Refined In-situ Grading and Evaluation Methods for Engineering Rock Mass During Drilling

Abstract

This invention discloses a method for refined in-situ classification and evaluation of engineering rock masses during drilling, relating to the field of geotechnical engineering exploration technology. The method includes: obtaining drilling parameters during the drilling process through digital drilling tests; constructing a drilling inversion model for rock mass compressive strength, a drilling prediction model for rock mass compressive strength, and a drilling identification model for structural surface parameters based on the drilling parameters, thereby obtaining the inverted rock mass compressive strength, predicted rock mass compressive strength, and structural surface parameters, respectively; calculating the equivalent compressive strength of the rock mass based on the inverted and predicted rock mass compressive strength, and simultaneously calculating the number of volume joints in the rock mass based on the structural surface parameters to obtain the rock mass integrity coefficient; inputting the equivalent compressive strength and rock mass integrity coefficient into the refined classification model during drilling to obtain the classification and evaluation results. This invention enables rapid in-situ classification during construction, which is of great significance for improving the safety of underground engineering.

Description

Technical Field

[0001] This invention relates to the field of geotechnical engineering technology, specifically to a method for refined in-situ classification and evaluation of engineering rock masses during drilling. Background Technology

[0002] Rock mass classification is a systematic method that classifies surrounding rock into different quality grades based on factors such as the physical and mechanical properties, structural characteristics, and engineering environment of the rock mass, using quantitative or qualitative indicators. Its purpose is to assess rock mass stability and guide engineering design and construction decisions. In the construction of underground engineering projects such as mine roadways, tunnels, and subways, the accuracy of rock mass classification directly affects the safety, economy, and construction efficiency of support design. Traditional rock mass classification methods mainly rely on core drilling and laboratory testing, which are not only time-consuming but also difficult to implement real-time dynamic assessment during construction.

[0003] Existing rock grading mainly relies on manual experience and field surveys, which suffers from strong subjectivity and low accuracy, making it difficult to meet the needs of refined construction under complex geological conditions. With the development of tunnel excavation technology, a large number of drilling face parameters and digital images can be acquired in real time. By using artificial intelligence technologies such as deep learning and machine learning, the rock grading level can be intelligently identified, avoiding human error and enabling more efficient processing and analysis of complex construction environment data. However, existing rock grading technologies still have the following limitations: (1) Traditional methods require rock mass parameters to be obtained through core samples and laboratory tests, resulting in a long data acquisition cycle, which cannot meet the needs of mechanized construction for real-time grading and affects the progress of the project. (2) Traditional drilling parameter inversion models usually require a long calculation time, which cannot meet the needs of real-time rapid identification, resulting in low grading efficiency. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention provides a method for refined in-situ classification and evaluation of engineering rock masses during drilling, which enables rapid in-situ classification during construction and is of great significance for improving the safety of underground engineering.

[0005] The technical solution of the present invention is as follows: In a first aspect of the present invention, a method for refined in-situ classification and evaluation of engineering rock masses during drilling is provided, comprising the following steps: The drilling parameters during the drilling process are obtained through digital drilling tests. These parameters include drilling torque, drilling pressure, drilling speed, and drill bit rotation speed. Based on the drilling parameters, a drilling inversion model for rock mass compressive strength, a drilling prediction model for rock mass compressive strength, and a drilling identification model for structural surface parameters are constructed respectively, thereby obtaining the drilling inversion rock mass compressive strength, the predicted rock mass compressive strength, and the structural surface parameters respectively; The equivalent compressive strength of the rock mass is calculated based on the inverted compressive strength and the predicted compressive strength of the rock mass during drilling. At the same time, the volume joint number of the rock mass is calculated based on the structural parameters, and the rock mass integrity coefficient is obtained. The equivalent compressive strength and rock mass integrity coefficient are input into the drilling-while-drilling refined classification model to obtain the drilling-while-drilling refined in-situ classification and evaluation results of the engineering rock mass.

[0006] In some embodiments of the present invention, the established rock mass compressive strength inversion model during drilling is as follows:

[0007]

[0008] in, s c The compressive strength of the rock mass; or e The cutting energy density; R 1 represents the cutting edge radius; N This refers to the drill bit rotation speed; V This refers to the drilling speed; M This refers to the drilling torque. F Drilling pressure; R The radius of the drill bit; l The cutting edge length; α The angle between the cutting edge and the horizontal direction. c The cutting angle, m The coefficient of friction, a、b It is a constant.

[0009] In some embodiments of the present invention, the rock mass compressive strength prediction model while drilling adopts a GA-BP neural network prediction model, which is obtained by constructing a BP neural network model and optimizing it using a GA genetic algorithm. The rock mass compressive strength prediction model while drilling uses preprocessed drilling parameters as input data and predicted rock mass compressive strength as output data. In some embodiments of the present invention, the equivalent compressive strength of the rock mass is obtained by calculating the average of the rock mass compressive strength obtained by inversion during drilling and the predicted rock mass compressive strength.

[0010] In some embodiments of the present invention, the structure surface parameter identification model while drilling adopts a 1D-CNN model. The drilling parameters are preprocessed and used as input data, while the structure surface parameters are used as output data. Through structure surface parameter prediction, the structure surfaces are grouped, with identical structure surface features grouped together, and the total number of groups on the structure surface measurement lines is counted. N i Ungrouped joint number N Calculate the number of rock mass grouped joints per meter along the normal direction. S iNumber of non-grouped joints per cubic meter of rock mass S 0:

[0011]

[0012] in: S i This represents the number of structural surfaces per meter along the normal direction in the i-th group; N i Let be the total number of groups on the measurement lines of the i-th structural surface; The length of the survey line along the normal direction; S 0 represents the number of non-grouped joints per cubic meter of rock mass; N To measure the number of non-grouped joints on the line; L This represents the length of the measuring line.

[0013] In some embodiments of the present invention, the number of rock mass joints is... J v The calculation formula is as follows:

[0014] in: J v The number of volumetric joints in the rock mass, in units per m. 3 m represents the number of structural surface groups within the statistical region; Furthermore, according to the volumetric joint number of rock mass in the "Engineering Rock Mass Classification Standard" J v and integrity coefficient K v The correspondence was used to derive the integrity coefficient. K v .

[0015] In some embodiments of the present invention, the drilling-while-drilling refined classification model is as follows:

[0016] in, BQ These are the basic quality indicators of the rock mass; K v This is the rock mass integrity coefficient; s c The compressive strength of the rock mass; e、g、s All are constants.

[0017] In some embodiments of the present invention, by setting multiple critical values, the rock mass to be evaluated is divided into multiple grades based on the comparison relationship between the basic quality indicators of the rock mass and the critical values. The classification rules for rock mass grades are as follows: If the basic quality index of the rock mass exceeds the first critical value A1, it is judged as Grade I; If the basic quality index of the rock mass is greater than the second critical value A2 but does not exceed the first critical value A1, it is judged as Grade II; If the basic quality index of the rock mass is greater than the third critical value A3 but does not exceed the second critical value A2, it is judged as Grade III; If the basic quality index of the rock mass is greater than the fourth critical value A4 but does not exceed the third critical value A3, it is judged as Grade IV; If the basic quality index of the rock mass is greater than the fourth critical value A4 but does not exceed the fifth critical value A5, it is judged as Grade V; Among them, the first critical value A1, the second critical value A2, the third critical value A3, the fourth critical value A4, and the fifth critical value A5 decrease sequentially in descending order.

[0018] In some embodiments of the present invention, by setting thresholds, the rock mass that has been preliminarily evaluated is further refined based on the comparison relationship between the basic quality indicators of the rock mass and these critical values. A first threshold B1 is inserted between the second critical value A2 and the third critical value A3. When the basic quality index of the rock mass is between A2 and B1, it is judged as Grade III1; when the basic quality index of the rock mass is between B1 and A3, it is judged as Grade III2. A second threshold B2 and a third threshold B3 are inserted between the third critical value A3 and the fourth critical value A4. When the basic quality index of the rock mass is between A3 and B2, it is judged as level IV1; when the basic quality index of the rock mass is between B2 and B3, it is judged as level IV2; when the basic quality index of the rock mass is between B3 and A4, it is judged as level IV3. A fourth threshold B4 is inserted between the fourth critical value A4 and the fifth critical value A5. When the basic quality index of the rock mass is between A4 and B4, it is judged as Grade V1; when the basic quality index of the rock mass is between B4 and A5, it is judged as Grade V2. Among them, the first threshold B1, the second threshold B2, the third threshold B3 and the fourth threshold B4 decrease in descending order.

[0019] In some embodiments of the present invention, a three-dimensional model of rock mass grade information is constructed based on the results of refined in-situ classification and evaluation of engineering rock mass during drilling, specifically including: First, n advance boreholes are drilled into the engineering rock mass and numbered sequentially as ZK1, ZK2, ZK3, ..., ZK n ; Secondly, based on the drilling parameters, the rock mass parameters of each borehole are obtained, and the rock mass within each borehole depth is finely classified. Then, the grading results are used as spatial point data, and three-dimensional spatial interpolation is performed using geostatistical interpolation algorithms to generate a continuous rock mass grade distribution surface. Finally, the distribution surfaces of the same depth are connected to form a three-dimensional space volume at the rock mass level, thus constructing a three-dimensional model of rock mass level information.

[0020] One or more technical solutions of the present invention have the following beneficial effects: (1) The method for refined in-situ classification and evaluation of engineering rock mass during drilling proposed in this invention can obtain the compressive strength and structural features of rock mass in real time during the digital drilling test by establishing a rock mass compressive strength inversion model, a rock mass compressive strength prediction model, and a structural surface parameter identification model. This avoids the lag of traditional laboratory testing, reduces human measurement errors through automated data acquisition and analysis, and improves the objectivity of the results.

[0021] (2) When obtaining the compressive strength of the rock mass, the present invention uses two models: the rock mass compressive strength inversion model while drilling and the rock mass compressive strength prediction model while drilling. Specifically, the average value of the rock mass compressive strength obtained while drilling and the predicted rock mass compressive strength is taken as the compressive strength of the rock mass. The rock mass compressive strength obtained while drilling is based on the regression analysis method, while the predicted rock mass compressive strength is obtained based on the machine learning method. The average value of the strengths obtained by the two different principle models is taken to make the value of the rock mass compressive strength more accurate, which is conducive to improving the accuracy of the fine classification of the surrounding rock while drilling.

[0022] (3) This invention proposes a rock mass compressive strength prediction model based on GA-BP neural network, which can quickly predict the rock mass compressive strength according to the drilling parameters, and further improve the prediction accuracy of rock compressive strength.

[0023] This invention proposes a 1D-CNN-based structure parameter identification model that can automatically identify structure parameters, effectively extract subtle features of torque curves, improve the accuracy and reliability of joint identification, and reduce the subjectivity of manual interpretation.

[0024] (4) When obtaining the number of volume joints in rock mass, the present invention first groups the structural surfaces according to the structural surface parameters, and then calculates the number of grouped joints per meter along the normal direction and the number of ungrouped joints per cubic meter of rock mass according to the total number of grouped joints and the number of ungrouped joints on the structural surface measuring line. Finally, the calculation results of multiple groups are summed to obtain the number of volume joints in rock mass. This method is applicable to various types of rock mass, including rock mass with poor joint development, moderate joint development, developed joint development and very developed joint development, and can accurately reflect the joint distribution of rock mass.

[0025] (5) Based on the rock mass compressive strength and rock mass integrity coefficient, the present invention obtains the basic quality index BQ value of the rock mass. By establishing a surrounding rock classification model while drilling, the surrounding rock is classified and evaluated in a refined manner while drilling. Based on the results of the refined in-situ classification and evaluation of the engineering rock mass while drilling, a three-dimensional model of rock mass level information is constructed. Combined with digital drilling, deep learning and standard specifications, the intelligent acquisition and analysis of rock mass parameters can be realized. It can be used in complex geological conditions. The three-dimensional model of rock mass level information can be adjusted in real time during the drilling process to enhance the applicability of the project. Attached Figure Description

[0026] Figure 1 This is a flowchart of the method for refined in-situ classification and evaluation of engineering rock masses during drilling, as described in this invention. Detailed Implementation

[0027] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0028] Example 1 In a typical embodiment of the present invention, a method for refined in-situ classification and evaluation of engineering rock masses during drilling is proposed, such as... Figure 1 As shown, it includes the following steps: Step 1: Obtain drilling parameters during the drilling process through digital drilling tests; Step 2: Based on the drilling parameters, construct the drilling inversion model of rock mass compressive strength, the drilling prediction model of rock mass compressive strength, and the drilling identification model of structural surface parameters, and then obtain the drilling inversion rock mass compressive strength, the predicted rock mass compressive strength, and the structural surface parameters, respectively. Step 3: Calculate the equivalent compressive strength of the rock mass based on the inverted rock mass compressive strength and the predicted rock mass compressive strength. At the same time, calculate the volume joint number of the rock mass based on the structural surface parameters and obtain the rock mass integrity coefficient. Step 4: Input the equivalent compressive strength and rock mass integrity coefficient into the drilling fine classification model to obtain the drilling fine in-situ classification and evaluation results of the engineering rock mass.

[0029] Each step is explained in detail below: Step 1: Obtain drilling parameters during the drilling process through digital drilling tests.

[0030] Specifically, the drilling parameters include drilling torque, drilling pressure, drilling speed, and drill bit rotation speed.

[0031] Step 2: Based on the drilling parameters, construct the drilling inversion model of rock mass compressive strength, the drilling prediction model of rock mass compressive strength, and the drilling identification model of structural surface parameters, and then obtain the drilling inversion rock mass compressive strength, the predicted rock mass compressive strength, and the number of rock mass volume joints, respectively.

[0032] Furthermore, the established inversion model for the compressive strength of the rock mass during drilling is as follows:

[0033]

[0034] in, s c The compressive strength of the rock mass; or e The cutting energy density; R 1 represents the cutting edge radius; N This refers to the drill bit rotation speed; V This refers to the drilling speed; M This refers to the drilling torque. F Drilling pressure; R The radius of the drill bit; l The cutting edge length; α The angle between the cutting edge and the horizontal direction. c The cutting angle, m The coefficient of friction, a、b It is a constant.

[0035] The above-mentioned rock mass compressive strength inversion model is obtained by performing mechanical analysis on a double-wing anchor drill bit during drilling. Different drill bit structures experience different forces during drilling. Therefore, different drill bit structures correspond to different rock mass compressive strength inversion models. Since the double-wing anchor drill bit is a commonly used drill bit in drilling tests, the above-mentioned rock mass compressive strength inversion model has a wide range of applications and can be applied to most existing drilling testing systems.

[0036] In this embodiment, the rock mass compressive strength prediction model while drilling adopts the GA-BP neural network prediction model. First, the parameters while drilling are normalized. The rock mass compressive strength prediction model while drilling is obtained by constructing a BP neural network model and using the GA genetic algorithm optimization process. The rock mass compressive strength prediction model while drilling uses the preprocessed parameters as input data and the predicted rock mass compressive strength as output data.

[0037] Furthermore, when normalizing the drilling parameters, the mapminmax function is used to normalize the data.

[0038] The constructed BP neural network model processes the drilling parameters through the input layer, hidden layer, and output layer in a forward propagation to obtain the final rock mass compressive strength. Based on the error between the calculated output rock mass compressive strength and the target rock mass compressive strength, backpropagation is performed to adjust the model's weights and thresholds. After updating the weights and thresholds, the forward propagation, error calculation, backpropagation, and weight and threshold update operations are repeated to finally achieve the set target requirements.

[0039] In the hidden layer, the number of nodes is calculated using an empirical method; when calculating the error, an appropriate learning rate is set based on repeated attempts to bring the model to converge; when using the GA genetic algorithm for optimization, the strong global optimization capability of the genetic algorithm is utilized to obtain the optimal initial weights and thresholds of the BP neural network to construct the GA-BP neural network prediction model.

[0040] In this embodiment, the structure surface parameter identification model adopts a 1D-CNN model. The drilling parameters are preprocessed as input data, and the 1D-CNN model is trained to learn local mutation features. The joint location is identified by threshold determination, and finally, parameter inversion is performed to calculate the position, angle and width of the structure surface based on the descent amplitude and width of the drilling parameter curve.

[0041] When preprocessing the drilling parameters, outlier removal and sliding smoothing are performed first, followed by normalization. When training a 1D-CNN model to learn local mutation features, local patterns of torque decrease are captured to predict joint probabilities. During training, the input layer is a one-dimensional sequence of drilling parameter data, the convolutional layers have appropriate kernel sizes and numbers, the pooling layers use average pooling, and the output layer displays the position, angle, and width of the structural plane.

[0042] By predicting structural surface parameters, structural surfaces are grouped together, with those sharing the same structural surface features forming a group. The total number of groups on the structural surface measurement lines is then counted. N i Ungrouped joint number N Calculate the number of rock mass grouped joints per meter along the normal direction. S i Number of non-grouped joints per cubic meter of rock mass S 0;

[0043]

[0044] in: S i This represents the number of structural surfaces per meter along the normal direction in the i-th group; N i Let be the total number of groups on the measurement lines of the i-th structural surface; The length of the survey line along the normal direction; S 0 represents the number of non-grouped joints per cubic meter of rock mass; N To measure the number of non-grouped joints on the line; L This represents the length of the measuring line.

[0045] Furthermore, the number of joints in the rock mass volume J vThe calculation formula is as follows:

[0046] in: J v The number of volumetric joints in the rock mass, in units per m. 3 m represents the number of structural surface groups within the statistical region.

[0047] Step 3: Calculate the equivalent compressive strength of the rock mass based on the inverted rock mass compressive strength and the predicted rock mass compressive strength. At the same time, obtain the rock mass integrity coefficient based on the rock mass volume joint number.

[0048] Specifically, the equivalent compressive strength of the rock mass is obtained by calculating the average of the inverted compressive strength of the rock mass during drilling and the predicted compressive strength of the rock mass. The inverted compressive strength of the rock mass during drilling is obtained based on regression analysis, while the predicted compressive strength of the rock mass is obtained based on machine learning. The average of the strengths obtained from the two different principle models is taken to make the value of the rock mass compressive strength more accurate.

[0049] Furthermore, according to the volumetric joint number of rock mass in the "Engineering Rock Mass Classification Standard" J v and integrity coefficient K v The correspondence was used to derive the integrity coefficient. K v .

[0050] Step 4: Input the equivalent compressive strength and rock mass integrity coefficient into the drilling fine classification model to obtain the in-situ drilling fine classification results of the surrounding rock.

[0051] Specifically, the drilling-while-drilling refined classification model is as follows:

[0052] in, BQ These are the basic quality indicators of the rock mass; K v This is the rock mass integrity coefficient; s c The compressive strength of the rock mass; e、g、s All are constants.

[0053] Furthermore, by setting multiple critical values, the rock mass to be evaluated is divided into multiple grades based on the comparison relationship between the basic quality indicators of the rock mass and the critical values.

[0054] Specifically, if the basic quality index of the rock mass exceeds the first critical value A1, it is classified as Grade I; if the basic quality index of the rock mass is between the first critical value A1 and the second critical value A2 (greater than the second critical value A2 but not exceeding the first critical value A1), it is classified as Grade II; if the basic quality index of the rock mass is between the second critical value A2 and the third critical value A3 (greater than the third critical value A3 but not exceeding the second critical value A2), it is classified as Grade III; if the basic quality index of the rock mass is between the third critical value A3 and the fourth critical value A4 (greater than the fourth critical value A4 but not exceeding the third critical value A3), it is classified as Grade IV; if the basic quality index of the rock mass is between the fourth critical value A4 and the fifth critical value A5 (greater than the fourth critical value A4 but not exceeding the fifth critical value A5), it is classified as Grade V. Among them, the first critical value A1, the second critical value A2, the third critical value A3, the fourth critical value A4, and the fifth critical value A5 decrease sequentially in descending order.

[0055] Furthermore, when performing refined grading of the surrounding rock, multiple thresholds are set, and the evaluated rock mass is refinedly graded according to the relationship between the basic quality indicators of the rock mass and the corresponding thresholds.

[0056] A first threshold B1 is inserted between the second critical value A2 and the third critical value A3. When the basic quality index of the rock mass is between A2 and B1, it is judged as Grade III1; when the basic quality index of the rock mass is between B1 and A3, it is judged as Grade III2.

[0057] A second threshold B2 and a third threshold B3 are inserted between the third critical value A3 and the fourth critical value A4. When the basic quality index of the rock mass is between A3 and B2, it is judged as level IV1; when the basic quality index of the rock mass is between B2 and B3, it is judged as level IV2; when the basic quality index of the rock mass is between B3 and A4, it is judged as level IV3.

[0058] A fourth threshold B4 is inserted between the fourth critical value A4 and the fifth critical value A5. When the basic quality index of the rock mass is between A4 and B4, it is judged as Grade V1; when the basic quality index of the rock mass is between B4 and A5, it is judged as Grade V2.

[0059] Among them, the first threshold B1, the second threshold B2, the third threshold B3 and the fourth threshold B4 decrease in descending order.

[0060] The above critical values ​​and thresholds are determined according to the surrounding rock classification standards.

[0061] Furthermore, based on the results of refined in-situ classification and evaluation of the engineering rock mass during drilling, a three-dimensional model of rock mass classification information is constructed, specifically including: First, n advance boreholes are drilled into the engineering rock mass and numbered sequentially as ZK1, ZK2, ZK3, ..., ZK n ; Secondly, based on the drilling parameters, the rock mass parameters of each borehole are obtained, and the rock mass within each borehole depth is finely classified. Then, the grading results are used as spatial point data, and three-dimensional spatial interpolation is performed using geostatistical interpolation algorithms to generate a continuous rock mass grade distribution surface. Finally, the distribution surfaces of the same depth are connected to form a three-dimensional space volume at the rock mass level, thus constructing a three-dimensional model of rock mass level information.

[0062] The established three-dimensional model of rock mass level information enables intelligent acquisition and analysis of rock mass parameters, and can be used in complex geological conditions. The three-dimensional model of rock mass level information can be adjusted in real time during the drilling process, enhancing the applicability of the project.

[0063] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for refined in-situ classification and evaluation of engineering rock masses during drilling, characterized in that, Includes the following steps: The drilling parameters during the drilling process are obtained through digital drilling tests. These parameters include drilling torque, drilling pressure, drilling speed, and drill bit rotation speed. Based on the drilling parameters, a drilling inversion model for rock mass compressive strength, a drilling prediction model for rock mass compressive strength, and a drilling identification model for structural surface parameters are constructed respectively, thereby obtaining the drilling inversion rock mass compressive strength, the predicted rock mass compressive strength, and the structural surface parameters respectively; The established rock mass compressive strength inversion model during drilling is as follows: in, σ c The compressive strength of the rock mass; η e The cutting energy density; R 1 represents the cutting edge radius; N This refers to the drill bit rotation speed; V This refers to the drilling speed; M This refers to the drilling torque; F Drilling pressure; R The radius of the drill bit; l The cutting edge length; α The angle between the cutting edge and the horizontal direction. γ The cutting angle, μ The coefficient of friction, a、b It is a constant; The rock mass compressive strength prediction model during drilling adopts the GA-BP neural network prediction model, which is obtained by constructing a BP neural network model and optimizing it using the GA genetic algorithm. The rock mass compressive strength prediction model during drilling uses the preprocessed drilling parameters as input data and the predicted rock mass compressive strength as output data. The drilling-while-drilling parameter identification model adopts a 1D-CNN model. Preprocessed drilling parameters are used as input data, and structural surface parameters are used as output data. Based on structural surface parameter prediction, structural surfaces are grouped, with identical structural surface features grouped together. The number of joints per meter along the normal direction in the rock mass is calculated. S i Number of non-grouped joints per cubic meter of rock mass S 0: in: S i This represents the number of structural surfaces per meter along the normal direction in the i-th group; N i Let be the total number of groups on the measurement lines of the i-th structural surface; The length of the survey line along the normal direction; S 0 represents the number of non-grouped joints per cubic meter of rock mass; N To measure the number of non-grouped joints on the line; L This refers to the length of the survey line; The equivalent compressive strength of the rock mass is calculated based on the inverted compressive strength and the predicted compressive strength of the rock mass during drilling. At the same time, the volume joint number of the rock mass is calculated based on the structural parameters, and the rock mass integrity coefficient is obtained. The equivalent compressive strength and rock mass integrity coefficient are input into the drilling-while-drilling refined classification model to obtain the drilling-while-drilling refined in-situ classification and evaluation results of the engineering rock mass.

2. The method for refined in-situ classification and evaluation of engineering rock mass during drilling as described in claim 1, characterized in that, The equivalent compressive strength of the rock mass is obtained by calculating the average of the rock mass compressive strength obtained from drilling inversion and the predicted rock mass compressive strength.

3. The method for refined in-situ classification and evaluation of engineering rock mass during drilling as described in claim 1, characterized in that, The number of volume joints in the rock mass J v The calculation formula is as follows: in: J v is the number of volumetric joints in the rock mass, in units per cubic meter; m is the number of structural plane groups within the statistical area; Furthermore, according to the volumetric joint number of rock mass in the "Engineering Rock Mass Classification Standard" J v and integrity coefficient K v The correspondence was used to derive the integrity coefficient. K v .

4. The method for refined in-situ classification and evaluation of engineering rock mass during drilling as described in claim 1, characterized in that, The refined classification model during drilling is as follows: in, BQ These are the basic quality indicators of the rock mass; K v This is the rock mass integrity coefficient; σ c The compressive strength of the rock mass; e, g, s All are constants.

5. The method for refined in-situ classification and evaluation of engineering rock mass during drilling as described in claim 1, characterized in that, By setting multiple critical values ​​and comparing them with the basic quality indicators of the rock mass, the rock mass to be evaluated is divided into multiple grades. The classification rules for rock mass grades are as follows: If the basic quality index of the rock mass exceeds the first critical value A1, it is judged as Grade I; If the basic quality index of the rock mass is greater than the second critical value A2 but does not exceed the first critical value A1, it is judged as Grade II; If the basic quality index of the rock mass is greater than the third critical value A3 but does not exceed the second critical value A2, it is judged as Grade III; If the basic quality index of the rock mass is greater than the fourth critical value A4 but does not exceed the third critical value A3, it is judged as Grade IV; If the basic quality index of the rock mass is greater than the fourth critical value A4 but does not exceed the fifth critical value A5, it is judged as Grade V; Among them, the first critical value A1, the second critical value A2, the third critical value A3, the fourth critical value A4, and the fifth critical value A5 decrease sequentially in descending order.

6. The method for refined in-situ classification and evaluation of engineering rock mass during drilling as described in claim 5, characterized in that, By setting thresholds, the rock mass that has been preliminarily evaluated is further subdivided based on the comparison between the basic quality indicators of the rock mass and these critical values. A first threshold B1 is inserted between the second critical value A2 and the third critical value A3. When the basic quality index of the rock mass is between A2 and B1, it is judged as Grade III1; when the basic quality index of the rock mass is between B1 and A3, it is judged as Grade III2. A second threshold B2 and a third threshold B3 are inserted between the third critical value A3 and the fourth critical value A4. When the basic quality index of the rock mass is between A3 and B2, it is judged as level IV1; when the basic quality index of the rock mass is between B2 and B3, it is judged as level IV2; when the basic quality index of the rock mass is between B3 and A4, it is judged as level IV3. A fourth threshold B4 is inserted between the fourth critical value A4 and the fifth critical value A5. When the basic quality index of the rock mass is between A4 and B4, it is judged as Grade V1; when the basic quality index of the rock mass is between B4 and A5, it is judged as Grade V2. Among them, the first threshold B1, the second threshold B2, the third threshold B3 and the fourth threshold B4 decrease in descending order.

7. The method for refined in-situ classification and evaluation of engineering rock mass during drilling as described in claim 1, characterized in that, Based on the results of refined in-situ classification and evaluation of engineering rock masses during drilling, a three-dimensional model of rock mass classification information is constructed, specifically including: First, n advance boreholes are drilled into the engineering rock mass and numbered sequentially as ZK1, ZK2, ZK3, ..., ZK n ; Secondly, based on the drilling parameters, the rock mass parameters of each borehole are obtained, and the rock mass within each borehole depth is finely classified. Then, the grading results are used as spatial point data, and three-dimensional spatial interpolation is performed using geostatistical interpolation algorithms to generate a continuous rock mass grade distribution surface. Finally, the distribution surfaces of the same depth are connected to form a three-dimensional space volume at the rock mass level, thus constructing a three-dimensional model of rock mass level information.

Images