the object to be measured and the standard deviation of the parameters associated with the object to be measured, set the sum of the average value of the parameters associated with the object to be measured and the standard deviation of the parameters associated with the object to be measured as a third value, calculate the difference between the average value of the parameters associated with the object to be measured and the standard deviation of the parameters associated with the object to be measured, set the difference between the average value of the parameters associated with the object to be measured and the standard deviation of the parameters associated with the object to be measured as a fourth value, take the third value as the upper limit and the fourth value as the lower limit, and construct a second fluctuation range.

[0022] As a preferred solution of the geological survey prediction method based on multi-node association described in the present invention, step S3 includes the following sub-steps:

[0023] Step S31, obtaining rock type, rock cracks, rock microcracks, rock faults, geological type, underground temperature gradient and geological scale, and obtaining porosity, bulk density, density, water content, permeability, compressive strength, tensile strength, shear strength, Poisson's ratio, cohesion, ground stress value, ground stress direction, underground water seepage, ground water pressure, underground temperature, drilling speed, drilling pressure, drilling continuity and backflow volume;

[0024] Step S32, extracting the characteristic quantities of porosity, bulk density, density, water content, permeability, compressive strength, tensile strength, shear strength, Poisson's ratio, cohesion, ground stress value, ground stress direction, underground water seepage, ground water pressure, underground temperature, drilling speed, drilling pressure, drilling continuity and backflow volume through a one-dimensional convolutional neural network, respectively, and recording them as the first characteristic quantity; extracting the characteristic quantities of rock type, rock cracks, rock microcracks, underground temperature gradient and geological type through a two-dimensional convolutional neural network, respectively, and recording them as the second characteristic quantity; extracting the characteristic quantities of rock faults and geological scale through a three-dimensional convolutional neural network, respectively, and recording them as the third characteristic quantity;

[0025] Step S33: performing feature fusion on the first feature quantity, the second feature quantity and the third feature quantity to construct a corresponding relationship between the first feature quantity, the second feature quantity and the third feature quantity and the surrounding rock grade.

[0026] As a preferred scheme of the geological survey and prediction method based on multi-node correlation described in the present invention, the construction logic of the corresponding relationship is: obtain the first characteristic quantity, the second characteristic quantity and the third characteristic quantity, construct a characteristic matrix, obtain the surrounding rock grade, construct a grade matrix, and establish a corresponding relationship between the characteristic matrix and the grade matrix.

[0027] As a preferred solution of the geological survey prediction method based on multi-node correlation described in the present invention, the calculation expression of the corresponding relationship between the characteristic matrix and the level matrix is:

[0028] ;

[0029] in, is the rank matrix, is the feature matrix, is the correspondence between the feature matrix and the level matrix.

[0030] As a preferred solution of the geological survey prediction method based on multi-node association described in the present invention, step S4 includes the following sub-steps:

[0031] Step S41, dividing the first feature quantity, the second feature quantity and the third feature quantity into a training set and a test set according to a certain ratio;

[0032] Step S42, training a deep learning model for feedback of surrounding rock grade, wherein the deep learning model takes the first characteristic quantity, the second characteristic quantity and the third characteristic quantity as input, the predicted surrounding rock grade as output, and the actual surrounding rock grade as training target, and tests the deep learning model until the accuracy reaches a first accuracy threshold.

[0033] As a preferred solution of the geological survey prediction method based on multi-node association described in the present invention, step S5 includes the following sub-steps:

[0034] Step S51, traverse each node to obtain the predicted surrounding rock level corresponding to each node;

[0035] Step S52, calculating the comprehensive stability of the object to be tested according to the predicted surrounding rock level;

[0036] The comprehensive stability is calculated according to the multivariate linear regression analysis, and the calculation expression of the comprehensive stability is:

[0037] ;

[0038] in, For the overall stability, is a constant, From the first node to the The surrounding rock level of each node.

[0039] The beneficial effects of the present invention are as follows: by collecting various parameters, the geological analysis is made more comprehensive, the collected data is preprocessed, which is beneficial to the subsequent calculation of the data and reduces the workload of calculation, constructs the correspondence between the first characteristic quantity, the second characteristic quantity and the third characteristic quantity and the surrounding rock grade, fuses the characteristic quantities, reduces the working time of the deep learning model, and is beneficial to the stability of the deep learning model, tests the deep learning model, improves the accuracy of the deep learning model, calculates the comprehensive stability of the object to be tested according to the predicted surrounding rock grade, and enhances the integrity and scientificity of the evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 A schematic diagram of the basic flow of a geological survey prediction method based on multi-node correlation provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0041] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0042] Example 1, reference Figure 1 , as an embodiment of the present invention, provides a geological survey prediction method based on multi-node correlation, comprising the following steps:

[0043] Step S1, defining each sampling point as a node, and obtaining the parameters of the object to be tested, the image of the object to be tested, the associated image of the object to be tested, and the associated parameters of the object to be tested respectively;

[0044] Step S2, performing a first preprocessing on the self-image of the object to be measured and the associated image of the object to be measured, respectively extracting feature quantities of the self-image of the object to be measured and the associated image of the object to be measured after the first preprocessing, performing a second preprocessing on the self-parameters of the object to be measured and the associated parameters of the object to be measured, respectively extracting feature quantities of the self-parameters of the object to be measured and the associated parameters of the object to be measured after the second preprocessing;

[0045] Step S3, feature fusion of the first preprocessed image of the object to be measured, the feature quantity of the associated image of the object to be measured, the second preprocessed parameter of the object to be measured, and the feature quantity of the associated parameter of the object to be measured after the second preprocessing, to construct a corresponding relationship between the first feature quantity, the second feature quantity, and the third feature quantity and the surrounding rock grade;

[0046] Step S4, dividing the first characteristic quantity, the second characteristic quantity and the third characteristic quantity into a training set and a test set according to a certain ratio, training a deep learning model for feedback of the surrounding rock level, and testing the deep learning model;

[0047] Step S5, traverse each node, obtain the predicted surrounding rock level, and calculate the comprehensive stability of the object to be tested according to the predicted surrounding rock level.

[0048] The present invention collects various parameters to make the geological analysis more comprehensive, pre-processes the collected data, which is beneficial to the subsequent calculation of the data and reduces the workload of calculation, constructs the correspondence between the first characteristic quantity, the second characteristic quantity and the third characteristic quantity and the surrounding rock grade, fuses the characteristic quantities, reduces the working time of the deep learning model, and is beneficial to the stability of the deep learning model. The deep learning model is tested to improve the accuracy of the deep learning model, and the comprehensive stability of the object to be tested is calculated according to the predicted surrounding rock grade, thereby enhancing the integrity and scientificity of the evaluation.

[0049] The step S1 includes the following sub-steps:

[0050] Step S11, taking the first value as the interval distance, selecting a number of sampling points in the operation area, and defining each sampling point as a node;

[0051] Step S12, respectively obtaining the parameters of the object to be measured, the image of the object to be measured, the associated image of the object to be measured and the associated parameters of the object to be measured, wherein the image of the object to be measured includes a two-dimensional image of the core, a two-dimensional image of the tunnel face and a three-dimensional image of the tunnel face, the associated image of the object to be measured includes a two-dimensional image of the geophysical profile and a three-dimensional image of the geophysical profile, the parameters of the object to be measured include porosity, bulk density, density, water content, permeability, compressive strength, tensile strength, shear strength, Poisson's ratio and cohesion, and the associated parameters of the object to be measured include ground stress value, ground stress direction, underground water seepage, groundwater pressure, underground temperature, drilling speed, drilling pressure, drilling continuity and backflow volume.

[0052] In the specific implementation, the parameters of the object to be measured are obtained through indoor experiments, the image of the object to be measured is obtained through face photography, the associated parameters of the object to be measured are obtained through advanced geological forecasting, and the associated image of the object to be measured is obtained through geological survey. By collecting various parameters, the analysis of geology is more comprehensive.

[0053] The step S2 includes the following sub-steps:

[0054] Step S21, performing a second preprocessing on the parameters of the object to be tested and the parameters associated with the object to be tested, wherein the second preprocessing includes a normalization process and an outlier removal process, and the parameters of the object to be tested and the parameters associated with the object to be tested are classified within the range of 0 to 1;

[0055] Step S22, obtaining the self-image of the object to be measured and the associated image of the object to be measured, performing a first preprocessing on the self-image of the object to be measured and the associated image of the object to be measured, wherein the first preprocessing includes filtering processing and grayscale processing, extracting feature quantities of the self-image of the object to be measured after the first preprocessing, wherein the feature quantity of the two-dimensional image of the core is the rock type, the feature quantity of the two-dimensional image of the face is the rock cracks and rock microcracks, and the feature quantity of the three-dimensional image of the face is the rock fault;

[0056] Extracting the characteristic quantity of the associated image of the object to be measured after the first preprocessing, the characteristic quantity of the two-dimensional image of the geophysical exploration section is the underground temperature gradient and the geological type, and the characteristic quantity of the three-dimensional image of the geophysical exploration section is the geological scale;

[0057] Step S23, respectively extracting the feature quantities of the parameters of the object to be measured itself and the associated parameters of the object to be measured after the second preprocessing.

[0058] The outlier elimination process calculates the first fluctuation interval of the measured object's own parameter and the second fluctuation interval of the measured object's associated parameter respectively through the standard deviation formula, eliminates the data whose measured object's own parameter is not distributed in the first fluctuation interval, and eliminates the data whose measured object's associated parameter is not distributed in the second fluctuation interval;

[0059] The calculation logic of the fluctuation range of the parameter of the object to be measured is as follows: calculating the average value of the parameter of the object to be measured, calculating the standard deviation of the parameter of the object to be measured, calculating the sum of the average value of the parameter of the object to be measured and the standard deviation of the parameter of the object to be measured, setting the sum of the average value of the parameter of the object to be measured and the standard deviation of the parameter of the object to be measured as a first value, calculating the difference between the average value of the parameter of the object to be measured and the standard deviation of the parameter of the object to be measured, setting the difference between the average value of the parameter of the object to be measured and the standard deviation of the parameter of the object to be measured as a second value, taking the first value as the upper limit and the second value as the lower limit, and constructing the first fluctuation range;

[0060] The calculation logic of the fluctuation range of the parameters associated with the object to be measured is: calculate the average value of the parameters associated with the object to be measured, calculate the standard deviation of the parameters associated with the object to be measured, calculate the sum of the average value of the parameters associated with the object to be measured and the standard deviation of the parameters associated with the object to be measured, set the sum of the average value of the parameters associated with the object to be measured and the standard deviation of the parameters associated with the object to be measured as a third value, calculate the difference between the average value of the parameters associated with the object to be measured and the standard deviation of the parameters associated with the object to be measured, set the difference between the average value of the parameters associated with the object to be measured and the standard deviation of the parameters associated with the object to be measured as a fourth value, take the third value as the upper limit and the fourth value as the lower limit, and construct a second fluctuation range.

[0061] In a specific implementation, the filtering process is Gaussian filtering, which pre-processes the collected data, facilitates subsequent calculations of the data, and reduces the workload of calculations.

[0062] The step S3 comprises the following sub-steps:

[0063] Step S31, obtaining rock type, rock cracks, rock microcracks, rock faults, geological type, underground temperature gradient and geological scale, and obtaining porosity, bulk density, density, water content, permeability, compressive strength, tensile strength, shear strength, Poisson's ratio, cohesion, ground stress value, ground stress direction, underground water seepage, ground water pressure, underground temperature, drilling speed, drilling pressure, drilling continuity and backflow volume;

[0064] Step S32, extracting the characteristic quantities of porosity, bulk density, density, water content, permeability, compressive strength, tensile strength, shear strength, Poisson's ratio, cohesion, ground stress value, ground stress direction, underground water seepage, ground water pressure, underground temperature, drilling speed, drilling pressure, drilling continuity and backflow volume through a one-dimensional convolutional neural network, respectively, and recording them as the first characteristic quantity; extracting the characteristic quantities of rock type, rock cracks, rock microcracks, underground temperature gradient and geological type through a two-dimensional convolutional neural network, respectively, and recording them as the second characteristic quantity; extracting the characteristic quantities of rock faults and geological scale through a three-dimensional convolutional neural network, respectively, and recording them as the third characteristic quantity;

[0065] Step S33: performing feature fusion on the first feature quantity, the second feature quantity and the third feature quantity to construct a corresponding relationship between the first feature quantity, the second feature quantity and the third feature quantity and the surrounding rock grade.

[0066] The construction logic of the corresponding relationship is: obtain the first characteristic quantity, the second characteristic quantity and the third characteristic quantity, construct a characteristic matrix, obtain the surrounding rock grade, construct a grade matrix, and establish a corresponding relationship between the characteristic matrix and the grade matrix.

[0067] The calculation expression of the corresponding relationship between the feature matrix and the level matrix is:

[0068] ;

[0069] in, is the rank matrix, is the feature matrix, is the correspondence between the feature matrix and the level matrix.

[0070] In the specific implementation, the correspondence between the first characteristic quantity, the second characteristic quantity and the third characteristic quantity and the surrounding rock grade is constructed, and the characteristic quantities are fused to reduce the working time of the deep learning model, which is beneficial to the stability of the deep learning model.

[0071] The step S4 includes the following sub-steps:

[0072] Step S41, dividing the first feature quantity, the second feature quantity and the third feature quantity into a training set and a test set according to a certain ratio;

[0073] Step S42, training a deep learning model for feedback of surrounding rock grade, wherein the deep learning model takes the first characteristic quantity, the second characteristic quantity and the third characteristic quantity as input, the predicted surrounding rock grade as output, and the actual surrounding rock grade as training target, and tests the deep learning model until the accuracy reaches a first accuracy threshold.

[0074] In a specific implementation, the deep learning model is tested to improve the accuracy of the deep learning model.

[0075] The step S5 comprises the following sub-steps:

[0076] Step S51, traverse each node to obtain the predicted surrounding rock level corresponding to each node;

[0077] Step S52, calculating the comprehensive stability of the object to be tested according to the predicted surrounding rock level;

[0078] The comprehensive stability is calculated according to the multivariate linear regression analysis, and the calculation expression of the comprehensive stability is:

[0079] ;

[0080] in, For the overall stability, is a constant, From the first node to the The surrounding rock level of each node.

[0081] In the specific implementation, the comprehensive stability of the object to be tested is calculated according to the predicted surrounding rock grade, which enhances the integrity and scientific nature of the assessment.

[0082] The present invention collects various parameters to make the geological analysis more comprehensive, pre-processes the collected data, which is beneficial to the subsequent calculation of the data and reduces the workload of calculation, constructs the correspondence between the first characteristic quantity, the second characteristic quantity and the third characteristic quantity and the surrounding rock grade, fuses the characteristic quantities, reduces the working time of the deep learning model, and is beneficial to the stability of the deep learning model. The deep learning model is tested to improve the accuracy of the deep learning model, and the comprehensive stability of the object to be tested is calculated according to the predicted surrounding rock grade, thereby enhancing the integrity and scientificity of the evaluation.

[0083] It should be understood by those skilled in the art that the embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program codes. Among them, the storage medium may be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic memory, flash memory, magnetic disk or optical disk. These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0084] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A geological survey prediction method based on multi-node correlation, characterized in that: The following steps are involved: Step S1, defining each sampling point as a node, and obtaining the parameters of the object to be tested, the image of the object to be tested, the associated image of the object to be tested, and the associated parameters of the object to be tested respectively; Step S2, performing a first preprocessing on the self-image of the object to be measured and the associated image of the object to be measured, respectively extracting feature quantities of the self-image of the object to be measured and the associated image of the object to be measured after the first preprocessing, performing a second preprocessing on the self-parameters of the object to be measured and the associated parameters of the object to be measured, respectively extracting feature quantities of the self-parameters of the object to be measured and the associated parameters of the object to be measured after the second preprocessing; Step S3, feature fusion of the first preprocessed image of the object to be measured, the feature quantity of the associated image of the object to be measured, the second preprocessed parameter of the object to be measured, and the feature quantity of the associated parameter of the object to be measured after the second preprocessing, to construct a corresponding relationship between the first feature quantity, the second feature quantity, and the third feature quantity and the surrounding rock grade; Step S4, dividing the first characteristic quantity, the second characteristic quantity and the third characteristic quantity into a training set and a test set according to a certain ratio, training a deep learning model for feedback of the surrounding rock level, and testing the deep learning model; Step S5, traversing each node, obtaining the predicted surrounding rock level, and calculating the comprehensive stability of the object to be tested according to the predicted surrounding rock level; The step S1 includes the following sub-steps: Step S11, taking the first value as the interval distance, selecting a number of sampling points in the operation area, and defining each sampling point as a node; Step S12, respectively obtaining the parameters of the object to be measured, the image of the object to be measured, the associated image of the object to be measured and the associated parameters of the object to be measured, wherein the image of the object to be measured includes a two-dimensional image of the core, a two-dimensional image of the tunnel face and a three-dimensional image of the tunnel face, the associated image of the object to be measured includes a two-dimensional image of the geophysical profile and a three-dimensional image of the geophysical profile, the parameters of the object to be measured include porosity, bulk density, density, water content, permeability, compressive strength, tensile strength, shear strength, Poisson's ratio and cohesion, and the associated parameters of the object to be measured include ground stress value, ground stress direction, underground water seepage, groundwater pressure, underground temperature, drilling speed, drilling pressure, drilling continuity and backflow volume.

2. The geological survey prediction method based on multi-node association according to claim 1, characterized in that: The step S2 includes the following sub-steps: Step S21, performing a second preprocessing on the parameters of the object to be tested and the parameters associated with the object to be tested, wherein the second preprocessing includes a normalization process and an outlier removal process, and the parameters of the object to be tested and the parameters associated with the object to be tested are classified within the range of 0 to 1; Step S22, obtaining the self-image of the object to be measured and the associated image of the object to be measured, performing a first preprocessing on the self-image of the object to be measured and the associated image of the object to be measured, wherein the first preprocessing includes filtering processing and grayscale processing, extracting feature quantities of the self-image of the object to be measured after the first preprocessing, wherein the feature quantity of the two-dimensional image of the core is the rock type, the feature quantity of the two-dimensional image of the face is the rock cracks and rock microcracks, and the feature quantity of the three-dimensional image of the face is the rock fault; Extracting the characteristic quantity of the associated image of the object to be measured after the first preprocessing, the characteristic quantity of the two-dimensional image of the geophysical exploration section is the underground temperature gradient and the geological type, and the characteristic quantity of the three-dimensional image of the geophysical exploration section is the geological scale; Step S23, respectively extracting the feature quantities of the parameters of the object to be measured itself and the associated parameters of the object to be measured after the second preprocessing.

3. The geological survey prediction method based on multi-node association according to claim 2, characterized in that: The outlier elimination process calculates the first fluctuation interval of the measured object's own parameter and the second fluctuation interval of the measured object's associated parameter respectively through the standard deviation formula, eliminates the data whose measured object's own parameter is not distributed in the first fluctuation interval, and eliminates the data whose measured object's associated parameter is not distributed in the second fluctuation interval; The calculation logic of the fluctuation range of the parameter of the object to be measured is as follows: calculating the average value of the parameter of the object to be measured, calculating the standard deviation of the parameter of the object to be measured, calculating the sum of the average value of the parameter of the object to be measured and the standard deviation of the parameter of the object to be measured, setting the sum of the average value of the parameter of the object to be measured and the standard deviation of the parameter of the object to be measured as a first value, calculating the difference between the average value of the parameter of the object to be measured and the standard deviation of the parameter of the object to be measured, setting the difference between the average value of the parameter of the object to be measured and the standard deviation of the parameter of the object to be measured as a second value, taking the first value as the upper limit and the second value as the lower limit, and constructing the first fluctuation range; The calculation logic of the fluctuation range of the parameters associated with the object to be measured is: calculate the average value of the parameters associated with the object to be measured, calculate the standard deviation of the parameters associated with the object to be measured, calculate the sum of the average value of the parameters associated with the object to be measured and the standard deviation of the parameters associated with the object to be measured, set the sum of the average value of the parameters associated with the object to be measured and the standard deviation of the parameters associated with the object to be measured as a third value, calculate the difference between the average value of the parameters associated with the object to be measured and the standard deviation of the parameters associated with the object to be measured, set the difference between the average value of the parameters associated with the object to be measured and the standard deviation of the parameters associated with the object to be measured as a fourth value, take the third value as the upper limit and the fourth value as the lower limit, and construct a second fluctuation range.

4. The geological survey prediction method based on multi-node association according to claim 1, characterized in that: The step S3 comprises the following sub-steps: Step S31, obtaining rock type, rock cracks, rock microcracks, rock faults, geological type, underground temperature gradient and geological scale, and obtaining porosity, bulk density, density, water content, permeability, compressive strength, tensile strength, shear strength, Poisson's ratio, cohesion, ground stress value, ground stress direction, underground water seepage, ground water pressure, underground temperature, drilling speed, drilling pressure, drilling continuity and backflow volume; Step S32, extracting the characteristic quantities of porosity, bulk density, density, water content, permeability, compressive strength, tensile strength, shear strength, Poisson's ratio, cohesion, ground stress value, ground stress direction, underground water seepage, ground water pressure, underground temperature, drilling speed, drilling pressure, drilling continuity and backflow volume through a one-dimensional convolutional neural network, respectively, and recording them as the first characteristic quantity; extracting the characteristic quantities of rock type, rock cracks, rock microcracks, underground temperature gradient and geological type through a two-dimensional convolutional neural network, respectively, and recording them as the second characteristic quantity; extracting the characteristic quantities of rock faults and geological scale through a three-dimensional convolutional neural network, respectively, and recording them as the third characteristic quantity; Step S33: performing feature fusion on the first feature quantity, the second feature quantity and the third feature quantity to construct a corresponding relationship between the first feature quantity, the second feature quantity and the third feature quantity and the surrounding rock grade.

5. The geological survey prediction method based on multi-node association according to claim 4, characterized in that: The construction logic of the corresponding relationship is: obtain the first characteristic quantity, the second characteristic quantity and the third characteristic quantity, construct a characteristic matrix, obtain the surrounding rock grade, construct a grade matrix, and establish a corresponding relationship between the characteristic matrix and the grade matrix.

6. The geological survey prediction method based on multi-node association according to claim 5, characterized in that: The calculation expression of the corresponding relationship between the feature matrix and the level matrix is: ; in, is the rank matrix, is the feature matrix, is the correspondence between the feature matrix and the level matrix.

7. The geological survey prediction method based on multi-node association according to claim 1, characterized in that: The step S4 includes the following sub-steps: Step S41, dividing the first feature quantity, the second feature quantity and the third feature quantity into a training set and a test set according to a certain ratio; Step S42, training a deep learning model for feedback of surrounding rock grade, wherein the deep learning model takes the first characteristic quantity, the second characteristic quantity and the third characteristic quantity as input, the predicted surrounding rock grade as output, and the actual surrounding rock grade as training target, and tests the deep learning model until the accuracy reaches a first accuracy threshold.

8. The geological survey prediction method based on multi-node association according to claim 1, characterized in that: The step S5 comprises the following sub-steps: Step S51, traverse each node to obtain the predicted surrounding rock level corresponding to each node; Step S52, calculating the comprehensive stability of the object to be tested according to the predicted surrounding rock level; The comprehensive stability is calculated according to the multivariate linear regression analysis, and the calculation expression of the comprehensive stability is: ; in, For the overall stability, is a constant, From the first node to the The surrounding rock level of each node.

Citation Information

Patent Citations

  • Horizontal Well Geological Prediction Method and Device

    CN103498670B

  • An identification method and device for tunnel surrounding rock classification

    CN109886534A

  • TBM construction tunnel surrounding rock intelligent automatic grading method and system

    CN116758343A