Tunnel rock stratum large deformation analysis method and system based on three-dimensional modeling

Abstract

The embodiment of the present application relates to the technical field of data modeling analysis, and discloses a tunnel rock stratum large deformation analysis method and system based on three-dimensional modeling, which comprises the following steps: collecting initial three-dimensional point cloud data through three-dimensional laser scanning, constructing an initial three-dimensional geological model containing spatial form characteristics and rock stratum interface geometric topological relationship through space coordinate matching and surface reconstruction; performing finite element unstructured grid division on the model, and establishing a rock stratum mechanics analysis model in combination with physical mechanics parameters; inputting construction parameters and process information, sequentially activating unit nodes and solving force balance equations to simulate stress distribution; identifying potential deformation areas through iterative calculation, adjusting boundary conditions to predict deformation trend; and generating multi-view deformation early warning information by fusing three-dimensional space distribution cloud maps, time evolution curves and position labels, so that the accuracy and visualization degree of tunnel rock stratum deformation analysis are improved, and support is provided for construction safety.

Description

Technical Field

[0001] This invention relates to the field of data modeling and analysis technology, specifically to a method and system for analyzing large deformations of tunnel rock strata based on three-dimensional modeling. Background Technology

[0002] As a crucial component of transportation infrastructure, tunnel engineering's construction safety is closely linked to the stability of rock strata. When tunnels traverse complex geological conditions (such as weak or high-stress rock strata), large rock deformation is a major contributing factor to construction accidents. Currently, existing technologies for tunnel rock deformation analysis primarily focus on three aspects: geological modeling, mechanical analysis, and deformation monitoring. However, these technologies suffer from insufficient integration between geological and mechanical models, making it difficult to accurately reflect the mechanical response of complex rock strata; furthermore, deformation prediction lacks a dynamic iterative adjustment mechanism, failing to adapt to stress changes during construction. Summary of the Invention

[0003] This invention provides a method and system for analyzing large deformations of tunnel rock strata based on three-dimensional modeling.

[0004] In a first aspect, embodiments of the present invention provide a method for analyzing large deformations of tunnel rock strata based on three-dimensional modeling, applied to a tunnel rock strata large deformation analysis system, comprising:

[0005] Initial three-dimensional point cloud data of the tunnel rock strata were acquired by three-dimensional laser scanning. Based on the spatial coordinate matching and surface reconstruction processing of the initial three-dimensional point cloud data, an initial three-dimensional geological model was constructed. The initial three-dimensional geological model includes the spatial morphological features of the tunnel rock strata and the geometric topological relationship of the rock strata interface.

[0006] The initial three-dimensional geological model was subjected to unstructured mesh generation based on the finite element method, and a rock strata mechanical analysis model was established by combining the physical and mechanical parameters of the tunnel rock strata.

[0007] The construction parameters and process information during tunnel construction are input into the rock mechanics analysis model. The unit nodes of the corresponding areas are activated sequentially according to the process information. The stress distribution state of the tunnel rock strata during construction is simulated by solving the force balance equation of the unit nodes.

[0008] The rock strata mechanical analysis model is iteratively calculated. Potential deformation areas are determined by identifying shear stress concentration areas and stress change gradients. The constraint stiffness and constraint direction of the unit nodes in the potential deformation areas are adjusted to update the boundary conditions. The corrected construction parameters are reapplied to calculate the updated stress distribution state. Based on the updated stress distribution state, the deformation trend of the tunnel rock strata is predicted.

[0009] A three-dimensional spatial distribution cloud map and a time evolution curve are generated based on the deformation trend. The three-dimensional spatial location of the potential deformation region is marked based on the spatial expansion direction of the potential deformation region in the deformation trend. The three-dimensional spatial distribution cloud map, the time evolution curve and the three-dimensional spatial location marking of the potential deformation region are subjected to multi-dimensional feature fusion processing to generate deformation early warning information with multi-view display.

[0010] In one embodiment, the step of inputting construction parameters and process information during tunnel construction into a rock strata mechanics analysis model, sequentially activating unit nodes in corresponding regions according to the process information, and simulating the stress distribution state of the tunnel rock strata during construction by solving the force balance equations of the unit nodes includes:

[0011] The support parameters, excavation sequence, and construction loads extracted from the tunnel construction plan are classified, disassembled, and differentiated respectively.

[0012] The classified support parameters are converted into constraints of the rock mechanics analysis model; the disassembled excavation sequence is converted into the unit activation sequence of the model, and the unit nodes of the corresponding area are activated sequentially according to the excavation time interval of each area; the classified construction loads are converted into force application data of the model.

[0013] Activate the unit nodes of the tunnel excavation area in sequence according to the unit activation sequence, simulate the stress release of the rock strata during the excavation process, apply the corresponding uniformly distributed force or concentrated force to the activated unit nodes, and establish the force balance equation of the unit nodes based on the constitutive relation of the finite element element.

[0014] Solving the force equilibrium equations yields the normal and shear stress components at the element nodes, and the stress distribution inside the element is calculated by interpolation using the element shape function, generating stress data for each element.

[0015] The stress data of all units are globally assembled, the principal stress direction is calculated by combining the geometric topological relationship of the rock layer interface, the shear stress concentration area is determined by filtering the shear stress value threshold, and the stress change gradient is calculated by the spatial derivative of the stress value, generating a stress distribution state that includes the principal stress direction, the shear stress concentration area and the stress change gradient.

[0016] In one embodiment, the iterative calculation of the rock strata mechanical analysis model, identifying potential deformation regions by recognizing shear stress concentration areas and stress change gradients, adjusting the constraint stiffness and constraint direction of the unit nodes in the potential deformation regions to update the boundary conditions, reapplying the corrected construction parameters to calculate the updated stress distribution state, and predicting the deformation trend of the tunnel rock strata based on the updated stress distribution state includes:

[0017] The iteration termination condition is set as follows: the difference in stress distribution between two adjacent iterations is less than a preset range and the number of iterations does not exceed the upper limit.

[0018] In the first iteration, the initial stress distribution state is calculated based on the initial boundary conditions and construction parameters;

[0019] Shear stress concentration areas and stress change gradients are extracted from the initial stress distribution state. The boundaries of shear stress concentration areas are screened by the shear stress value exceeding the shear strength threshold of the rock strata. The stress transmission path is determined by the dominant direction vector analysis of the stress change gradient. The potential deformation area is determined by combining the boundaries of the shear stress concentration areas and the stress transmission path.

[0020] The initial constraint stiffness and constraint direction of the unit nodes in the potential deformation area are obtained. The constraint stiffness of the nodes with shear stress values ​​exceeding the rock shear strength threshold is increased, and the constraint stiffness of the nodes with shear stress values ​​below the rock shear strength threshold is decreased. The constraint direction is adjusted to be consistent with the dominant stress transmission direction to complete the boundary condition update.

[0021] The construction parameters are corrected based on the updated boundary conditions. The support reaction parameters are adjusted in areas where the constraint stiffness increases, and the earth pressure parameters are adjusted in areas where the constraint stiffness decreases. The corrected construction parameters are then applied to the element nodes.

[0022] The force equilibrium equations are re-solved to obtain the updated stress distribution state. The difference between the stress distribution and the initial stress distribution is calculated. If the difference does not meet the termination condition, the boundary conditions and parameter application steps are repeated until the termination condition is met.

[0023] Based on the stress distribution at termination, the strain distribution of the rock strata is calculated through the stress-strain constitutive relationship to generate the deformation trend; wherein, the displacement is obtained by integrating the strain, the derivative of the displacement with respect to time is the deformation rate, and the spatial gradient direction of the displacement is the direction of expansion of the potential deformation region.

[0024] In one embodiment, the step of extracting shear stress concentration regions and stress change gradients from the initial stress distribution state, screening the boundaries of shear stress concentration regions by shear stress values ​​exceeding the shear strength threshold of the rock strata, determining the stress transmission path through the dominant direction vector analysis of the stress change gradient, and determining potential deformation regions by combining the boundaries of the shear stress concentration regions and the stress transmission path includes:

[0025] Extract the shear stress values ​​of all elements from the initial stress distribution state, compare the shear stress values ​​with the shear strength of the rock layer, and screen out elements whose shear stress values ​​exceed the shear strength threshold of the rock layer;

[0026] The selected units are spatially clustered, and the boundaries of the shear stress concentration region are determined by the density clustering algorithm. The units within the boundary constitute the shear stress concentration region.

[0027] Calculate the stress change gradient of each element in the shear stress concentration region. The direction of the gradient vector is the direction of stress change. Statistically analyze the directional distribution of all gradient vectors. The direction with the highest frequency is the dominant direction of stress transmission.

[0028] Draw the stress transfer path along the dominant direction; the area covered by the path is the stress transfer influence area.

[0029] The overlapping area between the shear stress concentration region and the stress transfer influence region is defined as the potential deformation region, and the elements within the overlapping area are potential deformation elements.

[0030] In one embodiment, obtaining the initial constraint stiffness and constraint direction of the element nodes within the potential deformation region, increasing the constraint stiffness of nodes whose shear stress value exceeds the rock stratum shear strength threshold, decreasing the constraint stiffness of nodes whose shear stress value is lower than the rock stratum shear strength threshold, and adjusting the constraint direction to be consistent with the dominant stress transmission direction to complete the boundary condition update includes:

[0031] Traverse all element nodes within the potential deformation region and extract the node coordinates and corresponding shear stress values ​​for each element node;

[0032] Query the initial constraint stiffness and constraint direction of the unit node; where the initial constraint stiffness is set based on the original physical and mechanical parameters of the rock strata, and the initial constraint direction is set based on the tunnel axis direction;

[0033] For nodes where the shear stress exceeds the rock shear strength threshold, the constraint stiffness is increased to a set multiple of the initial value to simulate the constraint effect of reinforced support; for nodes where the shear stress is lower than the rock shear strength threshold, the constraint stiffness is reduced to a set multiple of the initial value to simulate the relaxation effect of stress release.

[0034] Calculate the dominant direction vector of stress transfer, and adjust the constraint direction of the element node to be consistent with the dominant direction vector so that the constraint direction matches the stress transfer direction;

[0035] The adjusted constraint stiffness and constraint direction are reassigned to the element nodes, and the boundary conditions of the potential deformation region are updated.

[0036] In one embodiment, the process of generating a three-dimensional spatial distribution cloud map and a time evolution curve based on the deformation trend, and marking the three-dimensional spatial location of potential deformation regions based on the spatial expansion direction of potential deformation regions in the deformation trend, performs multi-dimensional feature fusion processing on the three-dimensional spatial distribution cloud map, the time evolution curve, and the three-dimensional spatial location markings of potential deformation regions to generate deformation early warning information containing multi-view displays, including:

[0037] Extract the displacement, deformation rate, and spatial expansion direction information of the potential deformation area from the deformation trend, obtain the displacement and corresponding spatial coordinates of each unit node, obtain the deformation rate value of each time step, and the time step corresponds to the stage interval of the construction process.

[0038] Spatial feature processing is performed on the displacement data. The displacement features of the neighboring units are associated with the unit node as the center. The enhanced displacement value of the unit node is generated by aggregating the features of the neighboring units. Based on the spatial distribution features of the enhanced displacement value, a color mapping rule is set to map the enhanced displacement value into a color and render it onto the initial three-dimensional geological model to generate a three-dimensional spatial distribution cloud map of displacement containing neighborhood association features.

[0039] The deformation rate data is processed using time-series features. The deformation rate at the current time step is used as the core to associate the deformation rate features of multiple previous time steps. The enhanced deformation rate value at the current time step is generated through feature transfer within the time-series window. The trend features of the enhanced deformation rate value are fitted to generate a deformation rate time evolution curve containing time-series associated features. The curve contains the instantaneous value at the current time step and the trend change rate within the time-series window.

[0040] Based on the spatial expansion direction and spatial distribution characteristics of the potential deformation area, the displacement change trend of the unit nodes within the potential deformation area is associated, the three-dimensional spatial boundary of the potential deformation area is marked, and the boundary is marked on the initial three-dimensional geological model using a preset style.

[0041] A unified scene coordinate system is created, and the three-dimensional spatial distribution cloud map of displacement, the time evolution curve of deformation rate, and the three-dimensional spatial boundary annotation of potential deformation area are imported into the unified scene coordinate system and the spatiotemporal features of the unified scene coordinate system are correlated to construct a multi-view display framework covering different observation dimensions of tunnel rock strata deformation.

[0042] Within a multi-perspective display framework, the overall spatial distribution, temporal variation, and regional expansion characteristics of tunnel rock strata deformation are presented through the interconnected display of spatial distribution features, temporal variation features, and regional expansion features. This generates visualized deformation early warning information with multiple perspectives.

[0043] In one embodiment, the spatial feature processing of the displacement data, which associates the displacement features of neighboring units with the unit node as the center, and generates an enhanced displacement value for the unit node through feature aggregation of neighboring units, includes:

[0044] Centered on each unit node, a neighborhood range is set to cover a preset number of unit nodes around it, generating a set of neighboring units for each unit node. The size of the neighborhood range is determined based on the structural characteristics of the tunnel rock strata.

[0045] Extract the displacement value and corresponding spatial coordinates of each node in the neighborhood unit set, and calculate the spatial distance between each neighborhood unit and the central unit node based on the Euclidean distance in the three-dimensional coordinate system.

[0046] The weight coefficient of the neighborhood unit is set based on the spatial distance. The smaller the spatial distance, the larger the weight coefficient, and the larger the spatial distance, the smaller the weight coefficient. The value range of the weight coefficient is within a preset range.

[0047] Multiply the displacement value of each neighboring cell by the corresponding weight coefficient to obtain the weighted displacement value of the neighboring cell.

[0048] The weighted average of the weighted displacement values ​​of all neighboring units is aggregated to generate the enhanced displacement value of the central unit node; the enhanced displacement value integrates the displacement characteristics of the central unit node itself and the displacement characteristics of the neighboring units.

[0049] Traverse all unit nodes in the initial 3D geological model, repeat the steps of setting a neighborhood range covering a preset number of unit nodes around each unit node as the center, generating a set of neighboring units of each unit node, performing a weighted average aggregation process on the weighted displacement values ​​of all neighboring units, generating the enhanced displacement value of the central unit node, generating the enhanced displacement value of each unit node, and generating an enhanced displacement dataset containing spatial neighborhood association features.

[0050] Based on the spatial distribution characteristics of the enhanced displacement dataset, the differences in enhanced displacement values ​​of unit nodes in different regions are associated according to the lightness and darkness strategy and the gradient trend strategy of color mapping.

[0051] In one embodiment, the step of performing time-series feature processing on the deformation rate data, using the deformation rate of the current time step as the core and associating it with the deformation rate features of multiple previous time steps, and generating an enhanced deformation rate value for the current time step through feature transfer within a time-series window, includes:

[0052] Set the length of the timing window. The timing window covers the current time step and a preset number of previous time steps. The length of the timing window is determined based on the stage interval of the construction process.

[0053] Extract the deformation rate value for each time step within the time window, where the order of time steps corresponds to the sequence of construction processes;

[0054] Based on the order of time steps, a time series weight coefficient is set for each time step. The weight coefficient of the current time step is the largest, and the weight coefficients of the preceding time steps decrease sequentially over time. The sum of the time series weight coefficients is a preset value.

[0055] The weighted deformation rate value for each time step is obtained by combining the deformation rate value for each time step with the corresponding temporal weighting coefficient.

[0056] The weighted deformation rate values ​​of all time steps within the time window are aggregated by weighted summation to generate the enhanced deformation rate value of the current time step. The enhanced deformation rate value integrates the deformation rate characteristics of the current time step and the deformation rate characteristics of the previous time step.

[0057] Iterate through all time steps, repeat the step of setting the length of the time window to aggregate the weighted deformation rate values ​​of all time steps within the time window, generate the enhanced deformation rate value for each time step, and generate an enhanced deformation rate dataset containing time-series correlation features.

[0058] Based on the enhanced deformation rate dataset, the changes in enhanced deformation rate values ​​at the current time step and the previous time step are correlated. The trend characteristics of the enhanced deformation rate values ​​are fitted using a linear regression method to generate the trend change rate of the deformation rate time evolution curve. The trend change rate reflects the direction and speed of change of the enhanced deformation rate values.

[0059] In one embodiment, the creation of a unified scene coordinate system involves importing the three-dimensional spatial distribution cloud map of displacement, the time evolution curve of deformation rate, and the three-dimensional spatial boundary annotation of potential deformation areas into the unified scene coordinate system and performing spatiotemporal feature association of the unified scene coordinate system to construct a multi-view display framework covering different observation dimensions of tunnel rock strata deformation, including:

[0060] A unified scene coordinate system is established with the center point of the tunnel entrance, the tunnel extension direction, the horizontal direction of the vertical extension, and the direction perpendicular to the ground. The coordinate system unit is consistent with the initial three-dimensional geological model.

[0061] The enhanced displacement values ​​of the three-dimensional spatial distribution cloud map of displacement are mapped to the corresponding positions in the unified scene coordinate system according to the coordinates of the unit nodes. The enhanced displacement value of each unit node is associated with the displacement features of its neighboring units. The enhanced displacement values ​​are rendered as corresponding colors and superimposed on the initial three-dimensional geological model through the light and dark size strategy and the gradient trend strategy of the color mapping, so as to realize the spatial distribution features of displacement and the spatial position of the three-dimensional geological model.

[0062] The time axis of the deformation rate time evolution curve is associated with the tunnel construction progress in a unified scene coordinate system. The construction progress corresponds to the spatial extension position of the tunnel. The Y-axis of the deformation rate time evolution curve corresponds to the enhanced deformation rate value. Each time step node of the deformation rate time evolution curve is associated with the corresponding spatial position of the tunnel under the construction progress. The instantaneous enhanced deformation rate value of the current time step and the trend change rate within the time series window are marked to realize the spatiotemporal association between the temporal characteristics of deformation rate and the spatial position of tunnel construction.

[0063] The coordinates of the three-dimensional spatial boundary of the potential deformation region are converted into coordinates under a unified scene coordinate system. Based on the spatial expansion direction and spatial gradient characteristics of the potential deformation region in the deformation trend, the boundary of the potential deformation region is defined on the initial three-dimensional geological model. The boundary nodes are associated with the displacement change trend and time-series deformation rate characteristics of the corresponding units, so as to realize the multi-dimensional association between the spatial boundary of the potential deformation region and the displacement and rate characteristics.

[0064] By combining the spatial location association, the spatiotemporal association, and the multidimensional association, the spatial distribution of displacement, the temporal trend of deformation rate, and the spatial expansion characteristics of potential deformation area are integrated into a unified scene coordinate system to generate a multi-view display framework covering different observation dimensions of tunnel rock strata deformation spatial distribution, rate temporal sequence, and regional expansion.

[0065] Secondly, embodiments of the present invention provide a tunnel rock strata large deformation analysis system, comprising:

[0066] processor;

[0067] Storage device, on which computer programs are stored,

[0068] When the computer program is executed by the processor, the processor implements any of the described methods for analyzing large deformations of tunnel rock strata based on three-dimensional modeling.

[0069] This invention provides a readable storage medium storing a program or instructions, which, when executed by a processor, implement the steps of the method for analyzing large deformations of tunnel rock strata based on three-dimensional modeling.

[0070] Therefore, the embodiments of the present invention have the following beneficial effects: Initial three-dimensional point cloud data is acquired through three-dimensional laser scanning, and an initial three-dimensional geological model containing spatial morphological features and the geometric topological relationship of rock strata interfaces is constructed. A rock strata mechanical analysis model is established by combining finite element unstructured mesh generation and physical and mechanical parameters. Then, stress distribution is simulated by dynamically inputting construction parameters and process information and activating unit nodes. Furthermore, potential deformation areas are identified through iterative calculations, and boundary conditions are adjusted to predict deformation trends. Finally, multi-dimensional features are integrated to generate multi-view early warning information. From an overall logical perspective, a closed-loop process from three-dimensional geological modeling to mechanical analysis, deformation prediction, and then to visual early warning is achieved, breaking through the technical bottlenecks of traditional tunnel rock strata deformation analysis, such as the disconnect between geological and mechanical models, the lack of dynamic iterative adjustment in deformation prediction, and the single dimension of early warning information.

[0071] In detail, the high-precision spatial data from 3D laser scanning is deeply integrated with the mechanical modeling of finite element analysis. The actual stress evolution during construction is restored through a process-driven unit node activation mechanism. Then, potential deformation areas are accurately located through the collaborative identification of shear stress concentration areas and stress change gradients. Combined with the dynamic adjustment of boundary conditions, high-precision prediction of deformation trends is achieved. Finally, intuitive multi-view early warning information is generated through multi-dimensional feature fusion. This provides an integrated technical solution for tunnel construction safety, from data acquisition to decision support, improving the accuracy, dynamism, and visualization of large deformation analysis of tunnel rock strata, and effectively reducing construction risks. Attached Figure Description

[0072] Figure 1 This is a flowchart of a method for analyzing large deformations of tunnel rock strata based on three-dimensional modeling, provided in an embodiment of the present invention.

[0073] Figure 2 This is a block diagram of a tunnel rock strata large deformation analysis device based on three-dimensional modeling, provided in an embodiment of the present invention.

[0074] Figure 3 This is a schematic diagram of the basic structure of a tunnel rock strata large deformation analysis system provided in an embodiment of the present invention. Detailed Implementation

[0075] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0076] See Figure 1 As shown in the figure, this is a flowchart of a method for analyzing large deformations of tunnel rock strata based on three-dimensional modeling, provided by an embodiment of the present invention. This method can be applied to a tunnel rock strata large deformation analysis system. Figure 1 As shown, the method may include steps 110-150.

[0077] This invention provides a method for analyzing large deformation of tunnel rock strata based on three-dimensional modeling. The execution subject is a tunnel rock strata large deformation analysis system (computer equipment or server). The application scenario is the analysis of large deformation of rock strata in a mountain tunnel. The rock strata that the tunnel passes through are mainly interbedded sandy mudstone and shale. The rock strata dip angle is within a preset angle range. The tunnel is designed as a single-tube two-way traffic structure with a preset design length and a horseshoe-shaped excavation cross section.

[0078] Step 110: Acquire initial three-dimensional point cloud data of the tunnel rock strata through three-dimensional laser scanning, and construct an initial three-dimensional geological model based on the spatial coordinate matching and surface reconstruction processing of the initial three-dimensional point cloud data. The initial three-dimensional geological model includes the spatial morphological features of the tunnel rock strata and the geometric topological relationship of the rock strata interface.

[0079] In this embodiment of the invention, the tunnel rock strata large deformation analysis system controls a three-dimensional laser scanner to be set up at a preset position outside the tunnel entrance. The scanning resolution is set to a preset parameter, the scanning range covers a preset distance range from the tunnel entrance to the unexcavated area, and the scanning frequency is a preset frequency. Initial three-dimensional point cloud data is collected. The initial three-dimensional point cloud data includes surface point cloud data of sandy mudstone and shale. Each point cloud data includes X, Y, Z three-dimensional spatial coordinates and reflection intensity values.

[0080] Next, the system performs spatial coordinate matching processing on the initial 3D point cloud data. The ICP (Iterative Closest Point) algorithm is used to transform the point cloud data scanned from multiple stations to the same coordinate system. The coordinate system takes the center point of the tunnel entrance as the origin, the tunnel extension direction as the X-axis, the vertical extension horizontal direction as the Y-axis, and the vertical ground direction as the Z-axis. The matching accuracy is controlled within a preset range.

[0081] Then, the system performs surface reconstruction processing, using the Poisson surface reconstruction algorithm, setting the sampling density to a preset value, and triangulating the matched point cloud data to generate an initial three-dimensional geological model containing the interface between sandy mudstone and shale. In the model, the interface between sandy mudstone and shale is marked by different colors, and the geometric topological relationship of the interface is represented by the adjacency relationship of triangular facets. The spatial morphological features of the model include the strike, dip angle, and thickness distribution of the rock layers. For example, the thickness of sandy mudstone is within a preset range, and the thickness of shale is within a preset range.

[0082] Step 120: Perform unstructured mesh generation on the initial three-dimensional geological model based on finite element method, and establish a rock strata mechanical analysis model in combination with the physical and mechanical parameters of the tunnel rock strata.

[0083] In this embodiment of the invention, the tunnel rock strata large deformation analysis system calls the finite element analysis module to perform unstructured meshing processing on the initial three-dimensional geological model. Tetrahedral elements are used for meshing, and the mesh size is set to a preset value. The mesh is densified in the interface area between sandy mudstone and shale, with a densification coefficient set to a preset value. The number of mesh elements and nodes obtained are preset.

[0084] Next, the system acquires the physical and mechanical parameters of the tunnel rock strata. These parameters are obtained through indoor rock mechanics tests. The elastic modulus, Poisson's ratio, compressive strength, and shear strength of the sandy mudstone are preset values. The elastic modulus, Poisson's ratio, compressive strength, and shear strength of the shale are also preset values.

[0085] Then, the system establishes a rock strata mechanics analysis model by combining physical and mechanical parameters. The unit attributes of the model are assigned according to the rock strata type. Sandy mudstone units are assigned the physical and mechanical parameters of sandy mudstone, and shale units are assigned the physical and mechanical parameters of shale. The boundary conditions of the model are set as follows: the direction of the tunnel entrance is a free boundary, the direction of the unexcavated area is a fixed boundary, and the direction perpendicular to the ground is a constrained boundary.

[0086] Step 130: Input the construction parameters and process information of the tunnel construction process into the rock mechanics analysis model, activate the unit nodes of the corresponding areas in sequence according to the process information, and simulate the stress distribution state of the tunnel rock strata during the construction process by solving the force balance equation of the unit nodes.

[0087] In this embodiment of the invention, the tunnel construction parameters include support parameters, excavation sequence, and construction loads. The support parameters are a combination of steel arch support and shotcrete support, with the steel arch model being a preset model, the spacing being a preset distance, the shotcrete thickness being a preset thickness, and the strength grade being a preset grade. The excavation sequence is bench excavation, divided into upper bench, middle bench, and lower bench, with the upper bench excavation length, middle bench excavation length, and lower bench excavation length all being preset lengths. The construction loads are the earth pressure at the tunnel excavation face and the self-weight of the support structure, with the earth pressure and the self-weight of the support structure being preset values. The process information is: upper bench excavation → upper bench support → middle bench excavation → middle bench support → lower bench excavation → lower bench support.

[0088] Step 131: Classify, disassemble, and differentiate the support parameters, excavation sequence, and construction loads extracted from the tunnel construction plan.

[0089] In this embodiment of the invention, the tunnel rock strata large deformation analysis system extracts support parameters, excavation sequence, and construction loads from the tunnel construction plan. The support parameters are categorized into steel arch parameters and shotcrete parameters. Steel arch parameters include model, spacing, and cross-sectional dimensions, while shotcrete parameters include thickness, strength grade, and mix proportion. The excavation sequence is broken down into upper bench excavation, upper bench support, middle bench excavation, middle bench support, lower bench excavation, and lower bench support. The construction loads are differentiated into earth pressure loads and self-weight loads. Earth pressure loads include active earth pressure and passive earth pressure, while self-weight loads include the self-weight of the steel arch and the self-weight of the shotcrete.

[0090] Step 132: Convert the classified support parameters into constraints of the rock mechanics analysis model; convert the disassembled excavation sequence into the unit activation sequence of the model, and activate the unit nodes of the corresponding area in sequence according to the excavation time interval of each area; convert the classified construction loads into force application data of the model.

[0091] In this embodiment of the invention, the system converts the steel arch frame parameters into elastic constraint conditions of the model, with the constraint stiffness set to a preset value and the constraint direction being the tunnel radial direction; and converts the shotcrete parameters into surface constraint conditions of the model, with the constraint strength set to a preset value.

[0092] The disassembled excavation sequence is converted into a unit activation sequence. The upper bench excavation process corresponds to the unit node of the upper bench area in the activation model, the upper bench support process corresponds to the unit node of the upper bench support area in the activation model, the middle bench excavation process corresponds to the unit node of the middle bench area in the activation model, the middle bench support process corresponds to the unit node of the middle bench support area in the activation model, the lower bench excavation process corresponds to the unit node of the lower bench area in the activation model, and the lower bench support process corresponds to the unit node of the lower bench support area in the activation model. The time interval between excavations in each area is a preset number of days.

[0093] The differentiated construction loads are converted into force application data for the model. Earth pressure loads are converted into uniformly distributed forces applied to the excavation face unit nodes of the model, with active earth pressure and passive earth pressure set to preset values. Self-weight loads are converted into concentrated forces applied to the support structure unit nodes of the model, with the self-weight of the steel arch frame and the self-weight of the shotcrete set to preset values.

[0094] Step 133: Activate the unit nodes of the tunnel excavation area in sequence according to the unit activation sequence, simulate the stress release of the rock strata during the excavation process, apply the corresponding uniformly distributed force or concentrated force to the activated unit nodes, and establish the force balance equation of the unit nodes based on the constitutive relationship of the finite element element.

[0095] In this embodiment of the invention, the system first activates the unit nodes of the upper step excavation area according to the unit activation sequence. The activation method is to remove the constraint conditions of the unit nodes in the area and simulate the stress release of the rock strata during the excavation process. The release coefficient is a preset value.

[0096] Next, active earth pressure uniformly distributed force is applied to the activated upper step excavation area unit nodes, and concentrated force of steel arch frame self-weight and concentrated force of shotcrete self-weight are applied to the activated upper step support area unit nodes.

[0097] Then, the system establishes the force balance equations of the element nodes based on the constitutive relation of the finite element. The constitutive relation adopts the elastoplastic constitutive model, the yield criterion is the Mohr-Coulomb criterion, and the force balance equations are in the form of: the nodal force vector is equal to the element stiffness matrix multiplied by the nodal displacement vector. The nodal force vector includes the earth pressure load, the self-weight load and the support constraint reaction force. The element stiffness matrix is ​​calculated based on the physical and mechanical parameters of the element and the element shape. The nodal displacement vector is the unknown quantity to be solved.

[0098] Step 134: Solve the force balance equation to obtain the normal stress and shear stress components of the element nodes, and calculate the stress distribution inside the element by interpolation of the element shape function to generate stress data for each element.

[0099] In this embodiment of the invention, the system uses the Newton-Raphson iterative method to solve the force balance equation. The convergence condition for the iteration is that the difference in nodal displacement between two adjacent iterations is less than a preset value, and the number of iterations does not exceed a preset number. The normal stress components of the element nodes obtained by the solution include normal stress in the X direction, normal stress in the Y direction, and normal stress in the Z direction, and the shear stress components include shear stress in the XY direction, shear stress in the YZ direction, and shear stress in the XZ direction.

[0100] Next, the system calculates the stress distribution inside the element by interpolating the element shape function. The element shape function is a linear shape function of the tetrahedral element. The interpolation method is to substitute the stress components of the element nodes into the shape function to calculate the stress components at any point inside the element, and generate stress data for each element. The stress data includes the normal stress components and shear stress components at each point inside the element.

[0101] Step 135: Globally assemble the stress data of all units, calculate the principal stress direction by combining the geometric topological relationship of the rock layer interface, determine the shear stress concentration area by filtering the shear stress value threshold, and calculate the stress change gradient by the spatial derivative of the stress value to generate a stress distribution state that includes the principal stress direction, the shear stress concentration area and the stress change gradient.

[0102] In this embodiment of the invention, the system globally assembles the stress data of each unit according to the node number of the unit to generate a global stress dataset.

[0103] Next, combining the geometric and topological relationships of the rock strata interface, the principal stress directions are calculated through the eigenvalue decomposition of the stress tensor. The eigenvalue decomposition method is as follows: the stress tensor matrix is ​​converted into a diagonal matrix, the diagonal elements are the principal stress values, and the corresponding eigenvectors are the principal stress directions. The principal stress directions include the maximum principal stress direction, the intermediate principal stress direction, and the minimum principal stress direction.

[0104] Then, the system determines the shear stress concentration area by screening the shear stress value threshold. The shear stress value threshold is set as a preset proportion of the shear strength of sandy mudstone. Units with shear stress values ​​exceeding the threshold are screened out, and spatial clustering is performed on these units using the DBSCAN density clustering algorithm. The clustering radius is a preset value, and the minimum number of points is a preset number of points, thus obtaining the shear stress concentration area.

[0105] Meanwhile, the system calculates the stress change gradient by using the spatial derivative of the stress value. The spatial derivative is calculated as follows: a preset number of sampling points are taken inside the unit, and the ratio of the stress value difference between adjacent sampling points to the spatial distance is calculated to obtain the stress change gradient. The gradient direction is the direction of the fastest stress value change, and the gradient magnitude is the rate of stress value change.

[0106] Finally, the system generates a stress distribution state that includes the principal stress direction, shear stress concentration area, and stress change gradient. The principal stress direction is marked on the initial three-dimensional geological model by arrows, the shear stress concentration area is marked by different colors, and the stress change gradient is marked by contour lines.

[0107] Step 140: Iteratively calculate the rock strata mechanics analysis model, identify potential deformation areas by identifying shear stress concentration areas and stress change gradients, adjust the constraint stiffness and constraint direction of the unit nodes in the potential deformation areas to update the boundary conditions, reapply the corrected construction parameters to calculate the updated stress distribution state, and predict the deformation trend of the tunnel rock strata based on the updated stress distribution state.

[0108] In this embodiment of the invention, the system first sets the iteration termination condition as follows: the stress distribution difference between two adjacent iterations is less than a preset range and the number of iterations does not exceed the upper limit. The stress distribution difference between two adjacent iterations is calculated by the root mean square error. The preset range is a preset value, and the upper limit of the number of iterations is a preset number.

[0109] Step 141: In the first iteration, calculate the initial stress distribution state based on the initial boundary conditions and construction parameters.

[0110] In this embodiment of the invention, the system inputs the initial boundary conditions (the direction of the tunnel entrance is the free boundary, the direction of the unexcavated area is the fixed boundary, and the direction perpendicular to the ground is the constrained boundary) and the initial construction parameters (support parameters, excavation sequence, and construction load) into the rock mechanics analysis model, and calculates the initial stress distribution state according to the method of steps 130 to 135. The initial stress distribution state includes the initial shear stress concentration area, the initial principal stress direction, and the initial stress change gradient.

[0111] Step 142: Extract the shear stress concentration region and stress change gradient from the initial stress distribution state. Screen the boundary of the shear stress concentration region by the shear stress value exceeding the shear strength threshold of the rock layer. Determine the stress transmission path by analyzing the dominant direction vector of the stress change gradient. Combine the boundary of the shear stress concentration region and the stress transmission path to determine the potential deformation region.

[0112] In this embodiment of the invention, the system first extracts the shear stress values ​​of all units from the initial stress distribution state, compares the shear stress values ​​with the shear strength threshold (preset value) of sandy mudstone, and filters out units whose shear stress values ​​exceed the threshold.

[0113] Next, spatial clustering is performed on the selected units using the DBSCAN density clustering algorithm. The cluster radius is a preset value, and the minimum number of points is a preset number of points. The boundary of the shear stress concentration region is determined. The units within the boundary constitute the shear stress concentration region. For example, the shear stress concentration region is located at the left arch waist of the upper step of the tunnel. The region shape is an irregular polygon and contains a preset number of units.

[0114] Then, the system calculates the stress change gradient of each element in the shear stress concentration region. The direction of the gradient vector is the direction of stress change. The distribution of the directions of all gradient vectors is statistically analyzed. The direction with the highest frequency is the dominant direction of stress transmission. For example, the dominant direction is the radial inward direction of the tunnel.

[0115] Next, the system draws the stress transmission path along the dominant direction. The area covered by the path is the stress transmission influence area. The stress transmission path is marked by lines drawn on the initial three-dimensional geological model. The thickness of the lines is proportional to the magnitude of the stress change gradient.

[0116] Finally, the system identifies the overlapping area between the shear stress concentration area and the stress transfer influence area as the potential deformation area. The elements within the overlapping area are potential deformation elements. For example, the potential deformation area is located in the area from the left arch waist to the arch top of the tunnel upper step, containing a preset number of potential deformation elements.

[0117] Step 143: Obtain the initial constraint stiffness and constraint direction of the unit nodes in the potential deformation area. Increase the constraint stiffness of nodes whose shear stress value exceeds the rock layer shear strength threshold, and decrease the constraint stiffness of nodes whose shear stress value is lower than the rock layer shear strength threshold. Adjust the constraint direction to be consistent with the dominant stress transmission direction to complete the boundary condition update.

[0118] In this embodiment of the invention, the system traverses all element nodes within the potential deformation region and extracts the node coordinates and corresponding shear stress values ​​of each element node.

[0119] Next, the system queries the initial constraint stiffness and constraint direction of the unit node. The initial constraint stiffness is set based on the original physical and mechanical parameters of the sandy mudstone. For example, the initial constraint stiffness is a preset value. The initial constraint direction is set based on the tunnel axis direction. For example, the initial constraint direction is radially outward from the tunnel.

[0120] Then, the system increases the constraint stiffness of nodes whose shear stress value exceeds the shear strength threshold of sandy mudstone by a preset multiple of the initial value, such as twice the initial value, to simulate the constraint effect of reinforced support; and reduces the constraint stiffness of nodes whose shear stress value is lower than the shear strength threshold of sandy mudstone by a preset proportion of the initial value, such as 0.8 times the initial value, to simulate the relaxation effect of stress release.

[0121] At the same time, the system adjusts the constraint direction to be consistent with the dominant stress transmission direction. For example, if the dominant stress transmission direction is radially inward from the tunnel, the constraint direction is adjusted to radially inward from the tunnel to match the constraint direction with the stress transmission direction.

[0122] Finally, the system reassigns the adjusted constraint stiffness and constraint direction to the element nodes, updating the boundary conditions of the potential deformation region.

[0123] Step 144: Correct the construction parameters according to the updated boundary conditions. Adjust the support reaction parameters in areas where the constraint stiffness increases, and adjust the earth pressure parameters in areas where the constraint stiffness decreases. Apply the corrected construction parameters to the element nodes.

[0124] In this embodiment of the invention, the system adjusts the support reaction parameters corresponding to the regions where the constraint stiffness increases (node ​​regions where the shear stress value exceeds the threshold). The support reaction parameters are directly proportional to the constraint stiffness. For example, if the constraint stiffness increases to twice the initial value, then the support reaction parameters are increased to twice the initial value. For the regions where the constraint stiffness decreases (node ​​regions where the shear stress value is below the threshold), the system adjusts the earth pressure parameters corresponding to the regions where the constraint stiffness decreases. The earth pressure parameters are inversely proportional to the constraint stiffness. For example, if the constraint stiffness decreases to 0.8 times the initial value, then the earth pressure parameters are decreased to 0.8 times the initial value.

[0125] Next, the system applies the corrected construction parameters (adjusted support reaction parameters and earth pressure parameters) to the corresponding unit nodes. The application method is to convert the support reaction parameters into concentrated forces and apply them to the unit nodes, and to convert the earth pressure parameters into uniformly distributed forces and apply them to the unit nodes.

[0126] Step 145: Resolve the force equilibrium equation to obtain the updated stress distribution state, calculate the difference between the stress distribution and the initial stress distribution. If the difference does not meet the termination condition, repeat the steps of adjusting the boundary conditions and applying parameters until the termination condition is met.

[0127] In this embodiment of the invention, the system re-solves the force balance equation according to the method of steps 133 to 135 to obtain the updated stress distribution state. The updated stress distribution state includes the updated shear stress concentration region, the updated principal stress direction, and the updated stress change gradient.

[0128] Next, the system calculates the difference between the updated stress distribution state and the initial stress distribution state. The difference is obtained by calculating the root mean square error. The root mean square error is calculated by: calculating the sum of the squares of the differences of the stress components corresponding to all elements, dividing by the number of elements, and then taking the square root.

[0129] If the difference does not meet the termination condition (the difference is less than a preset value and the number of iterations does not exceed a preset number), the system repeats steps 142 to 144 to adjust the boundary conditions and apply parameters until the termination condition is met. For example, after the first iteration, if the difference is the preset value and greater than the preset value of the termination condition, the system repeats the steps to adjust the boundary conditions and apply parameters for the second iteration. After the second iteration, if the difference is the preset value and less than the preset value of the termination condition, the termination condition is met, and the iteration stops.

[0130] Step 146: Based on the stress distribution state at the time of termination, calculate the strain distribution of the rock strata through the stress-strain constitutive relationship to generate the deformation trend; wherein, the displacement is obtained by integrating the strain, the derivative of the displacement with respect to time is the deformation rate, and the spatial gradient direction of the displacement is the direction of expansion of the potential deformation region.

[0131] In this embodiment of the invention, the system calculates the strain distribution of the rock strata based on the stress distribution state at termination using the stress-strain constitutive relationship. The stress-strain constitutive relationship adopts an elastoplastic constitutive model, and the strain components include normal strain components and shear strain components. The normal strain components are calculated by dividing the normal stress components by the elastic modulus, and the shear strain components are calculated by dividing the shear stress components by the shear modulus. The shear modulus is calculated based on the elastic modulus and Poisson's ratio.

[0132] Next, the system integrates the strain to obtain the displacement. The integration method is to integrate the strain components in space. The integration path is from the element node to any point inside the element to obtain the displacement of any point inside the element. The displacement includes the displacement in the X direction, the displacement in the Y direction, and the displacement in the Z direction.

[0133] Then, the system calculates the deformation rate by taking the derivative of the displacement with respect to time. The deformation rate is calculated by calculating the difference in displacement between two adjacent time steps and then dividing it by the time step interval, which is a preset number of days. The deformation rate includes the deformation rate in the X direction, the deformation rate in the Y direction, and the deformation rate in the Z direction.

[0134] Simultaneously, the system calculates the spatial gradient direction of the displacement to obtain the potential deformation region expansion direction. The spatial gradient direction is calculated by calculating the rate of change of the displacement in space. The direction with the largest rate of change is the potential deformation region expansion direction. For example, the potential deformation region expansion direction is the radial inward direction of the tunnel.

[0135] Finally, the system generates deformation trends, which include displacement distribution, deformation rate distribution, and the direction of potential deformation region expansion.

[0136] Step 150: Generate a three-dimensional spatial distribution cloud map and a time evolution curve based on the deformation trend, and mark the three-dimensional spatial position of the potential deformation region based on the spatial expansion direction of the potential deformation region in the deformation trend. Perform multi-dimensional feature fusion processing on the three-dimensional spatial distribution cloud map, the time evolution curve and the three-dimensional spatial position marking of the potential deformation region to generate deformation early warning information with multi-view display.

[0137] In this embodiment of the invention, the system first extracts the displacement, deformation rate and spatial expansion direction information of the potential deformation area from the deformation trend, obtains the displacement and corresponding spatial coordinates of each unit node, and obtains the deformation rate value of each time step. The time step corresponds to the stage interval of the construction process. For example, the time step is a preset number of days, corresponding to the upper step excavation stage, the upper step support stage, etc.

[0138] Step 151: Perform spatial feature processing on the displacement data, associate the displacement features of the neighboring units with the unit node as the center, generate the enhanced displacement value of the unit node by aggregating the features of the neighboring units, set the color mapping rules based on the spatial distribution features of the enhanced displacement value, map the enhanced displacement value into color and render it onto the initial three-dimensional geological model, and generate a three-dimensional spatial distribution cloud map of displacement containing the neighbor association features.

[0139] In this embodiment of the invention, the system takes each unit node as the center and sets a neighborhood range that covers a preset number of unit nodes around it, for example, a preset number of 6, to generate a set of neighboring units of the unit node. The size of the neighborhood range is determined based on the structural characteristics of the tunnel rock strata, for example, the neighborhood range is a preset distance.

[0140] Next, the system extracts the displacement value and corresponding spatial coordinates of each unit node in the neighborhood unit set. Based on the Euclidean distance in the three-dimensional coordinate system, it calculates the spatial distance between each neighborhood unit and the central unit node. The Euclidean distance is calculated by summing the squares of the coordinate differences between the neighborhood unit and the central unit node in the X, Y, and Z directions, and then taking the square root.

[0141] Then, the system sets the weight coefficient of the neighborhood unit based on the spatial distance. The smaller the spatial distance, the larger the weight coefficient, and the larger the spatial distance, the smaller the weight coefficient. The value range of the weight coefficient is within a preset range, for example, the preset range is 0.1 to 0.9. The weight coefficient is calculated as follows: the weight coefficient is equal to 1 divided by P, where P is equal to spatial distance * preset coefficient + 1.

[0142] Next, the system multiplies the displacement value of each neighboring unit by its corresponding weight coefficient to obtain the weighted displacement value of the neighboring unit. Then, the system performs a weighted average aggregation process on the weighted displacement values ​​of all neighboring units to generate the enhanced displacement value of the central unit node. The weighted average is calculated by adding the weighted displacement values ​​of all neighboring units and then dividing by the number of neighboring units. For example, the displacement value of the central unit node is a preset value, the displacement values ​​of the neighboring units are preset values, the corresponding weight coefficients are preset values, the weighted displacement values ​​are preset values, and the enhanced displacement value is a preset value.

[0143] Next, the system traverses all unit nodes in the initial 3D geological model, repeating the steps of setting a neighborhood range covering a preset number of unit nodes around each unit node as the center, generating a set of neighboring units for each unit node, performing a weighted average aggregation process on the weighted displacement values ​​of all neighboring units, generating the enhanced displacement value of the central unit node, generating the enhanced displacement value of each unit node, and generating an enhanced displacement dataset containing spatial neighborhood association features.

[0144] Then, the system sets color mapping rules based on the spatial distribution characteristics of the enhanced displacement dataset. The color mapping rules are: the smaller the enhanced displacement value, the lighter the color; the larger the enhanced displacement value, the darker the color, with the color range from blue (light) to red (dark).

[0145] Finally, the system maps the enhanced displacement values ​​to colors and renders them onto the initial 3D geological model, generating a 3D spatial distribution cloud map of the displacement containing neighborhood association features. In the cloud map, blue areas represent smaller displacements, and red areas represent larger displacements. For example, the area from the left arch waist to the arch top of the tunnel step is red, indicating a larger displacement.

[0146] Step 152: Perform time-series feature processing on the deformation rate data. Using the deformation rate of the current time step as the core, associate the deformation rate features of multiple previous time steps. Generate the enhanced deformation rate value of the current time step through feature transfer within the time-series window. Fit the trend features of the enhanced deformation rate value to generate a deformation rate time evolution curve containing time-series associated features. The curve contains the instantaneous value of the current time step and the trend change rate within the time-series window.

[0147] In this embodiment of the invention, the system sets the length of the timing window, which covers the current time step and a preset number of previous time steps, for example, the preset number is 5. The length of the timing window is determined based on the stage interval of tunnel construction, for example, the stage interval is a preset number of days.

[0148] Next, the system extracts the deformation rate value for each time step within the time window. The order of the time steps corresponds to the sequence of the construction process. For example, time step 1 corresponds to day 1 of the upper bench excavation stage, time step 2 corresponds to day 2 of the upper bench excavation stage, and so on.

[0149] Then, the system sets the time-series weight coefficient for each time step based on the order of the time steps. The weight coefficient of the current time step is the largest, and the weight coefficients of the preceding time steps decrease sequentially over time. The sum of the time-series weight coefficients is a preset value, for example, the preset value is 1. The weight coefficient is calculated as follows: the weight coefficient of the current time step is the preset value, the weight coefficient of the first preceding time step is the preset value, the weight coefficient of the second preceding time step is the preset value, and so on.

[0150] Next, the system multiplies the deformation rate value of each time step by the corresponding time-series weight coefficient to obtain the weighted deformation rate value for each time step. Then, the system performs a weighted summation aggregation on the weighted deformation rate values ​​of all time steps within the time-series window to generate the enhanced deformation rate value for the current time step. The weighted summation is calculated by adding the weighted deformation rate values ​​of all time steps. For example, the deformation rate value of the current time step is a preset value, the deformation rate values ​​of the previous 5 time steps are preset values, the corresponding time-series weight coefficients are preset values, the weighted deformation rate values ​​are preset values, and the enhanced deformation rate value is a preset value.

[0151] Next, the system iterates through all time steps, repeatedly setting the length of the time series window to aggregate the weighted deformation rate values ​​for all time steps within the window, generating an enhanced deformation rate value for each time step and creating an enhanced deformation rate dataset containing temporal correlation features. Then, the system fits the trend characteristics of the enhanced deformation rate values ​​using a linear regression method. The linear regression is performed with the time step as the independent variable and the enhanced deformation rate value as the dependent variable, fitting a linear equation. The slope of the linear equation represents the rate of change of the trend, reflecting the direction and speed of change in the enhanced deformation rate value. For example, a positive rate of change indicates an upward trend in the enhanced deformation rate value; a negative rate of change indicates a downward trend.

[0152] Finally, the system generates a deformation rate time evolution curve containing temporal correlation features. The X-axis of the curve is the time step, and the Y-axis is the enhanced deformation rate value. The curve includes the instantaneous enhanced deformation rate value at the current time step and the trend change rate within the time window. For example, the instantaneous value at the current time step is a preset value, and the trend change rate is a preset value.

[0153] Step 153: Based on the spatial expansion direction and spatial distribution characteristics of the potential deformation area, associate the displacement change trend of the unit nodes within the potential deformation area, mark the three-dimensional spatial boundary of the potential deformation area, and mark the boundary on the initial three-dimensional geological model using a preset style.

[0154] In this embodiment of the invention, the system, based on the spatial expansion direction of the potential deformation area (radial inward from the tunnel) and the spatial distribution characteristics of the displacement (the area from the left arch waist to the arch top of the tunnel upper step has a larger displacement), associates the displacement trend of the unit nodes within the potential deformation area (the displacement shows an upward trend), and marks the three-dimensional spatial boundary of the potential deformation area. The boundary is determined by connecting the unit nodes with the largest displacement within the potential deformation area to form a closed polygonal boundary. Next, the system marks the boundary on the initial three-dimensional geological model using a preset style, which is a red dashed line with a preset width, for example, 2mm.

[0155] Step 154: Create a unified scene coordinate system, import the three-dimensional spatial distribution cloud map of displacement, the time evolution curve of deformation rate, and the three-dimensional spatial boundary annotation of potential deformation area into the unified scene coordinate system, and perform spatiotemporal feature association of the unified scene coordinate system to construct a multi-view display framework covering different observation dimensions of tunnel rock strata deformation.

[0156] In this embodiment of the invention, the system establishes a unified scene coordinate system with the center point of the tunnel entrance, the tunnel extension direction, the vertical extension horizontal direction, and the vertical ground direction. The coordinate system unit is consistent with the initial three-dimensional geological model, for example, the unit is meters.

[0157] Next, the system maps the enhanced displacement values ​​of the three-dimensional spatial distribution cloud map of displacement to the corresponding positions in the unified scene coordinate system according to the coordinates of the unit nodes. The enhanced displacement value of each unit node is associated with the displacement features of its neighboring units. The enhanced displacement value is rendered as the corresponding color and superimposed on the initial three-dimensional geological model through the light and dark size strategy and the gradient trend strategy of the color mapping, so as to realize the spatial distribution features of displacement and the spatial position of the three-dimensional geological model.

[0158] Then, the system associates the time axis of the deformation rate time evolution curve with the tunnel construction progress in a unified scene coordinate system. The construction progress corresponds to the spatial extension position of the tunnel. For example, time step 1 corresponds to 0 meters of tunnel excavation, time step 2 corresponds to a preset number of meters of tunnel excavation, and so on. The Y-axis of the deformation rate time evolution curve is associated with the enhanced deformation rate value. Each time step node of the deformation rate time evolution curve is associated with the corresponding spatial position of the tunnel under the construction progress, and the instantaneous enhanced deformation rate value of the current time step and the trend change rate within the time series window are marked, so as to realize the spatiotemporal association between the deformation rate time series characteristics and the spatial position of tunnel construction.

[0159] Next, the system converts the coordinates of the three-dimensional spatial boundary annotation of the potential deformation region into coordinates under a unified scene coordinate system. Based on the spatial expansion direction and spatial gradient characteristics of the potential deformation region in the deformation trend, the boundary of the potential deformation region is defined on the initial three-dimensional geological model. The boundary nodes are associated with the displacement change trend and time-series deformation rate characteristics of the corresponding units, realizing the multi-dimensional association between the spatial boundary of the potential deformation region and the displacement and rate characteristics.

[0160] Finally, the system combines the spatial location association, the spatiotemporal association, and the multidimensional association to integrate the spatial distribution of displacement, the temporal trend of deformation rate, and the spatial expansion characteristics of potential deformation area into a unified scene coordinate system, generating a multi-view display framework covering different observation dimensions of tunnel rock strata deformation spatial distribution, rate temporal sequence, and regional expansion. The multi-view display framework includes a three-dimensional spatial view, a temporal view, and a regional expansion view.

[0161] Step 155: Under the multi-view display framework, through the correlation display of spatial distribution characteristics, temporal change characteristics and regional expansion characteristics, the overall spatial distribution of tunnel rock strata deformation, the temporal trend of rate and the three-dimensional expansion correlation of potential deformation areas are presented, and visualization of deformation early warning information containing multiple perspectives is generated.

[0162] In this embodiment of the invention, under a multi-view display framework, the system displays a three-dimensional spatial distribution cloud map of displacement through spatial distribution features, presenting the overall spatial distribution of tunnel rock strata deformation. For example, the area from the left arch waist to the arch top of the tunnel upper step is red, indicating a large displacement. The system also displays a time evolution curve of deformation rate through temporal change features, presenting the temporal trend of deformation rate. For example, an upward trend in the curve indicates that the deformation rate is increasing. Finally, the system displays three-dimensional spatial boundary markings of potential deformation areas through region expansion features, presenting the three-dimensional expansion relationship of potential deformation areas. For example, the boundary expands radially inward towards the tunnel.

[0163] Finally, the system generates visualized deformation early warning information from multiple perspectives. The early warning information is in the form of a three-dimensional visualization model, which includes a three-dimensional spatial distribution cloud map of displacement, a time evolution curve of deformation rate, and three-dimensional spatial boundary markings of potential deformation areas. The content of the early warning information includes the location, deformation rate, and expansion direction of the potential deformation area. For example, the potential deformation area is located from the left arch waist to the arch top of the tunnel upper step, the deformation rate is a preset value, and the expansion direction is radially inward from the tunnel.

[0164] As an optional embodiment, the method further includes:

[0165] Step 210: During tunnel construction, three-dimensional point cloud data of the current tunnel rock strata are collected by three-dimensional laser scanning at preset intervals, and the collection range covers the rock strata surface of the excavated area and the unexcavated area of ​​the tunnel.

[0166] In this embodiment of the invention, the system controls a 3D laser scanner to be set up at a preset location in the excavated area of ​​the tunnel every preset time (e.g., 7 days), sets the scanning resolution to a preset parameter, and the scanning range covers a preset distance range covering both the excavated and unexcavated areas of the tunnel. It collects 3D point cloud data of the current tunnel rock strata, and the collection range covers the surface of the rock strata in both the excavated (e.g., 100 meters excavated) and unexcavated (e.g., 50 meters unexcavated) areas of the tunnel.

[0167] Step 220: Register the current 3D point cloud data with the initial 3D point cloud data, identify the corresponding feature points through the feature point matching algorithm, calculate the coordinate transformation matrix of the feature points, and convert the current point cloud data into coordinates in the initial coordinate system.

[0168] In this embodiment of the invention, the system uses the SIFT (Scale Invariant Feature Transform) algorithm to extract feature points from the current 3D point cloud data and the initial 3D point cloud data. The number of feature points is a preset number, such as 1000. Next, the system identifies corresponding feature points using a feature point matching algorithm. The matching method involves calculating the descriptor similarity of feature points; feature points with a similarity greater than a preset value are identified as the corresponding feature points. Then, the system calculates the coordinate transformation matrix of the feature points. The coordinate transformation matrix is ​​calculated by fitting the coordinates of the corresponding feature points using the least squares method to obtain a rotation matrix and a translation vector. The coordinate transformation matrix is ​​a combination of the rotation matrix and the translation vector. Finally, the system converts the current point cloud data to coordinates in the initial coordinate system by multiplying the coordinates of the current point cloud data by the rotation matrix and adding the translation vector.

[0169] Step 230: Calculate the coordinate difference between the current point cloud data and the initial 3D geological model. Adjust the surface of the initial 3D geological model based on the coordinate difference. Use the adjusted model as the current 3D geological model. The current model reflects the actual spatial morphology of the rock strata after tunnel construction.

[0170] In this embodiment of the invention, the system calculates the coordinate difference between the current point cloud data and the initial 3D geological model. The coordinate difference is calculated by comparing the coordinates of each point in the current point cloud data with the coordinates of the corresponding point in the initial 3D geological model, thus obtaining a coordinate difference dataset. Next, the system adjusts the surface of the initial 3D geological model based on the coordinate difference. The adjustment method involves using the moving least squares method to fit the coordinate difference dataset onto the surface of the initial 3D geological model, adjusting the vertex coordinates of the surface to minimize the deviation between the adjusted surface and the current point cloud data. Finally, the system uses the adjusted model as the current 3D geological model. The current model reflects the actual spatial morphology of the rock strata after tunnel construction; for example, the position of the left arch waist of the tunnel step in the current model is offset inward by a preset distance compared to the initial model.

[0171] Step 240: Import the current 3D geological model into the rock mechanics analysis model, replace the original initial model, and re-generate the unstructured mesh, keeping the mesh size consistent with the initial model.

[0172] In this embodiment of the invention, the system imports the current three-dimensional geological model into the rock strata mechanics analysis model, replacing the original initial model. Next, the system re-generates the unstructured mesh using tetrahedral elements, maintaining the same mesh size as the initial model, for example, 0.5 meters. The mesh is then refined at the interface between sandy mudstone and shale, with a preset refinement coefficient.

[0173] Step 250: Collect construction parameters and process information for the current construction stage, and update the constraints, element activation sequences, and force application data of the rock mechanics analysis model.

[0174] In this embodiment of the invention, the system collects construction parameters and process information for the current construction stage, which is the middle bench excavation stage. Construction parameters include middle bench support parameters (steel arch frame model is a preset model, spacing is a preset distance, and shotcrete thickness is a preset thickness), excavation sequence (middle bench excavation → middle bench support), and construction loads (earth pressure is a preset value, and the self-weight of the support structure is a preset value). Next, the system updates the constraints of the rock mechanics analysis model, converting the middle bench support parameters into constraint conditions with a preset constraint stiffness and a constraint direction in the tunnel radial direction; it updates the unit activation sequence, converting the middle bench excavation process into unit nodes in the middle bench region of the activation model, and converting the middle bench support process into unit nodes in the middle bench support region of the activation model; and it updates the force application data, converting the middle bench construction load into uniformly distributed forces and concentrated forces applied to the corresponding unit nodes.

[0175] Step 260: Run the updated rock mechanics analysis model to simulate the stress distribution state at the current construction stage and predict the target deformation trend through iterative calculation.

[0176] In this embodiment of the invention, the system runs the updated rock mechanics analysis model according to the method of steps 130 to 146, simulates the stress distribution state of the current construction stage, and predicts the target deformation trend through iterative calculation. The target deformation trend includes the displacement distribution, deformation rate distribution and potential deformation area expansion direction of the current construction stage.

[0177] Step 270: Compare the target deformation trend with the deformation trend, calculate the changes in displacement and deformation rate. If the changes exceed a preset threshold, regenerate deformation warning information containing the current multi-view perspective.

[0178] In this embodiment of the invention, the system compares the target deformation trend with the deformation trend generated in step 146, and calculates the change in displacement (current displacement minus initial displacement) and the change in deformation rate (current deformation rate minus initial deformation rate). If the change exceeds a preset threshold (e.g., the change in displacement exceeds 5 mm, and the change in deformation rate exceeds 0.5 mm / day), the system regenerates deformation warning information containing the current multi-view data according to the methods in steps 150 to 155.

[0179] Step 280: Store the current 3D geological model, mechanical analysis model, deformation trend and early warning information to the tunnel engineering database to generate a deformation analysis log during the construction process.

[0180] In this embodiment of the invention, the system stores the current 3D geological model, the updated rock strata mechanical analysis model, the target deformation trend, and the regenerated deformation early warning information into a tunnel engineering database. The database is stored in a preset format, such as JSON. Next, the system generates a deformation analysis log during construction. The log includes the construction time, the current 3D geological model version, the mechanical analysis model parameters, the deformation trend, and the early warning information. The log is in tabular form, with columns including construction time, model version, parameters, deformation trend, and early warning information.

[0181] As an optional embodiment, the method further includes:

[0182] Step 310: Obtain real-time feedback data on tunnel construction adjustments after generating deformation early warning information. The real-time feedback data includes the adjusted support parameter sequence, construction procedure change records, and rock surface displacement monitoring sequence for the corresponding time period.

[0183] In this embodiment of the invention, the system acquires real-time feedback data on tunnel construction adjustments after generating deformation early warning information. The adjusted support parameter sequence is to increase the spacing of the steel arch frame in the area from the left arch waist to the arch crown of the tunnel upper step (adjusted from a preset distance to a preset distance) and increase the thickness of the shotcrete (adjusted from a preset thickness to a preset thickness). The construction procedure change record is to delay the middle step excavation procedure by a preset number of days. The rock surface displacement monitoring sequence for the corresponding time period is the displacement of the area from the left arch waist to the arch crown of the tunnel upper step, which is monitored by a total station. The monitoring frequency is once a day, and the monitoring period is 14 days after the early warning information is generated.

[0184] Step 320: Extract the adjustment support feature vector, process change feature matrix and displacement monitoring feature cluster from the real-time feedback data, and align the adjustment support feature vector with the support constraint features in the original deformation early warning information in the feature space.

[0185] In this embodiment of the invention, the system extracts adjustment support feature vectors from real-time feedback data. These feature vectors include the spacing of the steel arch frames, the thickness of the shotcrete, and the support stiffness; the vector dimension is 3. The system also extracts a process change feature matrix, which includes the process name, change time, and change content; the matrix dimension is a preset number of rows × a preset number of columns. Finally, the system extracts displacement monitoring feature clusters, which include the coordinates of monitoring points, the displacement amount, and the monitoring time; the number of clusters is a preset number. Next, the system aligns the adjustment support feature vectors with the support constraint features in the original deformation warning information in the feature space. The alignment method uses Principal Component Analysis (PCA) to transform the adjustment support feature vectors and support constraint features into the same feature space; the feature space dimension is a preset dimension, such as 2D.

[0186] Step 330: Input the aligned adjustment support feature vector and process change feature matrix into the feature association layer and displacement monitoring feature cluster of the original rock strata mechanical analysis model for feature matching to obtain the feature matching difference degree.

[0187] In this embodiment of the invention, the system inputs the aligned adjustment support feature vector and the process change feature matrix into the feature association layer of the original rock strata mechanical analysis model. The feature association layer is a fully connected layer, with the input dimension being the sum of the dimensions of the adjustment support feature vector and the process change feature matrix, and the output dimension being a preset dimension, such as 5 dimensions. Next, the system performs feature matching between the output of the feature association layer and the displacement monitoring feature cluster. The matching method is to calculate the cosine similarity, with the similarity value ranging from -1 to 1. The feature matching difference is 1 minus the cosine similarity, with the difference value ranging from 0 to 2.

[0188] Step 340: Generate model correction factors based on feature matching difference. The model correction factors correspond to the weight coefficients of feature extraction weights and constraint stiffness adjustment rules for potential deformation regions in the model.

[0189] In this embodiment of the invention, the system generates a model correction factor based on the feature matching difference. The model correction factor is calculated as follows: the model correction factor equals the feature matching difference multiplied by a preset coefficient, and the coefficient ranges from 0 to 1. The model correction factor corresponds to adjusting the feature extraction weights of potential deformation regions in the model (e.g., adjusting the feature extraction weights of shear stress concentration regions from a preset value to a preset value) and the weight coefficients of the constraint stiffness adjustment rules (e.g., adjusting the weight coefficient for increasing constraint stiffness from a preset value to a preset value).

[0190] Step 350: Apply the model correction factor to the rock mechanics analysis model, update the feature association layer weights and constraint adjustment logic of the rock mechanics analysis model, re-analyze the real-time three-dimensional point cloud data of the current tunnel rock strata, and generate corrected deformation early warning information. The corrected early warning information includes the adjusted potential deformation area boundary, deformation rate trend, and support optimization suggestions.

[0191] In this embodiment of the invention, the system applies the model correction factor to the rock mechanics analysis model, updates the weights of the feature-related layers (by multiplying the weight matrix by the model correction factor) and the constraint adjustment logic (by multiplying the weight coefficients of the constraint stiffness adjustment rules by the model correction factor). Next, the system re-analyzes the real-time three-dimensional point cloud data of the current tunnel rock strata (the current three-dimensional point cloud data collected in step 210), and generates corrected deformation warning information according to the methods in steps 110 to 155. The corrected warning information includes the adjusted potential deformation area boundary (the boundary shrinks radially outward from the tunnel by a preset distance), the deformation rate trend (the deformation rate decreases from a preset value to a preset value), and support optimization suggestions (maintaining the current support parameters and periodically monitoring displacement).

[0192] The tunnel rock strata large deformation analysis method based on three-dimensional modeling in this invention constructs an initial three-dimensional geological model by acquiring initial three-dimensional point cloud data through three-dimensional laser scanning, establishes a rock strata mechanical analysis model by performing unstructured mesh division, simulates stress distribution by inputting construction parameters and process information, iteratively calculates and predicts deformation trends, generates multi-view visualized deformation early warning information, and updates the model and corrects the early warning information in real time during construction, which can accurately analyze the large deformation of tunnel rock strata.

[0193] In summary, this invention acquires initial 3D point cloud data using 3D laser scanning and constructs an initial 3D geological model that includes spatial morphological features and the geometric topological relationships of rock strata interfaces. A rock strata mechanical analysis model is then established using finite element unstructured mesh generation and physical and mechanical parameters. Furthermore, stress distribution is simulated through dynamic input of construction parameters and process information, and activation of unit nodes. Iterative calculations are used to identify potential deformation areas and adjust boundary conditions to predict deformation trends. Finally, multi-dimensional features are integrated to generate multi-perspective early warning information. From an overall logical perspective, this invention achieves a closed-loop process from 3D geological modeling to mechanical analysis, deformation prediction, and then to visual early warning, overcoming the technical bottlenecks of traditional tunnel rock strata deformation analysis, such as the disconnect between geological and mechanical models, the lack of dynamic iterative adjustment in deformation prediction, and the single dimension of early warning information.

[0194] In detail, the high-precision spatial data from 3D laser scanning is deeply integrated with the mechanical modeling of finite element analysis. The actual stress evolution during construction is restored through a process-driven unit node activation mechanism. Then, potential deformation areas are accurately located through the collaborative identification of shear stress concentration areas and stress change gradients. Combined with the dynamic adjustment of boundary conditions, high-precision prediction of deformation trends is achieved. Finally, intuitive multi-view early warning information is generated through multi-dimensional feature fusion. This provides an integrated technical solution for tunnel construction safety, from data acquisition to decision support, improving the accuracy, dynamism, and visualization of large deformation analysis of tunnel rock strata, and effectively reducing construction risks.

[0195] See Figure 2 As shown, the tunnel rock strata large deformation analysis device based on 3D modeling includes:

[0196] The geological model construction module is used to acquire initial three-dimensional point cloud data of tunnel rock strata through three-dimensional laser scanning, and construct an initial three-dimensional geological model based on the spatial coordinate matching and surface reconstruction processing of the initial three-dimensional point cloud data. The initial three-dimensional geological model includes the spatial morphological features of the tunnel rock strata and the geometric topological relationship of the rock strata interface.

[0197] The mechanical analysis and modeling module is used to perform unstructured mesh generation based on finite element method on the initial three-dimensional geological model, and to establish a rock strata mechanical analysis model in combination with the physical and mechanical parameters of the tunnel rock strata.

[0198] The stress distribution analysis module is used to input construction parameters and process information during tunnel construction into the rock mechanics analysis model, activate the unit nodes of the corresponding areas according to the process information, and simulate the stress distribution state of the tunnel rock strata during construction by solving the force balance equations of the unit nodes.

[0199] The deformation trend prediction module is used to perform iterative calculations on the rock strata mechanical analysis model. It identifies potential deformation areas by identifying shear stress concentration areas and stress change gradients, adjusts the constraint stiffness and constraint direction of the unit nodes in the potential deformation areas to update the boundary conditions, reapplies the corrected construction parameters to calculate the updated stress distribution state, and predicts the deformation trend of the tunnel rock strata based on the updated stress distribution state.

[0200] The deformation early warning visualization module is used to generate a three-dimensional spatial distribution cloud map and a time evolution curve based on the deformation trend, and to mark the three-dimensional spatial position of the potential deformation region based on the spatial expansion direction of the potential deformation region in the deformation trend. The three-dimensional spatial distribution cloud map, the time evolution curve and the three-dimensional spatial position marking of the potential deformation region are subjected to multi-dimensional feature fusion processing to generate deformation early warning information that includes multi-view display.

[0201] See Figure 3As shown in the figure, this is a schematic diagram of the basic structure of a tunnel rock strata large deformation analysis system 200 provided in an embodiment of the present invention. The tunnel rock strata large deformation analysis system 200 includes:

[0202] Processor 201;

[0203] Storage device 202, on which computer program 2020 is stored;

[0204] Processor 201 and storage device 202 are communicatively connected via bus 203;

[0205] When the computer program 2020 is executed by the processor 201, the processor 201 implements any of the three-dimensional modeling-based tunnel rock strata large deformation analysis methods described above.

[0206] Based on the above, a readable storage medium is provided, on which a program or instructions are stored, and when the program or instructions are executed by a processor, the steps of the above method are implemented.

[0207] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems or apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and relevant parts can be referred to the method section.

Claims

1. A method for analyzing large deformations of tunnel rock strata based on three-dimensional modeling, characterized in that, include: Initial three-dimensional point cloud data of the tunnel rock strata were acquired by three-dimensional laser scanning. Based on the spatial coordinate matching and surface reconstruction processing of the initial three-dimensional point cloud data, an initial three-dimensional geological model was constructed. The initial three-dimensional geological model includes the spatial morphological features of the tunnel rock strata and the geometric topological relationship of the rock strata interface. The initial three-dimensional geological model was subjected to unstructured mesh generation based on finite element method, and a rock strata mechanical analysis model was established by combining the physical and mechanical parameters of the tunnel rock strata. The construction parameters and process information during tunnel construction are input into the rock mechanics analysis model. The unit nodes of the corresponding areas are activated sequentially according to the process information. The stress distribution state of the tunnel rock strata during construction is simulated by solving the force balance equation of the unit nodes. The rock strata mechanical analysis model is iteratively calculated. Potential deformation areas are determined by identifying shear stress concentration areas and stress change gradients. The constraint stiffness and constraint direction of the unit nodes in the potential deformation areas are adjusted to update the boundary conditions. The corrected construction parameters are reapplied to calculate the updated stress distribution state. Based on the updated stress distribution state, the deformation trend of the tunnel rock strata is predicted. A three-dimensional spatial distribution cloud map and a time evolution curve are generated based on the deformation trend. The three-dimensional spatial location of the potential deformation region is marked based on the spatial expansion direction of the potential deformation region in the deformation trend. The three-dimensional spatial distribution cloud map, the time evolution curve and the three-dimensional spatial location marking of the potential deformation region are subjected to multi-dimensional feature fusion processing to generate deformation early warning information with multi-view display.

2. The method for analyzing large deformations of tunnel rock strata based on three-dimensional modeling as described in claim 1, characterized in that, The process involves inputting construction parameters and procedure information during tunnel construction into a rock mechanics analysis model, sequentially activating corresponding unit nodes according to the procedure information, and simulating the stress distribution state of the tunnel rock strata during construction by solving the force balance equations of the unit nodes. This includes: The support parameters, excavation sequence, and construction loads extracted from the tunnel construction plan are classified, disassembled, and differentiated respectively. The classified support parameters are converted into constraints of the rock mechanics analysis model; the disassembled excavation sequence is converted into the unit activation sequence of the model, and the unit nodes of the corresponding area are activated sequentially according to the excavation time interval of each area; the classified construction loads are converted into force application data of the model. Activate the unit nodes of the tunnel excavation area in sequence according to the unit activation sequence, simulate the stress release of the rock strata during the excavation process, apply the corresponding uniformly distributed force or concentrated force to the activated unit nodes, and establish the force balance equation of the unit nodes based on the constitutive relation of the finite element element. Solving the force equilibrium equations yields the normal and shear stress components at the element nodes, and the stress distribution inside the element is calculated by interpolation using the element shape function, generating stress data for each element. The stress data of all units are globally assembled, the principal stress direction is calculated by combining the geometric topological relationship of the rock layer interface, the shear stress concentration area is determined by filtering the shear stress value threshold, and the stress change gradient is calculated by the spatial derivative of the stress value, generating a stress distribution state that includes the principal stress direction, the shear stress concentration area and the stress change gradient.

3. The method for analyzing large deformations of tunnel rock strata based on three-dimensional modeling as described in claim 1, characterized in that, The iterative calculation of the rock strata mechanical analysis model involves identifying potential deformation regions by recognizing shear stress concentration areas and stress change gradients, adjusting the constraint stiffness and constraint direction of the unit nodes in the potential deformation regions to update the boundary conditions, reapplying the corrected construction parameters to calculate the updated stress distribution state, and predicting the deformation trend of the tunnel rock strata based on the updated stress distribution state, including: The iteration termination condition is set as follows: the difference in stress distribution between two adjacent iterations is less than a preset range and the number of iterations does not exceed the upper limit. In the first iteration, the initial stress distribution state is calculated based on the initial boundary conditions and construction parameters; Shear stress concentration areas and stress change gradients are extracted from the initial stress distribution state. The boundaries of shear stress concentration areas are screened by the shear stress value exceeding the shear strength threshold of the rock strata. The stress transmission path is determined by the dominant direction vector analysis of the stress change gradient. The potential deformation area is determined by combining the boundaries of the shear stress concentration areas and the stress transmission path. The initial constraint stiffness and constraint direction of the unit nodes in the potential deformation area are obtained. The constraint stiffness of the nodes with shear stress values ​​exceeding the rock shear strength threshold is increased, and the constraint stiffness of the nodes with shear stress values ​​below the rock shear strength threshold is decreased. The constraint direction is adjusted to be consistent with the dominant stress transmission direction to complete the boundary condition update. The construction parameters are corrected based on the updated boundary conditions. The support reaction parameters are adjusted in areas where the constraint stiffness increases, and the earth pressure parameters are adjusted in areas where the constraint stiffness decreases. The corrected construction parameters are then applied to the element nodes. The force equilibrium equations are re-solved to obtain the updated stress distribution state. The difference between the stress distribution and the initial stress distribution is calculated. If the difference does not meet the termination condition, the boundary conditions and parameter application steps are repeated until the termination condition is met. Based on the stress distribution at termination, the strain distribution of the rock strata is calculated through the stress-strain constitutive relationship to generate the deformation trend; wherein, the displacement is obtained by integrating the strain, the derivative of the displacement with respect to time is the deformation rate, and the spatial gradient direction of the displacement is the direction of expansion of the potential deformation region.

4. The method for analyzing large deformations of tunnel rock strata based on three-dimensional modeling as described in claim 3, characterized in that, The process involves extracting shear stress concentration regions and stress gradients from the initial stress distribution state, screening the boundaries of shear stress concentration regions by ensuring the shear stress values ​​exceed the shear strength threshold of the rock strata, determining the stress transmission path through dominant direction vector analysis of the stress gradient, and identifying potential deformation regions by combining the boundaries of the shear stress concentration regions and the stress transmission paths. Extract the shear stress values ​​of all elements from the initial stress distribution state, compare the shear stress values ​​with the shear strength of the rock layer, and screen out elements whose shear stress values ​​exceed the shear strength threshold of the rock layer; The selected units are spatially clustered, and the boundaries of the shear stress concentration region are determined by the density clustering algorithm. The units within the boundary constitute the shear stress concentration region. Calculate the stress change gradient of each element in the shear stress concentration region. The direction of the gradient vector is the direction of stress change. Statistically analyze the directional distribution of all gradient vectors. The direction with the highest frequency is the dominant direction of stress transmission. Draw the stress transfer path along the dominant direction; the area covered by the path is the stress transfer influence area. The overlapping area between the shear stress concentration region and the stress transfer influence region is defined as the potential deformation region, and the elements within the overlapping area are potential deformation elements.

5. The method for analyzing large deformations of tunnel rock strata based on three-dimensional modeling as described in claim 3, characterized in that, The process involves obtaining the initial constraint stiffness and constraint direction of the element nodes within the potential deformation region, increasing the constraint stiffness of nodes whose shear stress exceeds the rock shear strength threshold, decreasing the constraint stiffness of nodes whose shear stress is below the rock shear strength threshold, and adjusting the constraint direction to align with the dominant stress transmission direction to complete the boundary condition update. This includes: Traverse all element nodes within the potential deformation region and extract the node coordinates and corresponding shear stress values ​​for each element node; Query the initial constraint stiffness and constraint direction of the unit node; where the initial constraint stiffness is set based on the original physical and mechanical parameters of the rock strata, and the initial constraint direction is set based on the tunnel axis direction; For nodes where the shear stress exceeds the rock shear strength threshold, the constraint stiffness is increased to a set multiple of the initial value to simulate the constraint effect of reinforced support; for nodes where the shear stress is lower than the rock shear strength threshold, the constraint stiffness is reduced to a set multiple of the initial value to simulate the relaxation effect of stress release. Calculate the dominant direction vector of stress transfer, and adjust the constraint direction of the element node to be consistent with the dominant direction vector so that the constraint direction matches the stress transfer direction; The adjusted constraint stiffness and constraint direction are reassigned to the element nodes, and the boundary conditions of the potential deformation region are updated.

6. The method for analyzing large deformations of tunnel rock strata based on three-dimensional modeling as described in claim 1, characterized in that, The process involves generating a three-dimensional spatial distribution cloud map and a time evolution curve based on the deformation trend, and marking the three-dimensional spatial location of potential deformation regions based on the spatial expansion direction of potential deformation regions in the deformation trend. The three-dimensional spatial distribution cloud map, the time evolution curve, and the three-dimensional spatial location markings of potential deformation regions are then subjected to multi-dimensional feature fusion processing to generate deformation early warning information that includes multi-view displays, including: Extract the displacement, deformation rate, and spatial expansion direction information of the potential deformation area from the deformation trend, obtain the displacement and corresponding spatial coordinates of each unit node, obtain the deformation rate value of each time step, and the time step corresponds to the stage interval of the construction process. Spatial feature processing is performed on the displacement data. The displacement features of the neighboring units are associated with the unit node as the center. The enhanced displacement value of the unit node is generated by aggregating the features of the neighboring units. Based on the spatial distribution features of the enhanced displacement value, a color mapping rule is set to map the enhanced displacement value into a color and render it onto the initial three-dimensional geological model to generate a three-dimensional spatial distribution cloud map of displacement containing neighborhood association features. The deformation rate data is processed using time-series features. The deformation rate at the current time step is used as the core to associate the deformation rate features of multiple previous time steps. The enhanced deformation rate value at the current time step is generated through feature transfer within the time-series window. The trend features of the enhanced deformation rate value are fitted to generate a deformation rate time evolution curve containing time-series associated features. The curve contains the instantaneous value at the current time step and the trend change rate within the time-series window. Based on the spatial expansion direction and spatial distribution characteristics of the potential deformation area, the displacement change trend of the unit nodes within the potential deformation area is associated, the three-dimensional spatial boundary of the potential deformation area is marked, and the boundary is marked on the initial three-dimensional geological model using a preset style. A unified scene coordinate system is created, and the three-dimensional spatial distribution cloud map of displacement, the time evolution curve of deformation rate, and the three-dimensional spatial boundary annotation of potential deformation area are imported into the unified scene coordinate system and the spatiotemporal features of the unified scene coordinate system are correlated to construct a multi-view display framework covering different observation dimensions of tunnel rock strata deformation. Within a multi-perspective display framework, the overall spatial distribution, temporal variation, and regional expansion characteristics of tunnel rock strata deformation are presented through the interconnected display of spatial distribution features, temporal variation features, and regional expansion features. This generates visualized deformation early warning information with multiple perspectives.

7. The method for analyzing large deformations of tunnel rock strata based on three-dimensional modeling as described in claim 6, characterized in that, The spatial feature processing of the displacement data, which associates the displacement features of neighboring units with the unit node as the center, and generates enhanced displacement values ​​for the unit node through feature aggregation of neighboring units, includes: Centered on each unit node, a neighborhood range is set to cover a preset number of unit nodes around it, generating a set of neighboring units for each unit node. The size of the neighborhood range is determined based on the structural characteristics of the tunnel rock strata. Extract the displacement value and corresponding spatial coordinates of each node in the neighborhood unit set, and calculate the spatial distance between each neighborhood unit and the central unit node based on the Euclidean distance in the three-dimensional coordinate system. The weight coefficient of the neighborhood unit is set based on the spatial distance. The smaller the spatial distance, the larger the weight coefficient, and the larger the spatial distance, the smaller the weight coefficient. The value range of the weight coefficient is within a preset range. Multiply the displacement value of each neighboring cell by the corresponding weight coefficient to obtain the weighted displacement value of the neighboring cell. The weighted average of the weighted displacement values ​​of all neighboring units is aggregated to generate the enhanced displacement value of the central unit node; the enhanced displacement value integrates the displacement characteristics of the central unit node itself and the displacement characteristics of the neighboring units. Traverse all unit nodes in the initial 3D geological model, repeat the steps of setting a neighborhood range covering a preset number of unit nodes around each unit node as the center, generating a set of neighboring units of each unit node, performing a weighted average aggregation process on the weighted displacement values ​​of all neighboring units, generating the enhanced displacement value of the central unit node, generating the enhanced displacement value of each unit node, and generating an enhanced displacement dataset containing spatial neighborhood association features. Based on the spatial distribution characteristics of the enhanced displacement dataset, the differences in enhanced displacement values ​​of unit nodes in different regions are associated according to the lightness and darkness strategy and the gradient trend strategy of color mapping.

8. The method for analyzing large deformations of tunnel rock strata based on three-dimensional modeling as described in claim 6, characterized in that, The step of performing time-series feature processing on deformation rate data, using the deformation rate at the current time step as the core and associating it with the deformation rate features of multiple previous time steps, generates an enhanced deformation rate value for the current time step through feature transfer within the time-series window, including: Set the length of the timing window. The timing window covers the current time step and a preset number of previous time steps. The length of the timing window is determined based on the stage interval of the construction process. Extract the deformation rate value for each time step within the time window; the order of the time steps corresponds to the sequence of the construction process. Based on the order of time steps, a time series weight coefficient is set for each time step. The weight coefficient of the current time step is the largest, and the weight coefficients of the preceding time steps decrease sequentially over time. The sum of the time series weight coefficients is a preset value. The weighted deformation rate value for each time step is obtained by combining the deformation rate value for each time step with the corresponding temporal weighting coefficient. The weighted deformation rate values ​​of all time steps within the time window are aggregated by weighted summation to generate the enhanced deformation rate value of the current time step. The enhanced deformation rate value integrates the deformation rate characteristics of the current time step and the deformation rate characteristics of the previous time step. Iterate through all time steps, repeat the step of setting the length of the time window to aggregate the weighted deformation rate values ​​of all time steps within the time window, generate the enhanced deformation rate value for each time step, and generate an enhanced deformation rate dataset containing time-series correlation features. Based on the enhanced deformation rate dataset, the changes in enhanced deformation rate values ​​at the current time step and the previous time step are correlated. The trend characteristics of the enhanced deformation rate values ​​are fitted using a linear regression method to generate the trend change rate of the deformation rate time evolution curve. The trend change rate reflects the direction and speed of change of the enhanced deformation rate values.

9. The method for analyzing large deformations of tunnel rock strata based on three-dimensional modeling as described in claim 6, characterized in that, The process of creating a unified scene coordinate system involves importing the three-dimensional spatial distribution cloud map of displacement, the time evolution curve of deformation rate, and the three-dimensional spatial boundary annotation of potential deformation areas into the unified scene coordinate system, and performing spatiotemporal feature correlation within the unified scene coordinate system. This constructs a multi-view display framework covering different observation dimensions of tunnel rock strata deformation, including: A unified scene coordinate system is established with the center point of the tunnel entrance, the tunnel extension direction, the horizontal direction of the vertical extension, and the direction perpendicular to the ground. The coordinate system unit is consistent with the initial three-dimensional geological model. The enhanced displacement values ​​of the three-dimensional spatial distribution cloud map of displacement are mapped to the corresponding positions in the unified scene coordinate system according to the coordinates of the unit nodes. The enhanced displacement value of each unit node is associated with the displacement features of its neighboring units. The enhanced displacement values ​​are rendered as corresponding colors and superimposed on the initial three-dimensional geological model through the light and dark size strategy and the gradient trend strategy of the color mapping, so as to realize the spatial distribution features of displacement and the spatial position of the three-dimensional geological model. The time axis of the deformation rate time evolution curve is associated with the tunnel construction progress in a unified scene coordinate system. The construction progress corresponds to the spatial extension position of the tunnel. The Y-axis of the deformation rate time evolution curve corresponds to the enhanced deformation rate value. Each time step node of the deformation rate time evolution curve is associated with the corresponding spatial position of the tunnel under the construction progress. The instantaneous enhanced deformation rate value of the current time step and the trend change rate within the time series window are marked to realize the spatiotemporal association between the temporal characteristics of deformation rate and the spatial position of tunnel construction. The coordinates of the three-dimensional spatial boundary of the potential deformation region are converted into coordinates under a unified scene coordinate system. Based on the spatial expansion direction and spatial gradient characteristics of the potential deformation region in the deformation trend, the boundary of the potential deformation region is defined on the initial three-dimensional geological model. The boundary nodes are associated with the displacement change trend and time-series deformation rate characteristics of the corresponding units, so as to realize the multi-dimensional association between the spatial boundary of the potential deformation region and the displacement and rate characteristics. By combining the spatial location association, the spatiotemporal association, and the multidimensional association, the spatial distribution of displacement, the temporal trend of deformation rate, and the spatial expansion characteristics of potential deformation area are integrated into a unified scene coordinate system to generate a multi-view display framework covering different observation dimensions of tunnel rock strata deformation spatial distribution, rate temporal sequence, and regional expansion.

10. A tunnel rock strata large deformation analysis system, characterized in that, include: processor; A storage device storing a computer program, which, when executed by the processor, causes the processor to implement the tunnel rock strata large deformation analysis method based on three-dimensional modeling as described in any one of claims 1-9.

Images