Deep vertical shaft soft rock large deformation comprehensive index evaluation method

Abstract

The invention provides a deep vertical shaft soft rock large deformation comprehensive index evaluation method comprising the following steps: carrying out geological environment analysis on a target area to obtain large deformation influence factors; calculating a large deformation evaluation index according to the large deformation influence factor; constructing a large deformation case database and a machine learning model, and performing risk prediction on the large deformation case database through the machine learning model to obtain a prediction result; dividing the soft rock large deformation grade according to the prediction result and the large deformation evaluation index; and executing different supporting schemes according to the soft rock large deformation grade. According to the method, the rock mass quality grading system is established through the multi-dimensional data and the large deformation evaluation indexes, meanwhile, the soft rock large deformation grade is divided according to the prediction result of the machine learning model on the large deformation case database, the soft rock large deformation prediction precision and the support design scientificity are improved, and the safety and efficiency of the deep vertical shaft project are guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of geotechnical engineering and geological disaster prevention and control, and in particular to a method for evaluating comprehensive indicators of large deformation of soft rocks in deep vertical shafts. Background Art

[0002] With the advancement of deep mineral resource mining, deep vertical shaft projects face complex geological environments such as high geostress, high osmotic pressure, and high ground temperature, leading to an increasingly prominent problem of large deformation in soft rock. Soft rock is susceptible to rheological, dilatant, and shear failure under the high stresses at depth. Traditional support methods often fail due to inadequate consideration of the nonlinear mechanical properties of the surrounding rock and multi-field coupling effects, leading to engineering disasters such as shaft wall collapse and damage to support structures.

[0003] Currently, existing technologies lack the ability to account for dynamic changes in temperature, groundwater levels, and the non-uniform distribution of geostress, resulting in significant deviations between predicted results and actual deformation patterns. Furthermore, support scheme design relies on empirical formulas, making it difficult to adapt to the differentiated needs of varying deformation levels. Furthermore, the lack of real-time monitoring and dynamic adjustment mechanisms results in low support efficiency and high costs. Therefore, it is crucial to develop a comprehensive evaluation method for large deformation in soft rock in deep vertical shafts. Summary of the Invention

[0004] The purpose of this invention is to provide a comprehensive index evaluation method for large deformation of soft rock in deep vertical shafts, which corrects rock quality indicators through multi-source data, combines machine learning to predict risk levels and match differentiated support schemes to achieve accurate prediction of large deformation of soft rock.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] A method for evaluating comprehensive indicators of large deformation of soft rock in deep vertical shafts includes the following steps:

[0007] Conduct geological environment analysis on the target area to obtain the factors affecting large deformation;

[0008] Calculate large deformation evaluation indicators based on large deformation influencing factors; large deformation evaluation indicators include: rock mass quality indicators, surrounding rock stability and stress state;

[0009] Build a large deformation case database and a machine learning model, and use the machine learning model to perform risk prediction on the large deformation case database to obtain prediction results;

[0010] The soft rock large deformation levels are divided according to the prediction results and large deformation evaluation indicators; the soft rock large deformation levels include: normal deformation, level I deformation, level II deformation and level III deformation;

[0011] Different support schemes are implemented according to the large deformation level of soft rock.

[0012] Optionally, the factors affecting large deformation include: geological conditions, ground stress, hydrological conditions, formation temperature, rock mechanical properties, rock physical indicators and mining data;

[0013] Rock mechanical properties include: elastic modulus, Poisson's ratio, rock uniaxial tensile strength, rock uniaxial compressive strength, rock shear strength and internal friction angle;

[0014] Rock physical indicators include: density, bulk density, porosity, permeability, mineral composition and thermophysical properties;

[0015] Mining data include: rock strength coefficient, drilling parameters, explosive parameters and detonation parameters.

[0016] Optionally, the ground stress is measured by hydraulic fracturing or casing stress relief method; the hydrological conditions are measured by electrical sounding method and seismic method.

[0017] Optionally, the calculation formula for the rock quality index is: BQ=100+3Rc+250Kv-100(K1+K2+K3+K4); where Rc is the saturated uniaxial compressive strength, Kv is the rock integrity coefficient, K1 is the groundwater influence correction coefficient, K2 is the main structural surface attitude influence correction coefficient, K3 is the initial stress state influence correction coefficient, and K4 is the temperature correction coefficient.

[0018] Optionally, surrounding rock stability includes difficulty coefficient and proximity to failure;

[0019] The calculation formula for the difficulty coefficient is: Among them, D f is the difficulty coefficient, H is the actual depth of the underground project, and Hcr is the critical depth of the deep project;

[0020] The calculation formula for destruction proximity is: Where w is the stress state parameter of the surrounding rock, and FD is the damage factor.

[0021] Optionally, the stress state includes: a strength stress index, a relative deformation index, and a comprehensive coefficient;

[0022] The calculation formula of strength stress index is: S = R / σ0; where S is the strength stress index, R is the uniaxial compressive strength, and σ0 is the initial ground stress;

[0023] The calculation formula of relative deformation index is: ε t =u / d; where, ε t is the relative deformation index, u is the wellbore displacement, and d is the wellbore radius;

[0024] The calculation formula of the comprehensive coefficient is: Among them, α is the comprehensive coefficient, R bis the uniaxial compressive strength of rock, λ is the lateral pressure coefficient, σ v is the initial ground stress.

[0025] Optionally, the steps of constructing a large deformation case database include:

[0026] Determine the database construction goals;

[0027] Collect data based on database construction goals to obtain historical data; data collection includes: historical case collection, on-site investigation and expert consultation;

[0028] Clean and organize historical data to obtain standardized data;

[0029] Each data in the standardized data is assigned a corresponding attribute, and the relationship between the attribute and the standardized data is determined to obtain a large deformation case database.

[0030] Optionally, the machine learning model includes: a random forest model, a LightBGM model, an XGBoost model, a K-nearest neighbor regression model, a catboost model, an SVR model, a Lasso regression model, and a ridge regression model.

[0031] Optionally, the classification criteria for large deformation levels of soft rock are:

[0032] When the comprehensive coefficient is greater than 60 or the relative deformation index is less than 3%, the large deformation level of soft rock is determined as normal deformation;

[0033] When 30<comprehensive coefficient ≤ 60 or 3% ≤ relative deformation index < 5%, the large deformation grade of soft rock is determined as Grade I deformation;

[0034] When 15<comprehensive coefficient ≤30 or 5% ≤ relative deformation index <8%, the large deformation level of soft rock is determined as level II deformation;

[0035] When the comprehensive coefficient is ≤15 or the relative deformation index is ≥8%, the large deformation level of soft rock is determined as level III deformation.

[0036] Optionally, the support scheme includes: cast concrete support, anchor rod and cable support, U-shaped steel bracket support and double-layer steel plate high-strength concrete arc plate bracket support.

[0037] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects: The present invention provides a method for evaluating comprehensive indicators of large deformation of soft rock in deep vertical shafts, which includes: conducting a geological environment analysis of the target area to obtain factors affecting large deformation; calculating a large deformation evaluation index based on the large deformation influencing factors; constructing a large deformation case database and a machine learning model, and using the machine learning model to perform risk prediction on the large deformation case database to obtain prediction results; classifying the large deformation level of soft rock based on the prediction results and the large deformation evaluation index; and implementing different support schemes based on the large deformation level of soft rock. This method integrates multi-dimensional data such as geological structure, ground stress, hydrological conditions, and formation temperature, and introduces a temperature correction coefficient to dynamically correct rock mass quality indicators. Simultaneously, it constructs an intelligent risk prediction system, overcoming the shortcomings of traditional static evaluation systems, which are insufficiently adaptable to deep non-uniform stress fields and thermal-mechanical coupling effects, and significantly improving the accuracy of large deformation prediction of soft rock. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0039] Figure 1 This is a flow chart of the comprehensive index evaluation method for large deformation of soft rock in deep vertical shafts of the present invention;

[0040] Figure 2 This is a flow chart of building a large deformation case database for the present invention. DETAILED DESCRIPTION

[0041] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0042] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0043] like Figure 1 As shown, the present invention provides a method for evaluating comprehensive indicators of large deformation of soft rock in deep vertical shafts, comprising the following steps:

[0044] Step 100: Analyze the geological environment of the target area to obtain factors affecting large deformation;

[0045] Specifically, factors affecting large deformation include: geological conditions, ground stress, hydrological conditions, formation temperature, rock mechanical properties, rock physical indicators and mining data.

[0046] Geological conditions are composed of geological structures and stratigraphic structures. Geological structures are the various deformation and displacement traces left in rock layers and rock masses due to tectonic changes and are the product of crustal movement. Strata are layered rocks formed at a certain time in geological history, a layer or group of rocks with a specific position in the crust.

[0047] In-situ stress is the naturally occurring stress state within underground rock masses or formations in the absence of external disturbances. It includes gravitational stress and tectonic stress. Gravity stress manifests as vertical stress, while tectonic stress manifests as horizontal stress. In this embodiment, in-situ stress is measured using hydraulic fracturing or core stress relief methods.

[0048] Hydrological conditions refer to the general conditions related to the formation, distribution, and change of groundwater, and are measured by electrical depth sounding and seismic methods. Formation temperature refers to the temperature of rocks at different depths underground. It increases with increasing depth, and its temperature distribution is affected by various factors such as heat flow inside the earth, thermal conductivity of rocks, and groundwater activity. Rock mechanical properties include: elastic modulus E, Poisson's ratio v, uniaxial tensile strength σ of rocks t , rock uniaxial compressive strength σ c , rock mass shear strength τ and internal friction angle The calculation formula is as follows:

[0049]

[0050] Where σ is stress, ε is strain, and ε x is the transverse strain, ε y is the axial strain, P t is the maximum axial tensile load, A is the cross-sectional area of ​​the specimen, P is the maximum axial pressure, c is the bond force on the structural surface, σ n is the normal stress acting on the structural surface. Rock physical indicators include: density ρ, bulk density γ, porosity n, permeability Q x , mineral composition and thermal properties, the calculation formula is as follows:

[0051]

[0052]

[0053] Where G is the mass of the rock being tested, V is the volume of the rock being tested, W is the weight of the rock being tested, and V P is the total volume of rock pores. Thermophysical properties include thermal conductivity, specific heat capacity and thermal diffusivity.

[0054] Mining data includes rock strength coefficient, drilling parameters, explosive parameters, and detonation parameters. Drilling parameters include hole spacing, row spacing, and blasthole depth. Explosive parameters include charge structure, charge quantity, and unit explosive consumption.

[0055] Rock mechanical properties include: elastic modulus, Poisson's ratio, rock uniaxial tensile strength, rock uniaxial compressive strength, rock shear strength and internal friction angle;

[0056] Rock physical indicators include: density, bulk density, porosity, permeability, mineral composition and thermophysical properties;

[0057] Mining data include: rock strength coefficient, drilling parameters, explosive parameters and detonation parameters.

[0058] Step 200: Calculating large deformation evaluation indicators based on large deformation influencing factors; large deformation evaluation indicators include: rock mass quality indicators, surrounding rock stability and stress state;

[0059] Specifically, in order to reflect the impact of rock strength degradation under high temperature environment on rock mass quality classification, the rock mass quality index BQ is dynamically corrected by introducing the rock temperature correction coefficient K4. The calculation formula is:

[0060] BQ=100+3Rc+250Kv-100(K1+K2+K3+K4);

[0061] Where Rc is the saturated uniaxial compressive strength, Kv is the rock mass integrity coefficient, K1 is the groundwater effect correction factor, K2 is the main structural surface orientation effect correction factor, K3 is the initial stress state effect correction factor, and K4 is the temperature correction factor. The rock mass integrity coefficient is the square of the ratio of the longitudinal wave velocity of the rock mass to the rock.

[0062] Specifically, surrounding rock stability includes difficulty coefficient and failure proximity;

[0063] The calculation formula for the difficulty coefficient is: Among them, D f is the difficulty coefficient, H is the actual depth of the underground project, and Hcr is the critical depth of the deep project; when D f When D is less than 1, it indicates that the engineering rock mass is in a linear working state and its stability can be controlled by conventional methods. f When ≥1, it indicates that the soft rock engineering rock group is in a nonlinear large deformation working state, and the hard rock engineering rock group will experience nonlinear dynamic phenomena such as rock burst.

[0064] The calculation formula for destruction proximity is: Where w is the stress state parameter of the surrounding rock, and FD is the damage factor. When FAI < 1, the rock is primarily in an elastic development stage, the stress-strain curve develops linearly, and the surrounding rock primarily stores energy, requiring no support to maintain stability. When 1 ≤ FAI < 2, the rock mass is in a plastic yield state, with stress exceeding the rock damage stress (the volumetric strain turning point). Under high stress, even if stress does not increase, cracks may develop within the surrounding rock. Maintaining the long-term stability of the surrounding rock requires appropriate support measures to increase the ultimate strength of the rock. When FAI ≥ 2, the rock mass is in a destructive state, with stress exceeding the ultimate strength, resulting in significant macroscopic cracks within the rock. Support measures should be implemented promptly during construction to ensure surrounding rock stability. A larger FAI value indicates more severe rock damage. Based on the FAI index classification principle, FAI = 1 can be used as the criterion for determining whether support is required after rock excavation.

[0065] Specifically, the stress state includes: strength stress index, relative deformation index and comprehensive coefficient.

[0066] The strength stress index is calculated by the strength stress ratio method, and the calculation formula is:

[0067] S=R / σ0;

[0068] Among them, S is the strength stress index, R is the uniaxial compressive strength, and σ0 is the initial ground stress. The strength stress indexes of the two examples in this embodiment are respectively obtained by the uniaxial compressive strength R of rock. b and the uniaxial compressive strength of rock mass R cm The calculation formula for the strength stress index of Example 1 is: S = R b / σ0; where R b Obtained from rock mechanics test. The strength stress index calculation formula of Example 2 is: S = R cm / σ0; where R cm It is calculated through empirical formula or theoretical formula. The calculation formula is: Where c is the rock mass cohesion, is the internal friction angle.

[0069] The calculation formula of relative deformation index is: ε t =u / d; where, ε t is the relative deformation index, u is the wellbore displacement, and d is the wellbore radius;

[0070] The comprehensive coefficient is obtained by comprehensively calculating the surrounding rock compressive strength, ground stress, elastic modulus and lateral pressure coefficient. The calculation formula is: Among them, α is the comprehensive coefficient, R b is the uniaxial compressive strength of rock, λ is the lateral pressure coefficient, σ vis the initial ground stress.

[0071] Step 300: constructing a large deformation case database and a machine learning model, and performing risk prediction on the large deformation case database using the machine learning model to obtain prediction results;

[0072] like Figure 2 As shown in Figure 2, the steps for building a large deformation case database include:

[0073] Step 301: Determine the database building target;

[0074] Specifically, the large deformation case database should include information on the engineering background, geological conditions, lithologic characteristics, and rockburst risk level of rockbursts, so as to provide comprehensive data support for the research and practice of large deformation in soft rocks.

[0075] Step 302: Collect data according to the database construction goal to obtain historical data. The data collection methods include:

[0076] Collection of historical cases: By consulting literature, archival records and news reports, historical rock burst engineering cases are collected, which should cover different types of engineering, soft rock large deformation events with different geological conditions and lithologic characteristics.

[0077] On-site investigation: Conduct field investigations on underground projects under construction or completed to understand the working conditions of large deformation, disaster levels, prevention and control measures, and other information, providing real and reliable data support for the establishment of the database.

[0078] Expert consultation: Through consultation and communication with experts and scholars in the field of large deformation, we can understand the latest progress and results of large deformation research, as well as the experts' opinions and suggestions on large deformation issues.

[0079] Step 303: Clean and organize the historical data to obtain standardized data;

[0080] Specifically, the collected data is cleaned and organized, removing duplicate, erroneous, and invalid data to ensure accuracy and reliability. Case studies are also categorized and organized based on factors such as the type of project where the rockburst occurred, geological conditions, and lithologic characteristics. The data within these cases is standardized to facilitate comparison and analysis.

[0081] Step 304: assign corresponding attributes to each data in the standardized data, and determine the relationship between the attributes and the standardized data to obtain a large deformation case database.

[0082] Specifically, for each data parameter, appropriate fields and attributes are designed. Fields cover various aspects of the case, such as project name, geological conditions, lithologic characteristics, and rockburst severity. Relationships between data parameters, such as one-to-one, one-to-many, or many-to-many relationships, are also determined to facilitate data association and query.

[0083] Specifically, the machine learning models include: random forest model, LightBGM model, XGBoost model, K nearest neighbor regression model, catboost model, SVR model, Lasso regression model and ridge regression model.

[0084] The XGBoost model is a tree-structured decision tree, where each internal node represents a test on a feature, each branch represents an output of this test, and each leaf node stores a category or a value. Its objective function consists of two parts: a loss function and a regularization term. The loss function of this embodiment is the mean square error (MSE). The regularization term is used to control the complexity of the model. In the XGBoost model, the number of leaf nodes and the L2 norm of the leaf node scores are penalized to prevent the model from overfitting. The XGBoost model uses an additive training method. At each step t, a new decision tree f is found. t , so that the objective function value after adding this decision tree is minimized. Assume that the prediction of the current model is The prediction after adding the new tree is where x i is the feature vector of the i-th sample.

[0085] The random forest model uses bagging to randomly sample the original dataset with replacement. Each decision tree in the model is trained on a different subset. During each tree split, a subset of all features is randomly selected to find the optimal split point. The results of multiple trees are combined using a majority vote (classification) and a mean (regression). For classification problems, this model uses the Gini coefficient or information gain to measure node purity and selects the feature that maximizes purity for splitting. For regression problems, the mean squared error is used to select the splitting feature.

[0086] The K-nearest neighbor regression model uses the averaging method to find the K sample points in the dataset that are most similar (closest to) a new sample point, and predicts the target value of the new sample based on the target value of these K sample points. The distance is calculated using Euclidean distance or Manhattan distance.

[0087] The LightBGM model divides the values ​​of continuous features into fixed discrete intervals and then calculates the gradient statistics for each interval, significantly reducing the number of data scans during model training and improving training speed. It also adopts a leaf-first strategy, selecting the leaf node with the largest split gain for splitting in each iteration, thereby reducing training time and improving model accuracy. The model retains samples with large gradients through unilateral gradient sampling and uses a mutually exclusive feature bundling algorithm to bundle mutually exclusive features together to form a new feature, thereby reducing feature dimensionality.

[0088] The objective function of the SVR model is: Where w is the weight vector of the hyperplane, b is the intercept, ξ i and is a slack variable used to allow some data points to be within the interval or error range, and C is a regularization parameter used to control the trade-off between the interval size and the error.

[0089] The objective function of the Lasso regression model is: The regularization parameter λ ≥ 0. The model uses the coordinate descent method to optimize the objective function.

[0090] The objective function of the ridge regression model is: The regularization parameter λ is ≥ 0. The regularization term is used to limit the size of the coefficient β to prevent overfitting. The model uses the normal equation method to optimize the objective function.

[0091] The catboost model introduces feature combination technology based on GBDT, combines original features to generate new features, and uses the interaction between features to improve the expressiveness of the model. The objective function expression of this model is: Among them, N is the number of samples, M is the number of features, and x i is a vector of features, F0(x) is the initial guess or baseline prediction, which in this example is the mean of the target variable in the dataset, represents the sum of all trees, M represents the total number of trees in the set, represents the sum of training samples, and N represents the total number of training samples.

[0092] Step 400: Classify the soft rock large deformation level according to the prediction results and the large deformation evaluation index; the soft rock large deformation level includes: normal deformation, level I deformation, level II deformation and level III deformation;

[0093] Specifically, the classification criteria for large deformation levels of soft rock are:

[0094] When the comprehensive coefficient is greater than 60 or the relative deformation index is less than 3%, the large deformation level of soft rock is determined to be normal deformation; the characteristics are as follows: the stability of the surrounding rock can be guaranteed by existing support measures and basically no adjustment is required; at this deformation level, since the rock remains stable and does not pose a threat to the project or personnel, special large deformation protection measures are usually not required.

[0095] When 30 < comprehensive coefficient ≤ 60 or 3% ≤ relative deformation index < 5%, the soft rock large deformation level is determined to be Level I deformation; the characteristics are: large and time-sensitive deformation of the surrounding rock, local cracking of the shotcrete, and partial loss of contact between the steel arch and the shotcrete; this deformation level has little impact on the overall stability of the project, but it is still necessary to increase the monitoring frequency and promptly clean up the flaked rock to prevent minor injuries to construction personnel or equipment.

[0096] When 15 < comprehensive coefficient ≤ 30 or 5% ≤ relative deformation index < 8%, the soft rock deformation grade is determined to be Grade II deformation; the characteristics are: large deformation of the surrounding rock with obvious time-effect, obvious cracking of the shotcrete, severe local damage, local bending of the steel arch frame, and bulging deformation of the bottom arch; at this deformation grade, a large number of cracks in the soft rock are connected, forming a large fracture surface, and the rock mass shows obvious block slip. Appropriate protective measures need to be taken to reduce potential risks.

[0097] When the comprehensive coefficient is ≤15 or the relative deformation index is ≥8%, the soft rock deformation level is determined to be Grade III deformation; the characteristics are: large deformation of the surrounding rock, significant timeliness and difficulty in convergence, large-scale cracking, damage and peeling of the shotcrete, large-scale deformation and distortion of the steel arch frame, and obvious bulging and deformation of the bottom arch. This deformation level will cause serious damage to the project structure, or even complete scrapping, posing a huge threat to the lives of construction workers. Construction needs to be stopped immediately for emergency rescue and reassessment of the project feasibility.

[0098] Step 500: Execute different support schemes according to the large deformation level of the soft rock.

[0099] Specifically, the support scheme includes: cast concrete support, anchor rod and cable support, U-shaped steel support and double-layer steel plate high-strength concrete arc plate support.

[0100] The thickness of the well wall must be known when pouring concrete support. When the well wall is under stress, the stress state of the inner edge of the well wall is used as the well wall design control condition, and the radial stress σ at the inner edge of the well wall is obtained based on the circular hole mechanical model analysis. r , vertical stress σ z and tangential stress σ θ , the calculation formulas are:

[0101] σ r =0

[0102]

[0103] σ z =μ W σ θ

[0104] Among them, p is the external load on the well wall, r a and r b are the inner radius and outer radius of the wellbore structure, μ w is the Poisson's ratio of the wellbore. The tangential stress should also meet the following conditions:

[0105]

[0106] Where γ0 is the structural importance coefficient of the wellbore, which is 1.1 in this embodiment, and v k is the safety factor of the shaft wall structure, which is 1.35 in this embodiment, and f' s is the design value of shaft wall strength considering the multiaxial strength of concrete, f' s The calculation formula is:

[0107] f′ s =β c (f s +ρ min f′ y );

[0108] Among them, β c is the concrete strength improvement coefficient under biaxial compression, which is 1.2 in this embodiment, f s is the design value of the well wall material strength, which is not specifically limited in this embodiment, ρ min is the minimum reinforcement ratio, f′ y is the design value of the compressive strength of ordinary steel bars, which is not specifically limited in this embodiment. Finally, the shaft wall thickness t and the shaft wall support resistance P are obtained. cm The calculation formulas are:

[0109]

[0110] Among them, τ cm is the shear strength of the supporting concrete material, t cm is the thickness of the first support layer, a cm is the failure angle of the first layer of supporting concrete, and the ultimate bearing strength q should also meet the following conditions:

[0111]

[0112] q>P0;

[0113] Among them, P0 is the original rock stress, b is the thickness of the bearing structure layer, r is the wellbore radius, P is the total support resistance of the support structure, k is the radial stress increase slope, which is 0 in unstable rock mass. and are the internal friction angle and cohesion of the wellbore support, respectively.

[0114] Anchor cable support is an active support method that can deform in coordination with the surrounding rock and make full use of the surrounding rock's own bearing capacity to resist deformation of the surrounding rock. The anchor cable is made of steel strands with a certain degree of bending flexibility. It has the characteristics of large anchoring depth, strong bearing capacity, and large preload force. In addition to the suspension, composite beam, composite arch, and wedging effects of ordinary anchors, the anchor cable is different from ordinary staggered rods in that it deeply anchors the top plate to produce a strong suspension effect. This suspension effect anchors the lower unstable rock layer in the upper stable rock layer, greatly enhancing the stability of the top plate. At the same time, the anchor cable can also apply preload force like an anchor rod to achieve active support of the surrounding rock. Anchor cable support can be used as temporary support and can also be combined with other structural forms to form composite support. In order to control the deformation of the flow layer, the anchor rod needs to pass through the flow layer and anchor in the plastic layer. The length of the anchor rod L bt and spacing D b The calculation formula is:

[0115] L bt =l b1 +l b2 +l b3 ;

[0116]

[0117] Among them, l b1 is the exposed length of the anchor rod, l b2 is the effective length of the anchor rod, which is the thickness of the surrounding rock stratification flow layer in this embodiment, l b3 is the anchor length of the anchor rod, in this embodiment, it is the thickness of the anchor rod extending into the plastic bearing layer, a is the wellbore excavation radius, and N is the number of anchor rods supporting the wellbore section. b The calculation formula is:

[0118]

[0119] Among them, Q b is the anchoring force of the anchor rod, which is determined by the anchor rod tensile and pull-out test, D1 is the anchor rod spacing, and D2 is the anchor rod row spacing. sr The calculation formula is:

[0120]

[0121] Among them, τ sris the shear strength of the metal mesh material, S sr is the cross-sectional area of ​​the metal mesh along the tunnel axis, β sr The shear angle of the metal mesh material.

[0122] U-shaped steel support is a passive support method that can provide high support for the surrounding rock, has a retractable and stable and reliable bearing capacity, and can adapt to the load and deformation of soft surrounding rock. The maximum resistance of the steel support of this support method is P max The calculation formula is:

[0123]

[0124] Among them, A s is the cross-sectional area of ​​the steel support, I s is the moment of inertia of the steel support section, σ s is the yield strength of steel, a is the equivalent radius of tunnel excavation, S is the spacing between supports along the length of the tunnel, θ is half the angle between the two pads of the steel support, x is the thickness of the steel support section, t B is the pad thickness.

[0125] Double-layer steel plate high-strength concrete arc plate support is a passive support, and its structure includes: backing material, metal mesh and sprayed concrete (or U-shaped steel), retractable joints and high-strength arc plate supports. The backing material, metal mesh and sprayed concrete are temporary supports for soft rock tunnels, playing a secondary role in the stability of the mine and tunnel, and cannot determine the ultimate stability of the soft rock tunnel; the retractable joints and high-strength arc plate supports constitute a double-layer steel plate high-strength concrete arc plate support, which provides permanent support for soft rock mines and tunnels and directly determines the ultimate stability of the soft rock tunnel. The ultimate bearing capacity of this support method is P b The calculation formula is:

[0126] P b b=σ g h g +σ ht h t ;

[0127] Where b is the outer radius of the arc plate support, h g is the total thickness of the steel plate, h t is the thickness of the concrete between the double steel plates, σ ht is the mean hoop stress of concrete in the ultimate state, σ g Steel is plate yield strength.

[0128] Specifically, when the soft rock's large deformation level is normal, anchor bolt support and cast concrete support are used. For anchor bolt support, Q235 anchor bolts are selected, with a length of 2m to 3m, a diameter of 20mm, and a spacing of 1.0 x 1.0m. The anchor force is 100kN. The anchor bolt tray is made of 10mm thick Q235 steel plate bent by pressure, with specifications of length x width x thickness = 200 x 200 x 10mm. For cast concrete support, C30 concrete is selected, and the thickness of the cast concrete is approximately 60cm.

[0129] When the soft rock reaches Grade I deformation, anchor bolts and cable anchors are used, along with poured concrete support. For anchor bolts and cable anchors, Q235 anchors with a length of 2m to 3m, a diameter of 20mm, and a spacing of 1.0 x 1.0m are used. The anchor force is 100kN, and the anchor tray is made of 10mm thick Q235 steel plate, with dimensions of 200 x 200 x 10mm (length x width x thickness). Low-relaxation steel strand anchor cables with a length of 4m to 6m, a diameter of 21.8mm, and a spacing of 1.0 x 1.0m are used. The anchor force is 260kN, and the prestress is 150kN. The anchor tray is made of 16mm thick Q235 steel plate, with dimensions of 120 x 120 x 16mm (length x width x thickness). For poured concrete support, C30 concrete is used, with a concrete thickness of approximately 80cm.

[0130] When the soft rock reaches Grade II deformation, anchor bolts, cast concrete, and U-shaped steel brackets are used. For anchor bolts and cable support, Q235 anchors with a length of 2m to 4m, a diameter of 20mm, and a spacing of 0.8 x 0.8m are used. The anchor force is 100kN, and the anchor tray is made of 10mm thick Q235 steel plate, with dimensions of 200 x 200 x 10mm. Low-relaxation steel strand anchor cable support is used, with a length of 6m to 8m, a diameter of 21.8mm, and a spacing of 0.8 x 0.8m. The anchor force is 260kN and the prestress is 150kN. The anchor tray is made of 16mm thick Q235 steel plate, with dimensions of 120 x 120 x 16mm. C40 concrete was selected for the concrete support, with a thickness of approximately 80 cm. This concrete can provide a certain degree of support to the rock surface after excavation. It can also be used in conjunction with anchor rods, anchor cables, and U-shaped steel supports to form a combined support system, achieving synergistic effects. The U-shaped steel support utilizes U29-type retractable steel supports with a spacing of 1.0 m.

[0131] When the soft rock reaches Grade III deformation, anchor support, poured concrete support, and double-layer steel plate high-strength concrete arc slab support are used. For anchor support, Q235 anchor rods with a length of 2m to 4m, a diameter of 20mm, and a spacing of 0.8 x 0.8m are used. The anchor force is 100kN. The anchor tray is made of 10mm thick Q235 steel plate, with a bending dimension of 200 x 200 x 10mm. Low-relaxation steel strand anchor cables are used for support, with a length of 6m to 8m, a diameter of 21.8mm, and a spacing of 0.8 x 0.8m. The anchor force is 260kN and the prestress is 150kN. The anchor tray is made of 16mm thick Q235 steel plate, with a dimension of 120 x 120 x 16mm. C50 concrete was selected for the concrete support, with a thickness of approximately 80cm. The double-layer steel plate high-strength concrete arc slab support used U36 steel scaffolding with a row spacing of 1.0m and C80 concrete support with a thickness of 460mm.

[0132] The beneficial effects of the present invention are as follows:

[0133] 1) By integrating multi-source data such as geological structure, geostress, hydrology, and temperature, and introducing a temperature correction coefficient to dynamically optimize rock mass quality indicators, the accuracy of large deformation prediction for soft rock has been significantly improved, solving the problem of the traditional static evaluation system's lack of adaptability to complex deep environments.

[0134] 2) By analyzing a historical engineering case database based on a machine learning model, dynamic identification and classification of soft rock deformation risks were achieved, reducing prediction errors;

[0135] 3) Scientific support plans are matched to different risk levels, which greatly improves support efficiency and reduces project costs.

[0136] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0137] The present invention uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only intended to help understand the method and core concept of the present invention. At the same time, those skilled in the art will find that the specific implementation methods and application scopes may vary based on the concept of the present invention. In summary, the contents of this specification should not be construed as limiting the present invention.

Claims

1. A method for evaluating comprehensive indicators of large deformation of soft rock in deep vertical shafts, characterized by: The steps include: Conduct geological environment analysis on the target area to obtain the factors affecting large deformation; Calculate a large deformation evaluation index based on the large deformation influencing factors; the large deformation evaluation index includes: rock mass quality index, surrounding rock stability and stress state; Constructing a large deformation case database and a machine learning model, and performing risk prediction on the large deformation case database using the machine learning model to obtain a prediction result; Classify the soft rock large deformation level according to the prediction results and the large deformation evaluation index; the soft rock large deformation level includes: normal deformation, level I deformation, level II deformation and level III deformation; Different support schemes are implemented according to the large deformation level of the soft rock.

2. The method for evaluating comprehensive indicators of large deformation of soft rock in deep vertical shafts according to claim 1 is characterized in that: The factors affecting large deformation include: geological conditions, ground stress, hydrological conditions, formation temperature, rock mechanical properties, rock physical indicators and mining data; The rock mechanical properties include: elastic modulus, Poisson's ratio, rock uniaxial tensile strength, rock uniaxial compressive strength, rock shear strength and internal friction angle; The rock physical indicators include: density, bulk density, porosity, permeability, mineral composition and thermophysical properties; The mining data includes: rock strength coefficient, drilling parameters, explosive parameters and detonation parameters.

3. The method for evaluating comprehensive indicators of large deformation of soft rock in deep vertical shafts according to claim 2, characterized in that: The ground stress is measured by hydraulic fracturing or casing stress relief method; the hydrological conditions are measured by electrical sounding method and seismic method.

4. The method for evaluating comprehensive indicators of large deformation of soft rock in deep vertical shafts according to claim 1 is characterized in that: The calculation formula of the rock mass quality index is: BQ=100+3Rc+250Kv-100(K1+K2+K3+K4); among them, Rc is the saturated uniaxial compressive strength, Kv is the rock mass integrity coefficient, K1 is the groundwater influence correction coefficient, K2 is the main structural surface attitude influence correction coefficient, K3 is the initial stress state influence correction coefficient, and K4 is the temperature correction coefficient.

5. The method for evaluating comprehensive indicators of large deformation of soft rock in deep vertical shafts according to claim 1 is characterized in that: Said surrounding rock stability includes difficulty coefficient and failure proximity; The calculation formula of the difficulty coefficient is: Among them, D f is the difficulty coefficient, H is the actual depth of the underground project, and Hcr is the critical depth of the deep project; The calculation formula of the destruction proximity is: Where w is the stress state parameter of the surrounding rock, and FD is the damage factor.

6. The method for evaluating comprehensive indicators of large deformation of soft rock in deep vertical shafts according to claim 1, characterized in that: The stress state includes: strength stress index, relative deformation index and comprehensive coefficient; The calculation formula of the strength stress index is: S=R / σ0; wherein S is the strength stress index, R is the uniaxial compressive strength, and σ0 is the initial ground stress; The calculation formula of the relative deformation index is: ε t =u / d; where, ε t is the relative deformation index, u is the wellbore displacement, and d is the wellbore radius; The calculation formula of the comprehensive coefficient is: Among them, α is the comprehensive coefficient, R b is the uniaxial compressive strength of rock, λ is the lateral pressure coefficient, σ v is the initial ground stress.

7. The method for evaluating comprehensive indicators of large deformation of soft rock in deep vertical shafts according to claim 1, characterized in that: The steps of constructing the large deformation case database include: Determine the database construction goals; Collect data according to the database construction goal to obtain historical data; the data collection includes: historical case collection, on-site investigation and expert consultation; Cleaning and arranging the historical data to obtain standardized data; Each data in the standardized data is assigned a corresponding attribute, and a relationship between the attribute and the standardized data is determined to obtain the large deformation case database.

8. The method for evaluating comprehensive indicators of large deformation of soft rock in deep vertical shafts according to claim 1, characterized in that: The machine learning models include: random forest model, LightBGM model, XGBoost model, K nearest neighbor regression model, catboost model, SVR model, Lasso regression model and ridge regression model.

9. The method for evaluating comprehensive indicators of large deformation of soft rock in deep vertical shafts according to claim 6, characterized in that: The classification standards for the large deformation levels of soft rock are as follows: When the comprehensive coefficient is greater than 60 or the relative deformation index is less than 3%, the large deformation level of the soft rock is determined as the normal deformation; When 30 < the comprehensive coefficient ≤ 60 or 3% ≤ the relative deformation index < 5%, the large deformation level of the soft rock is determined as the level I deformation; When 15 < the comprehensive coefficient ≤ 30 or 5% ≤ the relative deformation index < 8%, the large deformation level of the soft rock is determined as the level II deformation; When the comprehensive coefficient is ≤15 or the relative deformation index is ≥8%, the large deformation level of the soft rock is determined as the level III deformation.

10. The method for evaluating comprehensive indicators of large deformation of soft rock in deep vertical shafts according to claim 1, characterized in that: The support scheme includes: cast concrete support, anchor rod and cable support, U-shaped steel bracket support and double-layer steel plate high-strength concrete arc plate bracket support.

Images