Method for calculating radial gradient damage degree of surrounding rock of roadway and related equipment thereof

Abstract

The invention provides a roadway surrounding rock radial gradient damage degree calculation method and related equipment thereof, and the method comprises the steps: obtaining original data of damage factors in a target geological region, the original data being used for determining the damage degree in the target geological region; preprocessing the original data to generate a training data set for damage degree calculation; according to the corresponding boundary conditions in the training data set, the target geological area is layered to obtain a layering result, and the layering result comprises at least one of a broken area, a plastic softening area and an elastic area; based on a layering result, inputting corresponding original data in the training data set into an improved BP neural network model for processing to obtain damage data in the target geological area; a damage feature map is generated based on the damage data. Through the steps of the method, layered calculation and visual expression of the radial gradient damage degree of the surrounding rock of the roadway can be realized, and the accuracy and efficiency of damage assessment are improved.

Description

Technical Field

[0001] This invention relates to the field of dynamic mechanical property testing of solid materials, and in particular to a method, apparatus, electronic device and storage medium for calculating the radial gradient damage degree of roadway surrounding rock. Background Technology

[0002] During the construction of underground mines, transportation tunnels, and underground engineering projects, the surrounding rock of roadways is prone to varying degrees of damage and deformation under the combined effects of in-situ stress and excavation disturbance. The stability of the surrounding rock directly affects the safe operation of the roadway and the reliability of the support structure. Since the surrounding rock typically exhibits a gradually changing mechanical state from the inside to the outside in the radial direction, its damage degree has a distinct radial gradient distribution characteristic. Therefore, accurately calculating and assessing the radial gradient damage degree of the surrounding rock in roadways is an important research topic in the fields of geotechnical engineering and underground engineering safety.

[0003] Existing methods for calculating the degree of damage to surrounding rock in roadways are mostly based on a single mechanical index or empirical criterion for overall assessment, or rely on numerical simulation to analyze the overall stability of the surrounding rock. They generally have the problem of only evaluating the surrounding rock as a whole and failing to reflect the radial zoning characteristics. Although some methods introduce intelligent algorithms for damage prediction, they fail to effectively combine them with the layered structure of the surrounding rock, resulting in mixed damage mechanisms in different mechanical sections. At the same time, in terms of presenting damage results, existing technologies usually output in numerical or two-dimensional curve form, which is difficult to intuitively reflect the radial gradient damage distribution characteristics of the surrounding rock and has insufficient engineering applicability.

[0004] Therefore, existing methods for calculating the radial gradient damage degree of roadway surrounding rock have the problem of being unable to accurately and efficiently calculate the damage degree on a layered basis. Summary of the Invention

[0005] This invention provides a method for calculating the radial gradient damage degree of roadway surrounding rock, in order to solve the problem that existing methods for calculating the radial gradient damage degree of roadway surrounding rock cannot achieve accurate and efficient calculation of the damage degree on a layered basis.

[0006] In a first aspect, the present invention provides a method for calculating the radial gradient damage degree of roadway surrounding rock, the method comprising the following steps: Obtain raw data of damage factors within the target geological area; the raw data is used to determine the degree of damage within the target geological area. The raw data is preprocessed to generate a training dataset for calculating the degree of damage; Based on the boundary conditions in the training dataset, the target geological region is subjected to stratification to obtain a stratification result, which includes at least one of the fractured zone, plastic softening zone, and elastic zone. Based on the hierarchical results, the corresponding raw data in the training dataset is input into the improved BP neural network model for processing to obtain damage data within the target geological area; Based on the damage data, a damage feature map is generated.

[0007] Optionally, the raw data includes stress parameters, geometric parameters, rock mechanics parameters, and strain parameters. Obtaining the raw data of damage factors within the target geological area includes: The stress state of the surrounding rock within the target geological area is collected to obtain stress parameters, including the original rock stress within the target geological area. s 0. Radial stress of surrounding rock s r and tangential stress of surrounding rock s θ ; The geometric morphology of the tunnels and the spatial relationship between the surrounding rock within the target geological area are collected to obtain geometric parameters, including the radial distance between any point in the surrounding rock and the center of the tunnel. R and the radius of the alley a ; The mechanical properties of rock materials within the target geological region are tested to obtain rock mechanical parameters, including the elastic modulus. E Poisson's ratio u internal friction angle f and cohesion c ; The plastic deformation behavior of rocks within the target geological region was tested to obtain strain parameters, which include the radial plastic strain increment during uniaxial compression of the rock. and tangential plastic strain increment .

[0008] Optionally, the preprocessing of the original data to generate a training dataset for calculating the degree of damage includes: Based on preset rock strength criterion parameters, the original data is subjected to a first preprocessing to obtain a first preprocessing result; Based on the uniaxial compression test results with a preset expansion coefficient, the original data is subjected to a second preprocessing to obtain a second preprocessing result; The training dataset is generated based on the first preprocessing result and the second preprocessing result.

[0009] Optionally, the step of performing stratification processing on the target geological region based on the corresponding boundary conditions in the training dataset to obtain the stratification result includes: Determine the preset stress conditions corresponding to the fracture zone, plastic softening zone, and elastic zone; Based on the preset stress conditions, the stress range of the target geological area is determined, and the range of the corresponding area is delineated from the target geological area. Based on the extent of each region, corresponding fractured zones, plastic softening zones, and elastic zones are divided within the target geological region.

[0010] Optionally, before inputting the corresponding raw data from the training dataset into the improved BP neural network model for processing based on the hierarchical results to obtain the damage data within the target geological area, the method further includes: Obtain the basic BP neural network model; Based on the preset additional momentum term, the weight algorithm of the basic BP neural network model is adjusted to obtain new weight update data; Based on the new weight update data, the basic BP neural network model is updated, adjusted, and trained to obtain an improved BP neural network model after training is completed.

[0011] Optionally, based on the hierarchical results, the corresponding raw data in the training dataset is input into an improved BP neural network model for processing to obtain damage data within the target geological area, including: Based on the stratification results, determine the stratification type corresponding to the location to be calculated within the target geological area; Based on the hierarchical type, select the original data corresponding to the hierarchical type from the training dataset as the data to be calculated; The data to be calculated is input into the improved BP neural network model for processing, and the damage data corresponding to the location to be calculated is obtained according to the damage algorithm corresponding to the layer type.

[0012] Optionally, generating a damage feature map based on the damage data includes: Based on the radial position parameters corresponding to the location to be calculated, the damage data is mapped to the radial coordinate space of the target geological area to obtain damage distribution data; Based on the layering results, the damage data is labeled according to the fracture zone, plastic softening zone and elastic zone to obtain the region labeling data; Based on the damage distribution data and the corresponding regional labeling data, the damage feature map is generated, which is used to represent the degree of damage to the radial gradient of the surrounding rock of the roadway.

[0013] Secondly, the present invention also provides a device for calculating the radial gradient damage degree of roadway surrounding rock, the device comprising: The first acquisition module is used to acquire raw data of damage factors within the target geological area, and the raw data is used to determine the degree of damage within the target geological area. The first generation module is used to preprocess the original data to generate a training dataset for calculating the degree of damage. The first processing module is used to perform layering processing on the target geological region according to the boundary conditions corresponding to the training dataset, and obtain the layering result, wherein the layering result includes at least one of the fractured zone, the plastic softening zone and the elastic zone; The second processing module is used to input the corresponding raw data in the training dataset into the improved BP neural network model for processing based on the hierarchical results, so as to obtain the damage data in the target geological area. The second generation module is used to generate a damage feature map based on the damage data.

[0014] Thirdly, the present invention provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps in the method for calculating the radial gradient damage degree of roadway surrounding rock provided by the present invention.

[0015] Fourthly, the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps in the method for calculating the radial gradient damage degree of roadway surrounding rock provided by the invention.

[0016] This invention acquires raw data of damage factors within a target geological area, which is used to determine the degree of damage within the target geological area. The raw data is preprocessed to generate a training dataset for calculating the degree of damage. Based on the boundary conditions corresponding to the training dataset, the target geological area is layered to obtain a layering result, which includes at least one of a fractured zone, a plastic softening zone, and an elastic zone. Based on the layering result, the corresponding raw data from the training dataset is input into an improved BP neural network model for processing to obtain damage data within the target geological area. Based on the damage data, a damage feature map is generated. Through the above method steps, the radial gradient damage degree of the surrounding rock in a roadway can be calculated and visualized in a layered manner, improving the accuracy and efficiency of damage assessment. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a method for calculating the radial gradient damage degree of roadway surrounding rock provided in an embodiment of the present invention; Figure 2 This is a radial gradient damage feature map of the surrounding rock of a roadway provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an improved BP neural network model according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of a device for calculating the radial gradient damage degree of roadway surrounding rock provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] like Figure 1 As shown, Figure 1 This is a flowchart illustrating a method for calculating the radial gradient damage level of roadway surrounding rock according to an embodiment of the present invention. The method includes the following steps: 101. Obtain raw data on damage factors within the target geological area.

[0021] In this embodiment of the invention, the above-mentioned method for calculating the radial gradient damage degree of the surrounding rock of the roadway can be applied to a calculation platform for the radial gradient damage degree of the surrounding rock of the roadway. The calculation platform for the radial gradient damage degree of the surrounding rock of the roadway has functions such as damage degree data processing, damage degree data transmission and reception, and damage degree data memory storage. It can be built based on a server or server cluster. The server or server cluster can be an electronic device with damage degree data processing capability.

[0022] The aforementioned target geological area can refer to the spatial area corresponding to the underground tunnel that needs to be assessed for surrounding rock damage, as selected by the calculation platform for the radial gradient damage degree of the surrounding rock of the tunnel. This area can be the entire cross-section of the surrounding rock of a specific tunnel, or it can be a local section of the surrounding rock selected along the tunnel axis.

[0023] Generally, the spatial geometric range of the target geological area can be obtained by analyzing and confirming the engineering design drawings or monitoring and positioning information. For example, the calculation platform for the radial gradient damage degree of the surrounding rock in the roadway can select a 20-m long section of surrounding rock in a coal mine transportation roadway as the target geological area. This area extends radially outward from the roadway centerline to the entire intact rock mass, serving as the object area for this damage calculation and assessment.

[0024] The aforementioned damage factors can refer to a set of physical parameters that can directly or indirectly reflect the stress state, deformation state, and material mechanical properties of the surrounding rock in the tunnel, thereby affecting the degree of damage to the surrounding rock. These parameters include, but are not limited to, four major categories: stress parameters, geometric parameters, rock mechanics parameters, and strain parameters. They are used to comprehensively characterize the stress conditions, spatial relationships, material constitutive properties, and plastic deformation behavior of the surrounding rock in the target geological area.

[0025] The aforementioned raw data can refer to the basic data set directly collected or tested by the calculation platform for the radial gradient damage degree of the surrounding rock in the aforementioned roadway, without any calculation processing. This includes, but is not limited to, the original rock stress, radial stress, and tangential stress reflecting the stress state of the surrounding rock; the radial distance between any point of the surrounding rock and the center of the roadway, and the roadway radius reflecting the spatial position of the surrounding rock; the elastic modulus, Poisson's ratio, internal friction angle, and cohesion reflecting the basic mechanical properties of rock materials; and the radial plastic strain increment and tangential plastic strain increment reflecting the plastic deformation behavior of the rock. It can be understood that the aforementioned raw data can be used as the basic input data for subsequent preprocessing, stratification, and damage calculation.

[0026] 102. Preprocess the raw data to generate a training dataset for calculating the degree of damage.

[0027] In this embodiment of the invention, the above-mentioned platform for calculating the radial gradient damage degree of the surrounding rock of the tunnel can perform physical property conversion and parameter construction processing on the original data before the original data enters the damage calculation process.

[0028] Specifically, the above-mentioned calculation platform for the radial gradient damage degree of the surrounding rock in the tunnel constructs rock strength criterion parameters based on the internal friction angle and cohesion in the original data, which are used to characterize the shear strength characteristics of the surrounding rock; at the same time, it constructs the expansion coefficient based on the radial plastic strain increment and tangential plastic strain increment in the original data, which are used to characterize the volume change capacity of the surrounding rock in the plastic deformation stage, thus transforming the scattered original data into characteristic parameters of corresponding physical properties.

[0029] The aforementioned training dataset can refer to a set of data constructed by the aforementioned calculation platform for radial gradient damage of the surrounding rock of the tunnel based on the preprocessing results, which is used for subsequent damage calculation model to perform calculations. It can be understood that the aforementioned training dataset not only includes the rock strength criterion parameters and expansion coefficients that have been preprocessed and constructed, but also simultaneously retains the corresponding stress parameters, geometric parameters and rock mechanical parameters, so that each set of training samples can completely characterize the mechanical state of the surrounding rock at a certain radial position in the target geological area.

[0030] For example, the original rock stress, surrounding rock radial stress, tangential stress, radial distance, tunnel radius, elastic modulus, Poisson's ratio, internal friction angle, cohesion, strength criterion parameters, and expansion coefficient corresponding to a certain radial position can be combined into a complete training sample. Multiple samples together constitute a training dataset for damage calculation.

[0031] 103. Based on the boundary conditions in the training dataset, the target geological region is divided into layers to obtain the layering results.

[0032] In this embodiment of the invention, the boundary conditions mentioned above may refer to the stress criteria used by the calculation platform for the radial gradient damage degree of the surrounding rock in the roadway to distinguish between the fractured zone, the plastic softening zone and the elastic zone when performing the surrounding rock stratification process. These criteria may be composed of stress interval thresholds, with different stress intervals corresponding to the surrounding rock being in a fractured state, a plastic softening state or an elastic state, respectively, and are used to provide a quantitative basis for subsequent stratification processes.

[0033] For example, the calculation platform for the radial gradient damage degree of the surrounding rock in the above-mentioned roadway can be set to determine the fracture zone when the tangential stress of the surrounding rock exceeds the peak strength threshold, the plastic softening zone when the stress of the surrounding rock is in the strength reduction range, and the elastic zone when the stress of the surrounding rock is below the yield threshold. The above stress range constitutes the boundary conditions used for the layered treatment of the surrounding rock.

[0034] In one possible embodiment, the above-mentioned calculation platform for radial gradient damage of the surrounding rock in the tunnel is based on boundary conditions to determine the stress state of the surrounding rock at each radial position in the target geological area, and then performs mechanical state zoning of the surrounding rock accordingly. The stress interval corresponding to each radial position is determined point by point, and the surrounding rock is divided into fracture zone, plastic softening zone or elastic zone according to the determination results.

[0035] The above-mentioned stratification results include at least one of the fractured zone, plastic softening zone, and elastic zone. It can be noted that the corresponding range of each stratified region in the radial direction of the target geological area can also be given, which can be used as the stratification basis for subsequent damage algorithm selection and damage data calculation. For example, the radial stratification results of 0 to 0.8 m from the center of the tunnel are the fractured zone, 0.8 to 1.6 m are the plastic softening zone, and beyond 1.6 m are the elastic zone. This stratification result is used to indicate the mechanical state of the surrounding rock at different radial positions.

[0036] 104. Based on the hierarchical results, the corresponding raw data in the training dataset is input into the improved BP neural network model for processing to obtain damage data in the target geological area.

[0037] In this embodiment of the invention, the improved BP neural network model can refer to the neural network model obtained by introducing an additional momentum term into the weight update algorithm of the basic BP neural network model after adjusting and training the calculation platform for radial gradient damage of the surrounding rock in the roadway. This improved BP neural network model considers both the current error gradient and the historical weight change trend during training, thereby enhancing the convergence stability and computational efficiency of the model when dealing with nonlinear mapping problems of surrounding rock damage. It is used for high-precision fitting and numerical mapping of the damage calculation results corresponding to different layers.

[0038] Furthermore, the improved BP neural network model described above can also serve as a data-driven numerical fitting model for nonlinear mapping and smoothing correction of damage calculation results obtained under different stratification conditions of the surrounding rock, thereby enabling the damage data to have better continuity and stability in the radial direction.

[0039] The aforementioned platform for calculating the radial gradient damage degree of the surrounding rock in the roadway can also input the data to be calculated selected based on the stratification results into the improved BP neural network model, and the model performs forward inference calculation. Through the neural network weight mapping relationship, the multi-parameter damage characteristics of the surrounding rock are nonlinearly fitted and transformed to obtain the damage data in the corresponding area.

[0040] The aforementioned damage data can refer to quantitative output results used to characterize the degree of damage to the surrounding rock at various radial locations within the target geological area. This damage data can be continuous numerical values, reflecting the degree of evolution of the surrounding rock from an intact state to a damaged state.

[0041] 105. Generate a damage feature map based on the damage data.

[0042] In this embodiment of the invention, the above-mentioned platform for calculating the radial gradient damage level of the surrounding rock of the tunnel is based on damage data and corresponding radial position information. It performs spatial mapping, partition labeling and result integration processing on the damage data to form a damage feature map that can be used for engineering display and analysis.

[0043] The aforementioned damage feature map can refer to a visualization result map formed by spatially mapping the damage data at each radial position based on the radial coordinate space of the target geological area, and combining the layering results to complete the semantic annotation of the fractured zone, plastic softening zone and elastic zone. It is used to intuitively represent the gradient damage distribution state of the surrounding rock of the roadway in the radial direction.

[0044] Specifically, the aforementioned platform for calculating the radial gradient damage level of the surrounding rock in the tunnel, after acquiring the damage data corresponding to each radial position, maps the damage data to the radial coordinate space of the target geological area based on the radial position parameters corresponding to the position to be calculated, forming damage distribution data along the radial direction of the tunnel. Simultaneously, the platform, combined with the layering results obtained from the layering process, performs regional corresponding annotations on the damage data at each radial position, ensuring a one-to-one correspondence between the damage data and the fractured zone, plastic softening zone, and elastic zone. Further, the platform spatially arranges the damage data at each radial position as discrete scatter points and connects them sequentially from the inside to the outside along the radial direction, ultimately generating data as shown below. Figure 2 The radial gradient damage characteristic diagram of the surrounding rock in the tunnel shown is as follows: the area near the inner wall of the tunnel corresponds to the fractured zone, which has a high degree of damage; it gradually transitions to the plastic softening zone, where the degree of damage gradually decreases; and it further transitions to the elastic zone, where the degree of damage is relatively low. This visually reflects the gradient distribution characteristics of the surrounding rock damage in the radial direction.

[0045] In this embodiment of the invention, raw data of damage factors within a target geological area are acquired, and this raw data is used to determine the degree of damage within the target geological area. The raw data is preprocessed to generate a training dataset for calculating the degree of damage. Based on the boundary conditions corresponding to the training dataset, the target geological area is layered to obtain a layering result, which includes at least one of a fractured zone, a plastic softening zone, and an elastic zone. Based on the layering result, the corresponding raw data from the training dataset is input into an improved BP neural network model for processing to obtain damage data within the target geological area. Based on the damage data, a damage feature map is generated. Through the above method steps, the radial gradient damage degree of the surrounding rock in a roadway can be calculated and visualized in a layered manner, improving the accuracy and efficiency of damage assessment.

[0046] Optionally, in the step of acquiring the raw data of damage factors within the target geological area, the stress state of the surrounding rock within the target geological area can also be collected to obtain stress parameters, including the original rock stress within the target geological area. s0. Radial stress of surrounding rock s r and tangential stress of surrounding rock s θ The geometric morphology of the tunnels and the spatial relationship between the surrounding rock within the target geological area are collected to obtain geometric parameters, including the radial distance between any point in the surrounding rock and the center of the tunnel. R and the radius of the alley a The mechanical properties of rock materials within the target geological area are tested to obtain rock mechanical parameters, including the elastic modulus. E Poisson's ratio u internal friction angle f and cohesion c The plastic deformation behavior of rocks within the target geological area was tested to obtain strain parameters, including the radial plastic strain increment during uniaxial compression of the rock. and tangential plastic strain increment .

[0047] In this embodiment of the invention, the above-mentioned raw data may include, but is not limited to, parameter data such as stress parameters, geometric parameters, rock mechanics parameters, and strain parameters.

[0048] The aforementioned stress state of the surrounding rock can be understood as the stress state under the combined action of the original rock stress field and the disturbance caused by tunnel excavation. It reflects the magnitude and distribution characteristics of the stress borne by the surrounding rock at different radial locations. This can be obtained by collecting data on the stress state of the surrounding rock within the target geological area, for example, by deploying monitoring equipment such as stress sensors and geostress testing devices to acquire real-time or periodic data on the mechanical state of the surrounding rock within the target geological area. For instance, in a mine transport tunnel, the stress state of the surrounding rock near the tunnel wall is significantly lower than that of the deeper surrounding rock far from the tunnel center due to unloading, and the stress state shows a significant change along the radial direction.

[0049] The aforementioned stress parameters can refer to quantitative data collected by the calculation platform for the radial gradient damage degree of the surrounding rock in the aforementioned tunnel, used to characterize the stress state of the surrounding rock, including the original rock stress within the target geological area. s 0. Radial stress of surrounding rock s r and tangential stress of surrounding rock s θ .

[0050] For example, at a certain radial measuring point, the platform collects the original rock stress. s 0 is 18 MPa, radial stress of surrounding rock s r The tangential stress of the surrounding rock is 6 MPa. s θThe stress is 22 MPa, and the above stress parameters can be used as the stress input data for this measuring point.

[0051] The aforementioned tunnel geometry and spatial relationship with the surrounding rock can refer to the shape and size characteristics of the tunnel cross section and the spatial relationship of any calculated point in the surrounding rock relative to the tunnel center within the target geological area. This is used to reflect the spatial distribution characteristics of the surrounding rock in the radial direction. For example, if a tunnel cross section is circular and the tunnel diameter is 4 m, then the tunnel radius is 2 m. Different measuring points in the surrounding rock at distances of 0.8 m, 1.5 m, and 2.5 m from the tunnel center correspond to different radial positional relationships.

[0052] The aforementioned geometric parameters can refer to quantitative parameters obtained by the calculation platform for radial gradient damage of the surrounding rock of the roadway after collecting data on the roadway geometry and the spatial relationship between the surrounding rock and the roadway, including the radial distance R between any point of the surrounding rock and the center of the roadway and the roadway radius a.

[0053] The aforementioned mechanical properties of rock materials can refer to the basic mechanical behavior characteristics exhibited by the surrounding rocks in the aforementioned target geological area under stress, which are used to characterize the ability of rocks to resist deformation and failure. For example, sandstone, limestone and mudstone have significant differences in elastic modulus, Poisson's ratio, internal friction angle and cohesion due to differences in composition and structure. These differences constitute the mechanical properties of different rock materials.

[0054] In this embodiment, the above-mentioned platform for calculating the radial gradient damage degree of the surrounding rock of the tunnel can be used to conduct standard mechanical tests on rock samples obtained from the target geological area using laboratory rock sample testing equipment to obtain the mechanical properties of the rock material.

[0055] The aforementioned rock mechanics parameters can be understood as a set of quantitative parameters obtained by the calculation platform for the radial gradient damage degree of the surrounding rock of the tunnel after testing the mechanical properties of the rock material, including the elastic modulus. E Poisson's ratio u internal friction angle f and cohesion c This data can be used as important input data for subsequent structural strength criterion parameters.

[0056] For example, the elastic modulus E of a certain sandstone sample is 28 GPa, and the Poisson's ratio is... u The internal friction angle is 0.24. f The angle is 34°, and the cohesion is... c It is 2.1 MPa.

[0057] The aforementioned rock plastic deformation behavior can refer to the irreversible deformation process of rocks in the target geological area after the stress exceeds the elastic limit. It is used to reflect the deformation characteristics of the surrounding rock from the elastic stage to the plastic stage under compression.

[0058] In this embodiment, the above-mentioned calculation platform for the radial gradient damage degree of the surrounding rock of the tunnel can also be used to perform loading experiments on the rock sample through a uniaxial compression test device to obtain the radial and tangential deformation data of the rock in the plastic stage.

[0059] Specifically, during the uniaxial compression test, the radial and tangential plastic strain increments of the rock specimen during the loading process are simultaneously measured using strain gauges.

[0060] The aforementioned strain parameters can refer to quantitative parameters collected by the calculation platform for the radial gradient damage degree of the surrounding rock in the aforementioned tunnel during the testing of rock plastic deformation behavior. These parameters include the radial and tangential plastic strain increments during the uniaxial compression process of the rock, which are used to subsequently construct the rock dilatancy coefficient and participate in damage calculation. For example, in a certain uniaxial compression test, the radial plastic strain increment of the rock sample was 3.2 × 10⁻⁶. - ³, the tangential plastic strain increment is 1.1 × 10⁻⁶. - ³, the above strain parameters will be used as input data for calculating the expansion coefficient.

[0061] Optionally, the step of preprocessing the original data to generate a training dataset for damage degree calculation further includes performing a first preprocessing on the original data based on preset rock strength criterion parameters to obtain a first preprocessing result; performing a second preprocessing on the original data based on the uniaxial compression test results with a preset expansion coefficient to obtain a second preprocessing result; and generating a training dataset based on the first preprocessing result and the second preprocessing result.

[0062] In this embodiment of the invention, the aforementioned preset rock strength criterion parameter may refer to the calculation platform for the radial gradient damage degree of the surrounding rock of the tunnel based on the internal friction angle of the rock. f and cohesion c A pre-constructed set of parameters is used to characterize the resistance of surrounding rock to damage, and to reflect the strength characteristics of rock under different stress states.

[0063] Specifically, it can be constructed using the following formula:

[0064]

[0065] Furthermore, by applying the above formula to unify the physical constraints and feature transformations of the stress parameters and rock mechanics parameters in the original data, the corresponding first preprocessing result is obtained. This refers to the intermediate data result output by the calculation platform for the radial gradient damage degree of the surrounding rock of the tunnel after completing the first preprocessing. This result is feature data constrained by the rock strength criterion, used to characterize the relative damage trend of the surrounding rock at different radial positions. For example, after performing the first preprocessing on the stress parameters at a certain radial position, a feature value reflecting the degree of near-damage of the surrounding rock at that position under the current strength state can be obtained, which serves as the input basis for subsequent damage calculations.

[0066] The aforementioned pre-set expansion coefficient can refer to the parameter pre-constructed by the calculation platform for the radial gradient damage degree of the surrounding rock in the aforementioned roadway, based on the relationship between the radial plastic strain increment and the tangential plastic strain increment obtained during the uniaxial compression test of the rock, to characterize the volume change characteristics of the rock in the plastic stage, and to reflect the expansion capacity of the surrounding rock when plastic deformation occurs.

[0067] Specifically, it can be obtained through the following formula:

[0068] in, —Increment of radial plastic strain during uniaxial compression of rock; —Increment of tangential plastic strain during uniaxial compression of rock.

[0069] Furthermore, the strain parameters reflecting plastic deformation characteristics in the original data can be refined and uniformly scaled using the above formula to obtain the second preprocessing result. This refers to the intermediate data result output by the above-mentioned calculation platform for radial gradient damage degree of roadway surrounding rock after completing the second preprocessing. This result is strain characteristic data with constraints of rock plastic expansion characteristics introduced, used to reflect the volumetric deformation development law of surrounding rock in the plastic stage.

[0070] In one possible embodiment, the above-mentioned platform for calculating the radial gradient damage level of the surrounding rock of the tunnel is based on the first preprocessing result and the second preprocessing result, and fuses, registers and organizes the two types of intermediate feature data to form a standardized dataset for subsequent damage level calculation and model training.

[0071] Specifically, each sample in the above training dataset corresponds to the comprehensive characterization data of the surrounding rock at a certain radial position in the target geological area under the dual constraints of strength characteristics and plastic deformation characteristics.

[0072] For example, the strength feature data obtained by the first preprocessing at a certain radial measuring point can be combined with the plastic expansion feature data obtained by the second preprocessing to form a complete training sample, which together with the data at other radial locations constitutes the training dataset.

[0073] By employing the above methods and steps, the obtained data satisfies the constraints of the surrounding rock mechanics mechanism, thereby reducing the impact of the original data discreteness on the damage calculation results and improving the stability, accuracy, and convergence efficiency of subsequent training of the improved BP neural network model and damage calculation.

[0074] Optionally, the step of performing layered processing on the target geological region according to the boundary conditions in the training dataset to obtain the layered results further includes determining the preset stress conditions corresponding to the fractured zone, the plastic softening zone, and the elastic zone; judging the stress interval of the target geological region based on the preset stress conditions, and dividing the range of the corresponding region from the target geological region; and dividing the corresponding fractured zone, the plastic softening zone, and the elastic zone from the target geological region based on the range of each region.

[0075] In this embodiment of the invention, the above-mentioned calculation platform for the radial gradient damage degree of the surrounding rock of the tunnel can be pre-set according to the mechanical parameters of the surrounding rock material and the rock strength criterion parameters obtained by preprocessing, and can be used to distinguish the stress discrimination threshold conditions for the fracture zone, the plastic softening zone and the elastic zone. Generally speaking, in engineering design, it can also be uniformly set based on the mechanical properties of the rock material such as peak strength, yield strength and residual strength.

[0076] For example, the stress range corresponding to the tangential stress of the surrounding rock being greater than the peak strength of the rock can be used as the preset stress condition for the fracture zone; the range of surrounding rock stress between the yield strength and the peak strength can be used as the preset stress condition for the plastic softening zone; and the range of surrounding rock stress less than the yield strength can be used as the preset stress condition for the elastic zone.

[0077] The aforementioned preset stress conditions can refer to a set of stress threshold rules that have been pre-constructed for classifying the mechanical state of the surrounding rock. These stress conditions exist in the form of stress intervals and serve as the basis for subsequent stress interval judgments.

[0078] Specifically, the stress conditions for the corresponding layers of the three layered results can be set using the following formula: The stress condition in the elastic zone is satisfied as follows:

[0079] In the formula, s r The radial stress experienced by the surrounding rock unit; s θ The tangential stress experienced by the surrounding rock unit; s0 represents the original rock stress.

[0080] The stress conditions in the plastic softening zone are satisfied as follows:

[0081] In the formula, s r The radial stress experienced by the surrounding rock unit; s θ The tangential stress experienced by the surrounding rock unit; s 0 represents the original rock stress; M and N are the rock strength criterion parameters; χ is the rock dilatancy coefficient; Rs is the radius of the strain softening zone; and R is the distance between any point in the surrounding rock and the center point of the roadway.

[0082] The stress conditions in the fractured zone are satisfied as follows:

[0083] In the formula, s r The radial stress experienced by the surrounding rock unit; s θ χ represents the tangential stress on the surrounding rock unit; M and N are the rock strength criterion parameters; χ is the rock dilatancy coefficient; R is the distance between any point in the surrounding rock and the center point of the tunnel; a is the tunnel radius.

[0084] In one possible embodiment, the above-mentioned calculation platform for the radial gradient damage degree of the surrounding rock in the tunnel compares the surrounding rock stress parameters collected at each radial position in the target geological area with the corresponding preset stress conditions point by point, thereby determining the range to which the surrounding rock stress at each position belongs.

[0085] For example, when judging the surrounding rock measuring point 0.6 m from the center of the tunnel, the tangential stress value of the surrounding rock at that point is read and compared with three preset stress intervals. If the value falls into the stress interval corresponding to the fractured zone, the measuring point is determined to belong to the fractured zone.

[0086] In another possible embodiment, the above-mentioned calculation platform for radial gradient damage degree of the surrounding rock of the tunnel, based on the result of stress interval judgment, aggregates the spatial locations within the target geological area that meet the same stress interval judgment result, thereby forming the area range corresponding to different mechanical states.

[0087] For example, after determining the stress range of multiple radial measuring points, the spatial locations corresponding to all measuring points identified as fracture zones are summarized to form the preliminary range of the fracture zone; the same method is used to aggregate the regions of the plastic softening zone and the elastic zone.

[0088] The aforementioned area can refer to the spatial distribution range of each mechanical zone in the radial direction of the target geological area, which is finally determined by the calculation platform for the radial gradient damage degree of the surrounding rock of the aforementioned roadway after the division process is completed. This is used to clarify the specific location of different layers in the surrounding rock.

[0089] For example, it can be finally determined that: the range of 0 to 0.7 m from the center of the roadway corresponds to the fracture zone; the range of 0.7 to 1.5 m corresponds to the plastic softening zone; and the range beyond 1.5 m corresponds to the elastic zone. The above radial range constitutes the range of each layered area.

[0090] In this embodiment, after obtaining the range of each region, the calculation platform for the radial gradient damage degree of the surrounding rock in the tunnel ultimately forms three types of stratification results within the target geological area: a fractured zone, a plastic softening zone, and an elastic zone, according to the range of the regions. These stratification results not only characterize the differences in the mechanical state of the surrounding rock at different radial positions but also directly serve as the stratification basis for subsequent selection of stratified damage formulas and calculation of damage data.

[0091] Optionally, before inputting the corresponding raw data from the training dataset into the improved BP neural network model for processing based on the hierarchical results to obtain damage data within the target geological area, the steps also include: obtaining a basic BP neural network model; adjusting the weight algorithm of the basic BP neural network model according to a preset additional momentum term to obtain new weight update data; and updating and adjusting the basic BP neural network model based on the new weight update data to obtain an improved BP neural network model after training is completed.

[0092] In this embodiment of the invention, the aforementioned preset additional momentum term refers to the historical weight change weight parameter pre-set by the calculation platform for the radial gradient damage degree of the surrounding rock in the roadway when constructing the improved BP neural network model. This parameter is used to characterize the proportion of the weight change trend inherited from the previous round during the current weight update process. By setting the additional momentum term, the network can respond to the current error signal while maintaining a certain update inertia during training, thereby improving the oscillation problem during network training. For example, the additional momentum term can be set to 0.8, indicating that 80% of the current weight update is inherited from the previous weight update direction, and 20% comes from the current error information.

[0093] Furthermore, this can be illustrated using the following adjustment formula for the BP weight algorithm of the improved BP neural network:

[0094] in, or For learning factors; t Number of times; αThe potential factor, i.e. the aforementioned preset additional momentum term, determines the degree of influence of the previous learned weight change on the current weight update; Q This represents the sample output error. w ij For nodes i and j The weights between them.

[0095] The aforementioned weighting algorithm can refer to the update calculation rule used by the calculation platform for radial gradient damage degree of the surrounding rock of the roadway in the basic BP neural network model to adjust the connection weights of each layer of the network. The algorithm is based on the backpropagation relationship between the network output error and the input features, and gradually corrects the connection weights between the hidden layer and the output layer in order to realize the network's learning of the mapping relationship between "input features - damage data".

[0096] In one possible embodiment, the above-mentioned calculation platform for the radial gradient damage degree of the surrounding rock of the roadway introduces a preset additional momentum term into the weight update rule on the basis of the original basic BP neural network weight algorithm, so that the weight update process, which was originally determined only by the current error signal, becomes an update jointly determined by the "current error signal + historical weight change trend".

[0097] In other words, the weight changes from the previous round are superimposed during the original weight update process, so that the direction of the current weight adjustment depends not only on the current calculation error, but also on the direction of the previous adjustment, thus making the network training process smoother and more stable.

[0098] The final weight update data can refer to the weight change data calculated by the above-mentioned calculation platform for radial gradient damage degree of roadway surrounding rock after introducing additional momentum term and completing weight algorithm adjustment, according to the adjusted weight update rule. This data is used to indicate the specific magnitude and direction of the adjustment of the connection weights of each layer in the neural network in the current training round.

[0099] For example, if the updated data of the connection weight corresponding to a certain hidden layer node is calculated to be +0.015 in a certain training round, it means that the connection weight should be increased by 0.015 in this training round.

[0100] In another possible embodiment, the above-mentioned platform for calculating the radial gradient damage degree of the surrounding rock of the roadway performs multiple rounds of iterative updates on each connection weight in the basic BP neural network model based on the weight update data, and performs forward calculation and backward update processes by repeatedly inputting training dataset samples, so that the neural network gradually approaches the stable mapping relationship between "multi-source characteristic parameters of surrounding rock - damage data".

[0101] Specifically, multiple rounds of cyclic training can be performed on sample data at different radial positions in the training dataset. In each round of training, the network weights are corrected based on the latest weight update data until the network output error converges to a preset threshold range.

[0102] The improved BP neural network model described above can also be achieved through methods such as... Figure 3 The model structure shown is illustrated, comprising an input layer, a hidden layer, a computational layer, and an output layer. The input layer has 11 nodes, each corresponding to one of the 11 influencing factors. The hidden layer has 8 nodes: rock strength criterion parameters M and N, the rock's dilatancy coefficient χ, three stress conditions for regional stratification, the radius of the strain softening zone Rs, and the radius of the fractured zone Rb. The computational layer has 3 nodes: formulas for calculating the radial gradient damage degree of the surrounding rock in the fractured zone, plastic softening zone, and elastic stratification zone of the roadway. The output layer has 1 node: the radial gradient damage value of the surrounding rock in the roadway.

[0103] That is, after completing the introduction of a preset additional momentum term, the execution of weight algorithm adjustment, and the completion of update adjustment training, the final neural network model for damage data calculation is formed.

[0104] Optionally, the step of inputting the corresponding raw data from the training dataset into the improved BP neural network model for processing based on the stratification results to obtain damage data within the target geological area further includes: determining the stratification type corresponding to the location to be calculated within the target geological area based on the stratification results; selecting raw data corresponding to the stratification type from the training dataset as the data to be calculated based on the stratification type; inputting the data to be calculated into the improved BP neural network model for processing; and calculating the damage data corresponding to the location to be calculated based on the damage algorithm corresponding to the stratification type.

[0105] In this embodiment of the invention, the above-mentioned layering type can refer to the mechanical state category marked by the calculation platform for the radial gradient damage degree of the surrounding rock in the roadway according to the layering processing result, for different radial positions of the surrounding rock in the target geological area, which is used to characterize the stress and deformation stage of the surrounding rock at that position. Specifically, it includes the fractured layering corresponding to the fractured zone, the softening layering corresponding to the plastic softening zone, and the elastic layering corresponding to the elastic zone.

[0106] For example, if the surrounding rock within 0 to 0.7 m from the center of the tunnel is determined to be a fractured zone, then the stratification type corresponding to any calculation point within this range is "fractured stratification"; when the surrounding rock at 1.2 m from the center of the tunnel is classified as a plastic softening zone, the stratification type corresponding to this location is "softening stratification".

[0107] In one possible embodiment, before performing damage calculation at a specific location, the platform for calculating the radial gradient damage of the surrounding rock in the tunnel compares and confirms the current stratification type of the location based on the radial coordinates corresponding to that location and the already formed stratification results, thereby determining what stratification type the geological structure belongs to. For example, when the platform receives a radial distance R=0.5 m for a certain measurement point to be calculated, the platform matches the radial position with the interval range in the stratification results. If the position falls within the radial range of the fractured zone, the stratification type of the location to be calculated is determined to be a fractured stratification.

[0108] Furthermore, the following steps can be used to distinguish them: Discrimination R s ≤ R If yes, then the surrounding rock stratification belongs to the fractured stratification type; if not, then determine... R b ≤R<R s If yes, then the surrounding rock stratification belongs to the softening type; if not, then determine... a ≤ R < R b If so, then the surrounding rock stratification belongs to the elastic type.

[0109] The aforementioned location to be calculated may refer to a specific spatial location point selected by the calculation platform for the radial gradient damage degree of the surrounding rock of the roadway during the damage calculation process, which requires separate calculation of the damage degree. This location corresponds to a certain radial coordinate value within the target geological area.

[0110] The aforementioned data to be calculated can refer to a set of raw data selected from the training dataset by the calculation platform for radial gradient damage degree of the surrounding rock of the roadway after completing the determination of the layer type. This data matches the current location to be calculated and its corresponding layer type and is used as the input data for damage calculation at that location. For example, when a location to be calculated is determined to belong to the plastic softening zone, the platform selects the stress parameters, geometric parameters, rock mechanics parameters, first preprocessing results, and second preprocessing results corresponding to that location from the training dataset and inputs them into the subsequent model as the data to be calculated for that location.

[0111] The aforementioned damage algorithm can be a pre-defined calculation model or mathematical expression for different stratification types, used by the calculation platform for the radial gradient damage degree of the surrounding rock in the aforementioned roadway, to describe the evolution law of surrounding rock damage. Different stratification types correspond to different damage algorithms, used to reflect the differences in damage mechanisms among the three types of surrounding rock: fractured zone, plastic softening zone, and elastic zone. Specifically, it can be defined by the following formula: After determining the stratification type as described above, if the surrounding rock stratification is of the fractured type, the loss data can be calculated using the following formula:

[0112] in, D This represents the radial gradient damage value of the surrounding rock in the tunnel. s 0 represents the original rock stress; R s Let be the radius of the strain softening region. R is the distance between any point in the surrounding rock and the center point of the tunnel; M and N These are parameters for rock strength criteria. x is the dilatancy coefficient of the rock; u The Poisson's ratio for rocks.

[0113] If the surrounding rock stratification is of the softening type, the loss data can be calculated using the following formula:

[0114] in, D This represents the radial gradient damage value of the surrounding rock in the tunnel. s 0 represents the original rock stress; R b The radius of the fracture zone; R s Let be the radius of the strain softening region. R is the distance between any point in the surrounding rock and the center point of the tunnel; M and N The rock strength criterion parameters can be calculated using S21; x The dilatation factor of the rock can be calculated using S22; u Poisson's ratio of the rock; s r e-s This refers to the radial stress at the elastoplastic interface.

[0115] If the surrounding rock stratification is elastic, the loss data can be calculated using the following formula:

[0116] in, D This represents the radial gradient damage value of the surrounding rock in the tunnel. s 0 represents the original rock stress; R b The radius of the fracture zone; a The radius of the alleyway, R is the distance between any point in the surrounding rock and the center point of the tunnel; M and N These are parameters for rock strength criteria. x is the dilatancy coefficient of the rock; u Poisson's ratio of the rock; s r e-s This refers to the radial stress at the elastoplastic interface.

[0117] Optionally, the step of generating a damage feature map based on damage data further includes mapping the damage data to the radial coordinate space of the target geological area based on the radial position parameters corresponding to the location to be calculated, to obtain damage distribution data; according to the stratification results, the damage data is labeled according to the fractured zone, plastic softening zone and elastic zone, to obtain regional labeling data; and based on the damage distribution data and the corresponding regional labeling data, a damage feature map is generated, which is used to represent the degree of damage of the radial gradient of the surrounding rock of the roadway.

[0118] In this embodiment of the invention, the aforementioned radial position parameter can refer to the radial spatial coordinate parameter determined by the calculation platform for the radial gradient damage degree of the surrounding rock of the tunnel for each location to be calculated within the target geological area, used to characterize the location relative to the center of the tunnel. This parameter is usually represented by the radial distance R between any point in the surrounding rock and the center of the tunnel. For example, if the radial distance from a location to be calculated to the center of the tunnel is R = 0.6 m, then the radial position parameter corresponding to that location is 0.6 m, used to identify the specific position of the damage data on the radial coordinate axis.

[0119] The aforementioned damage distribution data can refer to the set of damage data with spatial location correspondence formed by the calculation platform for the radial gradient damage degree of the surrounding rock of the roadway, which maps the damage data corresponding to multiple locations to be calculated to the radial coordinate space of the target geological area based on the radial position parameters.

[0120] The aforementioned area labeling data refers to the labeling results generated by the calculation platform for the radial gradient damage degree of the surrounding rock in the aforementioned roadway, which assigns corresponding layer category labels to the damage data at each radial position in the damage distribution data based on the layering results. This labeling is used to indicate the mechanical zoning type to which the damage data belongs. For example, when the damage data at R=0.4 m is labeled as "fractured zone", the damage data at R=1.1 m is labeled as "plastic softening zone", and the damage data at R=1.8 m is labeled as "elastic zone", the above labeling information constitutes the area labeling data.

[0121] In one possible embodiment, the aforementioned platform for calculating the radial gradient damage level of the surrounding rock of the tunnel is based on the aforementioned damage distribution data and corresponding regional annotation data, and performs spatial visualization mapping and partition rendering processing on the damage data to form a graphical result for characterizing the radial gradient damage level of the surrounding rock of the tunnel.

[0122] For example, the platform displays the damage data corresponding to the fractured zone as a high-damage-level area, the damage data corresponding to the plastic softening zone as a medium-damage-level area, and the damage data corresponding to the elastic zone as a low-damage-level area, and arranges them from the inside to the outside according to their radial position, thereby generating a complete radial gradient damage feature map of the roadway surrounding rock.

[0123] like Figure 4 As shown, this embodiment of the invention also provides a calculation device 400 for the radial gradient damage degree of the surrounding rock of a roadway. The calculation device 400 for the radial gradient damage degree of the surrounding rock of the roadway includes: The first acquisition module 401 is used to acquire raw data of damage factors within the target geological area, and the raw data is used to determine the degree of damage within the target geological area. The first generation module 402 is used to preprocess the original data to generate a training dataset for calculating the degree of damage. The first processing module 403 is used to perform layering processing on the target geological region according to the boundary conditions corresponding to the training dataset to obtain a layering result, wherein the layering result includes at least one of the fractured zone, the plastic softening zone and the elastic zone. The second processing module 404 is used to input the corresponding raw data in the training dataset into the improved BP neural network model for processing based on the hierarchical results, so as to obtain the damage data in the target geological area. The second generation module 405 is used to generate a damage feature map based on the damage data.

[0124] Optionally, the first acquisition module 401 mentioned above includes: The first acquisition submodule is used to collect the stress state of the surrounding rock within the target geological area and obtain stress parameters, including the original rock stress within the target geological area. s 0. Radial stress of surrounding rock s r and tangential stress of surrounding rock s θ ; The second acquisition submodule is used to collect the geometric morphology of the tunnel and the spatial relationship between the surrounding rock within the target geological area, and obtain geometric parameters, including the radial distance between any point in the surrounding rock and the center of the tunnel. R and the radius of the alley a ; The third acquisition submodule is used to test the mechanical properties of rock materials within the target geological area to obtain rock mechanical parameters, including the elastic modulus. E Poisson's ratio u internal friction angle f and cohesionc ; The fourth acquisition submodule is used to test the plastic deformation behavior of rocks within the target geological area and obtain strain parameters, including the radial plastic strain increment during uniaxial compression of the rock. and tangential plastic strain increment .

[0125] Optionally, the first generation module 402 mentioned above includes: The first generation submodule is used to perform a first preprocessing on the original data based on preset rock strength criterion parameters to obtain a first preprocessing result; The second generation submodule is used to perform a second preprocessing on the original data based on the uniaxial compression test results with a preset expansion coefficient, so as to obtain a second preprocessing result. The third generation submodule is used to generate the training dataset based on the first preprocessing result and the second preprocessing result.

[0126] Optionally, the first processing module 403 mentioned above includes: The first processing submodule is used to determine the preset stress conditions corresponding to the fracture zone, plastic softening zone and elastic zone; The second processing submodule is used to determine the stress range of the target geological area based on the preset stress conditions, and to divide the range of the corresponding area from the target geological area. The third processing submodule is used to divide the target geological region into corresponding fractured zones, plastic softening zones, and elastic zones based on the range of each of the regions.

[0127] Optionally, the above-mentioned device further includes: The first building block is used to obtain the basic BP neural network model; The second construction module is used to adjust the weight algorithm of the basic BP neural network model according to the preset additional momentum term to obtain new weight update data; The third construction module is used to update and adjust the basic BP neural network model based on the new weight update data, and obtain an improved BP neural network model after training is completed.

[0128] Optionally, the second processing module 404 mentioned above includes: The fourth processing submodule is used to determine the layer type corresponding to the location to be calculated within the target geological area based on the layering results; The fifth processing submodule is used to select the original data corresponding to the hierarchical type from the training dataset as the data to be calculated, according to the hierarchical type. The sixth processing submodule is used to input the data to be calculated into the improved BP neural network model for processing, and to calculate the damage data corresponding to the location to be calculated based on the damage algorithm corresponding to the layer type.

[0129] Optionally, the second generation module 405 mentioned above includes: The fourth generation submodule is used to map the damage data to the radial coordinate space of the target geological area based on the radial position parameters corresponding to the location to be calculated, so as to obtain damage distribution data. The fifth generation submodule is used to perform region-based labeling of the damage data according to the layering results, namely the fracture zone, the plastic softening zone, and the elastic zone, to obtain region-labeled data; The sixth generation submodule is used to generate the damage feature map based on the damage distribution data and the corresponding regional annotation data. The damage feature map is used to represent the degree of damage of the radial gradient of the surrounding rock of the roadway.

[0130] like Figure 5 As shown, this embodiment of the invention also provides an electronic device 500, including a processor, which can execute any of the above-mentioned methods for calculating the radial gradient damage degree of roadway surrounding rock.

[0131] Specifically, it includes a processor 501 and a memory 502, as well as a computer program stored in the memory 502 and capable of running on the processor 501, which executes a method for calculating the radial gradient damage degree of the surrounding rock in the roadway, wherein: The processor 501 executes the calculator program storing the calculation method for the radial gradient damage degree of the surrounding rock in the roadway in memory 502, and performs the following steps: Obtain raw data of damage factors within the target geological area; the raw data is used to determine the degree of damage within the target geological area. The raw data is preprocessed to generate a training dataset for calculating the degree of damage; Based on the boundary conditions in the training dataset, the target geological region is subjected to stratification to obtain a stratification result, which includes at least one of the fractured zone, plastic softening zone, and elastic zone. Based on the hierarchical results, the corresponding raw data in the training dataset is input into the improved BP neural network model for processing to obtain damage data within the target geological area; Based on the damage data, a damage feature map is generated.

[0132] Optionally, the processor 501 executes the raw data, including stress parameters, geometric parameters, rock mechanics parameters, and strain parameters. The acquisition of raw data on damage factors within the target geological area includes: The stress state of the surrounding rock within the target geological area is collected to obtain stress parameters, including the original rock stress within the target geological area. s 0. Radial stress of surrounding rock s r and tangential stress of surrounding rock s θ ; The geometric morphology of the tunnels and the spatial relationship between the surrounding rock within the target geological area are collected to obtain geometric parameters, including the radial distance between any point in the surrounding rock and the center of the tunnel. R and the radius of the alley a ; The mechanical properties of rock materials within the target geological region are tested to obtain rock mechanical parameters, including the elastic modulus. E Poisson's ratio u internal friction angle f and cohesion c ; The plastic deformation behavior of rocks within the target geological region was tested to obtain strain parameters, which include the radial plastic strain increment during uniaxial compression of the rock. and tangential plastic strain increment .

[0133] Optionally, the processor 501 performs the preprocessing of the raw data to generate a training dataset for damage degree calculation, including: Based on preset rock strength criterion parameters, the original data is subjected to a first preprocessing to obtain a first preprocessing result; Based on the uniaxial compression test results with a preset expansion coefficient, the original data is subjected to a second preprocessing to obtain a second preprocessing result; The training dataset is generated based on the first preprocessing result and the second preprocessing result.

[0134] Optionally, the processor 501 performs the layering process on the target geological region according to the corresponding boundary conditions in the training dataset to obtain the layering result, including: Determine the preset stress conditions corresponding to the fracture zone, plastic softening zone, and elastic zone; Based on the preset stress conditions, the stress range of the target geological area is determined, and the range of the corresponding area is delineated from the target geological area. Based on the extent of each region, corresponding fractured zones, plastic softening zones, and elastic zones are divided within the target geological region.

[0135] Optionally, before the processor 501 executes the step of inputting the corresponding raw data from the training dataset into the improved BP neural network model for processing based on the hierarchical results to obtain the damage data within the target geological area, the method further includes: Obtain the basic BP neural network model; Based on the preset additional momentum term, the weight algorithm of the basic BP neural network model is adjusted to obtain new weight update data; Based on the new weight update data, the basic BP neural network model is updated, adjusted, and trained to obtain an improved BP neural network model after training is completed.

[0136] Optionally, the processor 501 executes the step of inputting the corresponding raw data from the training dataset into the improved BP neural network model based on the hierarchical results for processing, to obtain damage data within the target geological area, including: Based on the stratification results, determine the stratification type corresponding to the location to be calculated within the target geological area; Based on the hierarchical type, select the original data corresponding to the hierarchical type from the training dataset as the data to be calculated; The data to be calculated is input into the improved BP neural network model for processing, and the damage data corresponding to the location to be calculated is obtained according to the damage algorithm corresponding to the layer type.

[0137] Optionally, the processor 501 performs the step of generating a damage feature map based on the damage data, including: Based on the radial position parameters corresponding to the location to be calculated, the damage data is mapped to the radial coordinate space of the target geological area to obtain damage distribution data; Based on the layering results, the damage data is labeled according to the fracture zone, plastic softening zone and elastic zone to obtain the region labeling data; Based on the damage distribution data and the corresponding regional labeling data, the damage feature map is generated, which is used to represent the degree of damage to the radial gradient of the surrounding rock of the roadway.

[0138] This invention also provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the various processes of the method for calculating the radial gradient damage degree of roadway surrounding rock provided in this invention, or the method for calculating the radial gradient damage degree of roadway surrounding rock in an application, and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0139] Those skilled in the art will understand that implementing all or part of the processes in the above embodiments can be done by a computer program instructing related hardware, and can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0140] The above description discloses only preferred embodiments of the present invention and should not be construed as limiting the scope of the present invention. Therefore, equivalent variations made in accordance with the claims of the present invention are still within the scope of the present invention.

Claims

1. A method for calculating the radial gradient damage degree of surrounding rock in a roadway, characterized in that, include: Obtain raw data of damage factors within the target geological area; the raw data is used to determine the degree of damage within the target geological area. The raw data is preprocessed to generate a training dataset for calculating the degree of damage; Based on the boundary conditions in the training dataset, the target geological region is subjected to stratification to obtain a stratification result, which includes at least one of the fractured zone, plastic softening zone, and elastic zone. Based on the hierarchical results, the corresponding raw data in the training dataset is input into the improved BP neural network model for processing to obtain damage data within the target geological area; Based on the damage data, a damage feature map is generated.

2. The method for calculating the radial gradient damage degree of the surrounding rock of a roadway as described in claim 1, characterized in that, The raw data includes stress parameters, geometric parameters, rock mechanics parameters, and strain parameters. The acquisition of raw data on damage factors within the target geological area includes: The stress state of the surrounding rock within the target geological area is collected to obtain stress parameters, including the original rock stress within the target geological area. σ 0. Radial stress of surrounding rock σ r and tangential stress of surrounding rock σ θ ; The geometric morphology of the tunnels and the spatial relationship between the surrounding rock within the target geological area are collected to obtain geometric parameters, including the radial distance between any point in the surrounding rock and the center of the tunnel. R and the radius of the alley a ; The mechanical properties of rock materials within the target geological region are tested to obtain rock mechanical parameters, including the elastic modulus. E Poisson's ratio υ internal friction angle φ and cohesion c ; The plastic deformation behavior of rocks within the target geological region was tested to obtain strain parameters, which included the radial plastic strain increment during uniaxial compression of the rock. and tangential plastic strain increment .

3. The method for calculating the radial gradient damage degree of the surrounding rock of a roadway as described in claim 1, characterized in that, The preprocessing of the original data to generate a training dataset for damage degree calculation includes: Based on preset rock strength criterion parameters, the original data is subjected to a first preprocessing to obtain a first preprocessing result; Based on the uniaxial compression test results with a preset expansion coefficient, the original data is subjected to a second preprocessing to obtain a second preprocessing result; The training dataset is generated based on the first preprocessing result and the second preprocessing result.

4. The method for calculating the radial gradient damage degree of the surrounding rock of a roadway as described in claim 1, characterized in that, The step of performing stratification processing on the target geological region based on the corresponding boundary conditions in the training dataset to obtain stratification results includes: Determine the preset stress conditions corresponding to the fracture zone, plastic softening zone, and elastic zone; Based on the preset stress conditions, the stress range of the target geological area is determined, and the range of the corresponding area is delineated from the target geological area. Based on the extent of each region, corresponding fractured zones, plastic softening zones, and elastic zones are divided within the target geological region.

5. The method for calculating the radial gradient damage degree of the surrounding rock of a roadway as described in claim 1, characterized in that, Before inputting the corresponding raw data from the training dataset into the improved BP neural network model for processing based on the hierarchical results to obtain the damage data within the target geological area, the method further includes: Obtain the basic BP neural network model; Based on the preset additional momentum term, the weight algorithm of the basic BP neural network model is adjusted to obtain new weight update data; Based on the new weight update data, the basic BP neural network model is updated, adjusted, and trained to obtain an improved BP neural network model after training is completed.

6. The method for calculating the radial gradient damage degree of the surrounding rock of a roadway as described in claim 5, characterized in that, Based on the hierarchical results, the corresponding raw data in the training dataset is input into an improved BP neural network model for processing to obtain damage data within the target geological area, including: Based on the stratification results, determine the stratification type corresponding to the location to be calculated within the target geological area; Based on the hierarchical type, select the original data corresponding to the hierarchical type from the training dataset as the data to be calculated; The data to be calculated is input into the improved BP neural network model for processing, and the damage data corresponding to the location to be calculated is obtained according to the damage algorithm corresponding to the layer type.

7. The method for calculating the radial gradient damage degree of the surrounding rock of a roadway as described in claim 6, characterized in that, The step of generating a damage feature map based on the damage data includes: Based on the radial position parameters corresponding to the location to be calculated, the damage data is mapped to the radial coordinate space of the target geological area to obtain damage distribution data; Based on the layering results, the damage data is labeled according to the fracture zone, plastic softening zone and elastic zone to obtain the region labeling data; Based on the damage distribution data and the corresponding regional labeling data, the damage feature map is generated, which is used to represent the degree of damage to the radial gradient of the surrounding rock of the roadway.

8. A device for calculating the radial gradient damage degree of roadway surrounding rock, characterized in that, include: The first acquisition module is used to acquire raw data of damage factors within the target geological area, and the raw data is used to determine the degree of damage within the target geological area. The first generation module is used to preprocess the original data to generate a training dataset for calculating the degree of damage. The first processing module is used to perform layering processing on the target geological region according to the boundary conditions corresponding to the training dataset, and obtain the layering result, wherein the layering result includes at least one of the fractured zone, the plastic softening zone and the elastic zone; The second processing module is used to input the corresponding raw data in the training dataset into the improved BP neural network model for processing based on the hierarchical results, so as to obtain the damage data in the target geological area. The second generation module is used to generate a damage feature map based on the damage data.

9. An electronic device, characterized in that, include: The memory, the processor, and the computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the steps in the method for calculating the radial gradient damage degree of the surrounding rock of the roadway as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps in the method for calculating the radial gradient damage degree of the surrounding rock of a roadway as described in any one of claims 1 to 7.

Images