Method for monitoring large deformation stress release of soft rock tunnel

Abstract

The invention provides a soft rock tunnel large deformation stress release monitoring method, and belongs to the technical field of soft rock tunnel deformation monitoring. Stress component data and displacement vector data of continuous 180 days after excavation are collected, a principal stress value is obtained through coordinate transformation processing, and the stress release rate is calculated; a surrounding rock stress release spatial-temporal characteristic matrix is constructed, singular value decomposition is carried out to obtain a dominant stress release mode, a threshold judgment system based on the principal stress release rate is established to divide the surrounding rock state into a rapid release state, a slow release state and a stable state, and the monitoring frequency and the lining construction time are adjusted according to the judgment result. And a crack network topological graph model is constructed, a crack network connectivity index is calculated through a graph connectivity analysis algorithm, whether a chemical grouting repair program is started or not is judged, and the problem that the judgment of the lining construction opportunity is inaccurate due to the fact that spatial-temporal evolution characteristics in the soft rock tunnel stress release process are difficult to accurately recognize is solved.

Description

Technical Field

[0001] This invention belongs to the field of soft rock tunnel deformation monitoring technology, specifically, it relates to a method for monitoring stress release during large deformations in soft rock tunnels. Background Technology

[0002] In soft rock tunnel engineering, due to the low strength and significant rheological properties of soft rock, the surrounding rock stress undergoes a long release process after excavation. Existing monitoring technologies mainly employ a combination of single-point discrete stress monitoring and empirical judgment to assess surrounding rock stability. This involves deploying a limited number of stress gauges and convergence meters at the tunnel arch or sidewalls to acquire stress and displacement data for local areas of the monitoring section. Based on stress or displacement thresholds set through engineering experience, the suitability for lining construction is determined. However, existing methods have significant shortcomings in practical application. On the one hand, single-point monitoring data cannot reflect the spatial distribution differences and temporal evolution of stress release, leading to insufficient understanding of the overall characteristics of the stress release process. On the other hand, the judgment criteria based on fixed thresholds do not consider the dynamic changes in stress release rate, easily resulting in premature lining construction leading to excessive structural loads or delayed lining construction causing project delays. In other words, existing technologies suffer from the technical problem of inaccurate identification of the spatiotemporal evolution characteristics of stress release in soft rock tunnels, leading to inaccurate determination of the timing of lining construction. Summary of the Invention

[0003] In view of this, the present invention provides a method for monitoring stress release during large deformation in soft rock tunnels, which can solve the technical problem in the prior art where the spatiotemporal evolution characteristics of stress release in soft rock tunnels are difficult to accurately identify, leading to inaccurate determination of the timing of lining construction.

[0004] This invention is implemented as follows: A method for monitoring stress release during large deformation in soft rock tunnels is provided. Monitoring sections are set up every 5m to 15m along the tunnel axis behind the excavation face. At each monitoring section, triaxial stress sensors and displacement sensors are embedded at the arch crown, arch waist, sidewalls, and invert, forming a spatial stress-displacement monitoring network. Stress component data and displacement vector data are collected from each monitoring section over 180 consecutive days after excavation. Coordinate transformation is performed on the stress component data to obtain the principal stress direction angle and principal stress value. The stress release rate and displacement rate at each measuring point are calculated. A spatiotemporal characteristic matrix of surrounding rock stress release is constructed, and singular value decomposition is performed on the matrix to obtain the dominant stress release mode. A stress release rate threshold discrimination system is established. The monitoring frequency and lining construction timing are adjusted based on the discrimination results. A crack network topology model is constructed, and a graph connectivity analysis algorithm is used to calculate the crack network connectivity index and the size of the maximum connected subgraph.

[0005] Among them, the triaxial stress sensor is a pressure sensor that can simultaneously measure stress components in three orthogonal directions. Each triaxial stress sensor contains three mutually perpendicular measurement units, which correspond to the radial, tangential and axial stress components of the tunnel, respectively.

[0006] Among them, the displacement sensor is a sensing device that measures the three-dimensional displacement vector of the measuring point relative to a fixed reference point. The displacement sensor uses the laser ranging principle or the vibrating wire principle to achieve high-precision displacement monitoring.

[0007] Among them, the spatial stress-displacement monitoring network is a three-dimensional monitoring system composed of triaxial stress sensors and displacement sensors on multiple monitoring sections. The spatial stress-displacement monitoring network covers the stress field and displacement field of the surrounding rock within the influence range of tunnel excavation.

[0008] The coordinate transformation process is a mathematical process of converting stress component data in the engineering coordinate system to the principal stress coordinate system through a coordinate rotation transformation formula. The three coordinate axes of the principal stress coordinate system are aligned with the principal stress directions, and the shear stress component is zero in the principal stress coordinate system.

[0009] Among them, the principal stress direction angle is the rotation angle of the principal stress coordinate system relative to the engineering coordinate system, and the principal stress direction angle is obtained by solving the characteristic equation of the stress tensor; the principal stress value is the magnitude of the normal stress in the three orthogonal directions in the principal stress coordinate system after coordinate transformation, and the principal stress value is the three characteristic values ​​of the stress tensor.

[0010] Among them, the stress release rate is the ratio of the change in the principal stress value of the measuring point within a certain time period to the length of the time period and the initial principal stress value. The stress release rate reflects the rate at which the principal stress value decays over time; the displacement rate is the ratio of the cumulative displacement of the measuring point within a certain time period to the length of the time period.

[0011] Among them, the spatiotemporal characteristic matrix of surrounding rock stress release is a mathematical matrix that describes the evolution law of stress release rate process in time and space. The row vector of the spatiotemporal characteristic matrix of surrounding rock stress release is the time series of stress release rate of different monitoring sections, and the column vector of the spatiotemporal characteristic matrix of surrounding rock stress release is the spatial distribution of stress release rate of different monitoring sections at the same time.

[0012] Singular value decomposition (SVD) is a mathematical method that decomposes the spatiotemporal feature matrix of surrounding rock stress release into a product of three matrices. The main features of the spatiotemporal feature matrix of surrounding rock stress release are extracted through SVD, and the magnitude of the singular values ​​obtained by SVD reflects the importance of the corresponding modes.

[0013] Among them, the dominant stress release mode is the spatiotemporal evolution mode that contributes the most to the stress release process of the surrounding rock. The dominant stress release mode is determined by the left and right singular vectors corresponding to the maximum singular value obtained by singular value decomposition. The left singular vector represents the spatial distribution characteristics of the dominant stress release mode, and the right singular vector represents the temporal evolution characteristics of the dominant stress release mode.

[0014] The stress release rate threshold discrimination system is a discrimination standard for classifying the stress release state of surrounding rock based on the numerical range of principal stress release rate. When the principal stress release rate of the monitoring section exceeds 0.15 MPa / d, it is judged as a rapid release state; when the principal stress release rate of the monitoring section is in the range of 0.05 MPa / d to 0.15 MPa / d, it is judged as a slow release state; and when the principal stress release rate of the monitoring section is lower than 0.05 MPa / d, it is judged as a stable state.

[0015] Specifically, for monitoring sections in a rapid release state, the monitoring frequency is increased to once every 4 hours and lining construction is temporarily suspended; for monitoring sections in a slow release state, the monitoring frequency is increased to once every 12 hours and the feasibility of lining construction is assessed; and for monitoring sections in a stable state, the monitoring frequency is increased to once every 24 hours and lining construction is carried out.

[0016] Among them, the feasibility of lining construction is a comprehensive evaluation based on the stress release state of the surrounding rock to determine whether it is suitable to carry out lining construction. The feasibility assessment of lining construction includes stress release rate, displacement rate and surrounding rock deformation convergence analysis.

[0017] The crack network topology graph model is an analytical model that abstracts the crack system on the lining surface into a mathematical graph structure. The intersection of surface cracks in the constructed lining section is taken as graph nodes, the surface crack section is taken as graph edges, and the crack width is taken as edge weight.

[0018] Among them, the graph connectivity analysis algorithm is an algorithm to determine whether there is a path between any two graph nodes in the crack network topology graph model. The graph connectivity analysis algorithm traverses all graph nodes through depth-first search or breadth-first search to identify the set of interconnected graph nodes.

[0019] Among them, the crack network connectivity index is a dimensionless parameter that measures the overall connectivity of the crack network. The crack network connectivity index is calculated by dividing the number of graph nodes contained in the largest connected subgraph by the total number of graph nodes in the crack network. When the crack network connectivity index exceeds 0.6, the chemical grouting repair procedure is initiated.

[0020] This invention acquires stress component data and displacement vector data from multiple sections and measuring points by constructing a spatial stress-displacement monitoring network. It integrates the time series and spatial distribution of the stress release process into a spatiotemporal characteristic matrix of surrounding rock stress release, and uses singular value decomposition (SVD) to extract the dominant stress release modes, achieving a quantitative characterization of the spatiotemporal evolution characteristics of the stress release process. Addressing the limitations of existing technologies that rely on single-point data and fixed threshold discrimination, this invention establishes a dynamic threshold discrimination system based on the principal stress release rate. It divides the surrounding rock state into three stages: rapid release, slow release, and stable state. Corresponding monitoring frequencies and lining construction strategies are matched according to different states, ensuring that the timing of lining construction is adapted to the stress release process. Through dominant stress release mode analysis, the spatiotemporal evolution patterns that contribute most to the stress release process can be identified, accurately grasping the overall trend and local differences in stress release, avoiding the limitations of single-point monitoring. The stress release rate threshold discrimination system enables dynamic tracking of the stress release process and timely adjustment of engineering measures, avoiding the lag of fixed threshold discrimination. In summary, this invention solves the technical problem mentioned in the background art: the spatiotemporal evolution characteristics of stress release in soft rock tunnels are difficult to accurately identify, leading to inaccurate determination of the timing of lining construction. Attached Figure Description

[0021] Figure 1 This is a flowchart of the method of the present invention.

[0022] Figure 2 This is a spatiotemporal distribution map of the spatiotemporal characteristic matrix of surrounding rock stress release.

[0023] Figure 3 The graph shows the change in the principal stress release rate of section D2 over time.

[0024] Figure 4 This is a diagram showing the network topology of cracks on the lining surface.

[0025] Figure 5 This is a comparison image of the crack width at different locations before and after chemical grouting repair. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below.

[0027] like Figure 1 The diagram shows a flowchart of a method for monitoring stress release during large deformation in soft rock tunnels provided by this invention. The method includes the following steps:

[0028] S01. A monitoring section is set up every 5m to 15m along the tunnel axis behind the tunnel excavation face. Each monitoring section is equipped with triaxial stress sensors and displacement sensors at the arch crown, arch waist, sidewall and invert arch positions to form a spatial stress and displacement monitoring network.

[0029] S02. Collect stress component data and displacement vector data of each monitoring section for 180 consecutive days after excavation, perform coordinate transformation on the stress component data to obtain the principal stress direction angle and principal stress value, and calculate the stress release rate and displacement rate of each measuring point.

[0030] S03. Construct a spatiotemporal feature matrix of surrounding rock stress release. The row vector of the spatiotemporal feature matrix of surrounding rock stress release is the time series of stress release rate of different monitoring sections, and the column vector of the spatiotemporal feature matrix of surrounding rock stress release is the spatial distribution of stress release rate of different monitoring sections at the same time. Perform singular value decomposition on the spatiotemporal feature matrix of surrounding rock stress release to obtain the dominant stress release mode.

[0031] S04. Establish a stress release rate threshold discrimination system. When the principal stress release rate of the monitored section exceeds 0.15 MPa / d, it is judged as a rapid release state. When the principal stress release rate of the monitored section is in the range of 0.05 MPa / d to 0.15 MPa / d, it is judged as a slow release state. When the principal stress release rate of the monitored section is lower than 0.05 MPa / d, it is judged as a stable state.

[0032] S05. Adjust the monitoring frequency and lining construction timing based on the discrimination results of the rapid release state, the slow release state, and the stable state. For monitoring sections in the rapid release state, adopt an increased monitoring frequency of once every 4 hours and temporarily suspend lining construction. For monitoring sections in the slow release state, adopt a regular monitoring frequency of once every 12 hours and assess the feasibility of lining construction. For monitoring sections in the stable state, adopt a maintenance monitoring frequency of once every 24 hours and carry out lining construction operations.

[0033] S06. Construct a crack network topology graph model, taking the intersection of surface cracks in the lining section as graph nodes, surface crack sections as graph edges, and crack width as edge weights. Use a graph connectivity analysis algorithm to calculate the crack network connectivity index and the maximum connected subgraph size. When the crack network connectivity index exceeds 0.6, start the chemical grouting repair procedure.

[0034] Among them, the triaxial stress sensor is a pressure sensor capable of simultaneously measuring stress components in three orthogonal directions. Each triaxial stress sensor contains three mutually perpendicular measurement units, which correspond to the radial, tangential, and axial stress components of the tunnel, respectively. The displacement sensor is a sensing device that measures the three-dimensional displacement vector of a measuring point relative to a fixed reference point. The displacement sensor uses the laser ranging principle or the vibrating wire principle to achieve high-precision displacement monitoring.

[0035] The spatial stress-displacement monitoring network is a three-dimensional monitoring system composed of triaxial stress sensors and displacement sensors on multiple monitoring sections. This network covers the stress and displacement fields of the surrounding rock within the influence range of tunnel excavation. The stress component data consists of the stress component values ​​in three orthogonal directions measured by the triaxial stress sensors in an engineering coordinate system. The three axes of this engineering coordinate system are the radial, tangential, and axial directions of the tunnel. The displacement vector data consists of the displacement component values ​​of the measuring points in three orthogonal directions measured by the displacement sensors. These three orthogonal directions are consistent with the engineering coordinate system directions of the stress component data.

[0036] Coordinate transformation is the mathematical process of converting stress component data in the engineering coordinate system to the principal stress coordinate system using a coordinate rotation transformation formula. The three coordinate axes of the principal stress coordinate system are aligned with the principal stress directions, and the shear stress component is zero in this system. The principal stress direction angle is the rotation angle of the principal stress coordinate system relative to the engineering coordinate system, obtained by solving the characteristic equation of the stress tensor. The principal stress values ​​are the magnitudes of the normal stresses in the three orthogonal directions in the principal stress coordinate system after coordinate transformation; these are the three eigenvalues ​​of the stress tensor.

[0037] The stress release rate is the ratio of the change in principal stress value at a measuring point within a certain time period to the length of that time period and the initial principal stress value. The stress release rate reflects the rate at which the principal stress value decays over time. The formula for calculating the stress release rate is: the difference between the current principal stress value and the initial principal stress value, divided by the time interval, and then divided by the initial principal stress value. All terms in the formula are divided by 1 MPa for dimensionless processing. The displacement rate is the ratio of the cumulative displacement at a measuring point within a certain time period to the length of that time period. The displacement rate reflects the speed of deformation development in the surrounding rock.

[0038] The spatiotemporal characteristic matrix of surrounding rock stress release is a mathematical matrix describing the evolution of the stress release rate process in both time and space dimensions. The number of rows in the matrix equals the number of monitoring sections, and the number of columns equals the number of monitoring time points. Each element of the matrix represents the normalized stress release rate of the corresponding monitoring section at the corresponding time. The stress release rate time series is a one-dimensional array formed by arranging the stress release rates of a single monitoring section at multiple times in chronological order. The spatial distribution of stress release rates is a one-dimensional array formed by arranging the stress release rates of multiple monitoring sections at the same time according to their spatial location.

[0039] Singular value decomposition (SVD) is a mathematical method that decomposes the spatiotemporal characteristic matrix of surrounding rock stress release into a product of three matrices. SVD extracts the main features of the surrounding rock stress release spatiotemporal characteristic matrix, and the magnitude of the singular values ​​obtained reflects the importance of the corresponding modes. The dominant stress release mode is the spatiotemporal evolution pattern that contributes the most to the surrounding rock stress release process. The dominant stress release mode is determined by the left and right singular vectors corresponding to the largest singular value obtained from SVD. The left singular vector represents the spatial distribution characteristics of the dominant stress release mode, and the right singular vector represents the temporal evolution characteristics of the dominant stress release mode.

[0040] The stress release rate threshold discrimination system is a discrimination standard for classifying the stress release state of surrounding rock based on the numerical range of principal stress release rates. This system includes thresholds for rapid release, slow release, and stable states. The principal stress release rate is the rate of change of the principal stress value over time, calculated by dividing the difference between two adjacent monitored principal stress values ​​by the time interval.

[0041] The rapid release state is the stage where the surrounding rock stress is continuously released at a high rate. In this state, the deformation rate of the surrounding rock is large and does not converge, and the load on the lining structure continues to increase. The slow release state is a transitional stage where the surrounding rock stress is gradually released at a lower rate. In this state, the deformation rate decreases but is not yet fully stable. The stable state is the stage where the surrounding rock stress is basically released and the deformation tends to converge. In this stable state, the load on the lining structure tends to be constant.

[0042] The encrypted monitoring frequency is a high-frequency data acquisition frequency used for monitoring sections in the rapid release state, and this frequency is used to capture the rapid changes in the stress release process. The regular monitoring frequency is a medium-frequency data acquisition frequency used for monitoring sections in the slow release state, and this frequency is used to track the gradual trend of the stress release process. The maintenance monitoring frequency is a low-frequency data acquisition frequency used for monitoring sections in the stable state, and this frequency is used to verify the stability of the stress release process.

[0043] The feasibility assessment of lining construction is a comprehensive evaluation based on the stress release state of the surrounding rock to determine whether lining construction is suitable. This assessment includes stress release rate, displacement rate, and surrounding rock deformation convergence analysis. The lining construction operation involves pouring secondary lining concrete after the surrounding rock stress has basically stabilized. This lining construction operation is carried out in the tunnel section corresponding to the stable state monitoring section.

[0044] The crack network topology model is an analytical model that abstracts the crack system on the lining surface into a mathematical graph structure. In this model, nodes represent crack intersections or endpoints, edges represent crack segments connecting two nodes, and edge weights represent the geometric characteristic parameters of the cracks. A surface crack intersection is the point where two or more cracks on the lining surface intersect. A surface crack segment is the portion of the crack connecting two adjacent surface crack intersections. The crack width is the vertical distance between the two edges of the crack, measured using a crack width measuring instrument or image processing methods.

[0045] Graph connectivity analysis algorithms are used to determine whether a path exists between any two graph nodes in a crack network topology model. These algorithms traverse all graph nodes using either depth-first search or breadth-first search to identify sets of interconnected nodes. The crack network connectivity index is a dimensionless parameter that measures the overall connectivity of the crack network. It is calculated by dividing the number of nodes in the largest connected subgraph by the total number of nodes in the crack network. A higher connectivity index indicates stronger connectivity within the crack network.

[0046] The maximum connected subgraph size is the number of graph nodes contained in the connected subgraph with the most graph nodes in the crack network topology model. A path exists between any two graph nodes in the maximum connected subgraph, and this maximum connected subgraph represents the most important through-structure in the crack system. The chemical grouting repair procedure is a process of filling and reinforcing cracks using chemical materials. This procedure involves injecting epoxy resin or polyurethane chemical grout into the cracks, allowing the grout to solidify and restore the integrity and impermeability of the lining structure.

[0047] The specific implementation methods of the above steps are described in detail below.

[0048] The specific implementation of step S01 is as follows: First, determine the spatial coordinates of the tunnel excavation face. Starting from the excavation face, select a monitoring section position every 5m to 15m along the opposite direction of the tunnel axis. The interval distance is determined according to the severity of the surrounding rock deformation. A 5m interval is used in areas of severe deformation, and a 15m interval is used in areas of mild deformation. At the arch crown, left arch waist, right arch waist, left wall, right wall, and invert arch positions of each monitoring section, drill holes and bury triaxial stress sensors and displacement sensors. The drilling depth is 50cm to 80cm. The sensors are fixed inside the surrounding rock by grouting. The axes of the three measuring units of the triaxial stress sensor are aligned with the radial, tangential, and axial directions of the tunnel, respectively. The fixed reference point of the displacement sensor is set in stable rock mass far away from the excavation influence range. All sensors at the monitoring sections are connected to the monitoring host through data acquisition cables to form a spatial stress and displacement monitoring network. The spatial stress and displacement monitoring network covers the entire process of surrounding rock stress release within 180 days after excavation.

[0049] The specific implementation of step S02 is as follows: the data acquisition system automatically reads the output values ​​of the triaxial stress sensor and displacement sensor on each monitoring section according to a preset time interval. The time interval is set to 4 hours for the first 30 days after excavation, 12 hours for the 31st to 90th day, and 24 hours for the 91st to 180th day. The collected stress component data is classified and stored according to the monitoring section number and the location of the measuring point. A stress tensor is constructed for the three stress component values ​​of each measuring point. By solving the characteristic equation of the stress tensor, three eigenvalues ​​and corresponding eigenvectors are obtained. The three eigenvalues ​​are the principal stress values, the direction of the eigenvector is the principal stress direction, and the angle between the principal stress direction and the axis of the engineering coordinate system is the principal stress direction angle. The difference between the principal stress value at the current moment and the principal stress value on the first day after excavation is calculated. The difference is divided by the time interval and then by the initial principal stress value to obtain the stress release rate. The three components of the displacement vector data are vector synthesized to obtain the total displacement. The total displacement is divided by the time interval to obtain the displacement rate.

[0050] The specific implementation of step S03 involves normalizing the stress release rate data of all monitoring sections at each monitoring time. The normalization method involves dividing each stress release rate value by the maximum value of all stress release rate values, constructing a spatial-temporal feature matrix of surrounding rock stress release with the number of rows equal to the number of monitoring sections and the number of columns equal to the number of monitoring time points. The element in the i-th row and j-th column of the spatial-temporal feature matrix represents the normalized stress release rate value of the i-th monitoring section at the j-th time. Singular value decomposition (SVD) is then performed on the spatial-temporal feature matrix of surrounding rock stress release. The spatiotemporal characteristic matrix of surrounding rock stress release is decomposed into the product of a left singular matrix, a singular value diagonal matrix, and a right singular matrix. The singular value with the largest value in the singular value diagonal matrix and its corresponding left and right singular vectors are extracted. The values ​​of each element of the left singular vector reflect the contribution weight of different monitoring sections to the dominant stress release mode, and the values ​​of each element of the right singular vector reflect the evolution intensity of the dominant stress release mode at different times. By analyzing the left singular vector, the spatial location of the most intense stress release is identified, and by analyzing the right singular vector, the time period of the most active stress release is identified.

[0051] The specific implementation of step S04 involves calculating the difference in principal stress values ​​between two adjacent monitoring sessions for each monitoring section. Dividing this difference by the time interval between the two monitoring sessions yields the principal stress release rate. A first threshold of 0.15 MPa / d and a second threshold of 0.05 MPa / d are set. When the principal stress release rate is greater than 0.15 MPa / d, the monitoring section is marked as being in a rapid release state. When the principal stress release rate is between 0.05 MPa / d and 0.15 MPa / d, the monitoring section is marked as being in a slow release state. When the principal stress release rate is less than 0.05 MPa / d, the monitoring section is marked as being in a stable state. The first and second thresholds are determined statistically based on historical data of surrounding rock stress release measured on-site. This stress release rate threshold discrimination system discretizes the continuously changing stress release process into three state levels, providing a quantitative basis for subsequent monitoring frequency adjustments and construction decisions.

[0052] The specific implementation of step S05 is as follows: based on the stress release state markings of each monitoring section obtained in step S04, the data acquisition interval is adjusted to 4 hours for monitoring sections marked as rapid release state, to 12 hours for monitoring sections marked as slow release state, and to 24 hours for monitoring sections marked as stable state. Simultaneously, the lining construction conditions of the corresponding tunnel section are determined. For tunnel sections in the rapid release state, a temporary suspension of lining construction is issued, and monitoring continues until the stress release rate decreases to below 0.15 MPa / d. For tunnel sections in the slow release state, a feasibility assessment program for lining construction is initiated. This assessment program comprehensively analyzes three indicators: stress release rate, displacement rate, and deformation convergence. When all three indicators meet the design requirements, lining construction is permitted. For tunnel sections in the stable state, a direct instruction to commence lining construction is issued. This graded monitoring and dynamic construction strategy achieves refined control of the surrounding rock stress release process.

[0053] The specific implementation of step S06 involves using a crack width measuring instrument or high-resolution image processing method to conduct a crack survey on the surface of the lining section, recording the starting coordinates, ending coordinates, and crack width of each crack, identifying the intersection points of all surface cracks and numbering them as graph nodes, defining the surface crack segments between two adjacent graph nodes as graph edges, assigning the crack width corresponding to the graph edge as the edge weight, constructing a crack network topology graph model based on the graph nodes and graph edges, using a depth-first search algorithm to traverse all reachable graph nodes starting from any graph node, grouping all interconnected graph nodes into a connected subgraph, counting the number of graph nodes in all connected subgraphs, and identifying the connected subgraph with the most graph nodes as the maximum connected subgraph, calculating the ratio of the number of graph nodes in the maximum connected subgraph to the total number of graph nodes in the crack network to obtain the crack network connectivity index, and determining that the crack system has formed a through leakage channel when the crack network connectivity index exceeds 0.6, immediately initiating a chemical grouting repair procedure to fill and seal the crack.

[0054] It should be noted that the first key technical idea of ​​this invention is to establish a spatiotemporal decoupling method for stress release based on singular value decomposition. By constructing the monitoring data into a spatiotemporal feature matrix and performing singular value decomposition, the complex spatiotemporal coupled stress release process is decomposed into the spatial distribution characteristics and temporal evolution characteristics of the dominant modes. Compared with the traditional single-point time history curve analysis method, the spatiotemporal decoupling method can identify the main control law and key areas of stress release from a global perspective, avoiding the randomness and limitations of single-point data. It significantly improves the depth of understanding and prediction accuracy of the stress release process of large deformation in soft rock, and provides a scientific mathematical analysis tool for optimizing monitoring schemes and selecting construction timing.

[0055] The second key technical idea of ​​this invention is to construct a dynamic monitoring and construction decision-making system based on stress release rate classification. By setting rapid release state thresholds and slow release state thresholds, the continuous stress release process is discretized into three state levels. Different monitoring frequencies and construction strategies are adopted for different state levels. Compared with the traditional method of fixed monitoring frequency and experience-based judgment of construction timing, the dynamic decision-making system realizes the optimized allocation of monitoring resources and quantitative control of construction risks. It not only ensures the timely capture of rapidly changing processes, but also avoids the waste of resources in the stable stage. At the same time, the quantitative thresholds eliminate the subjective arbitrariness of construction decisions.

[0056] The third key technical idea of ​​this invention is to introduce a graph theory-based quantitative assessment method for crack network connectivity. This method abstracts the crack system on the lining surface into a topological graph model and uses a graph connectivity analysis algorithm to calculate the connectivity index. Compared with the traditional single-index evaluation method of crack width and depth, the topological analysis method can assess the overall hazard and leakage risk of the crack system from the perspective of network connectivity. It reveals the key impact of the interconnection between cracks on structural safety, provides a quantitative scientific basis for the timing and scope of chemical grouting repair, and avoids resource waste caused by over-repair and safety hazards caused by under-repair.

[0057] The synergistic effect of the three key technical approaches mentioned above lies in constructing a full-chain quantitative analysis system from stress release process monitoring to lining structure damage assessment. Spatiotemporal decoupling analysis reveals the deep-seated laws of stress release, dynamic hierarchical monitoring captures the real-time status of stress release, and topological connectivity assessment quantifies the structural damage caused by stress release. These three aspects support each other to form a closed-loop management system for monitoring, early warning, and engineering response. Compared with traditional experience-driven and single-indicator discrimination methods, the synergistic system realizes scientific monitoring and precise control of the entire process of stress release during large deformation in soft rock tunnels, significantly reducing the risk of lining structure cracking and construction safety hazards.

[0058] It should be noted that this invention also solves the following technical problem: In existing technologies, the assessment of the development status of lining cracks relies on manual inspection and measurement of individual crack parameters, making it difficult to systematically evaluate the overall connectivity and structural hazard of the crack network, resulting in a lack of quantitative basis for crack repair decisions. This invention constructs a crack network topology graph model, abstracting the lining surface crack system into a mathematical graph structure composed of graph nodes, edges, and edge weights. It uses a graph connectivity analysis algorithm to calculate the crack network connectivity index and the size of the largest connected subgraph, quantifying the connectivity of the crack system as a connectivity index value. When the crack network connectivity index exceeds 0.6, it indicates that the crack has formed a continuous structure, at which point a chemical grouting repair procedure is initiated for timely treatment. This crack network analysis based on graph theory avoids the subjectivity and bias of manual inspection, achieving a systematic assessment of the crack development status and a scientific determination of the repair timing.

[0059] Specifically, the principle of this invention is as follows: The invention solves the aforementioned technical problems by transforming dispersed monitoring data into characteristic parameters that reflect the essential laws of stress release through mathematical transformation. First, by deploying triaxial stress sensors and displacement sensors at key locations across multiple monitoring sections, a monitoring network covering the tunnel space is established. This spatial layout ensures that the monitoring data comprehensively reflects the distribution of the surrounding rock stress and displacement fields. Second, the collected stress component data undergoes coordinate transformation to obtain principal stress values, calculates the stress release rate, and constructs a spatiotemporal characteristic matrix of surrounding rock stress release. The row and column vectors of this matrix describe the temporal and spatial dimensions of the stress release process, respectively. Singular value decomposition can extract the dominant evolutionary pattern from the complex spatiotemporal data. This pattern reflects the main characteristics and trends of the stress release process. Finally, a dynamic threshold discrimination system is established based on the principal stress release rate. The surrounding rock state is divided into different stages according to the magnitude of the stress release rate. Differentiated monitoring frequencies and construction strategies are adopted for different stages to ensure that the monitoring plan and construction arrangements match the stress release process of the surrounding rock. This systematic approach, from data acquisition to feature extraction and state determination, enables accurate identification of the spatiotemporal evolution characteristics of stress release and scientific determination of the timing of lining construction.

[0060] The following provides a specific embodiment 1 of the present invention, and the specific implementation of each step in this embodiment 1 is described in detail below.

[0061] The specific implementation methods of steps S01 and S05 are the same as those described above, and will not be repeated in detail here.

[0062] The specific implementation of step S02 involves automatically reading the output values ​​of the triaxial stress sensor and displacement sensor on each monitoring section through a data acquisition system at preset time intervals. The time interval is set to 4 hours for the first 30 days after excavation, 12 hours for days 31 to 90, and 24 hours for days 91 to 180. The collected stress component data is categorized and stored according to the monitoring section number and measuring point location. A stress tensor is constructed for the three stress component values ​​at each measuring point, and coordinate transformation is performed. The coordinate transformation uses the stress tensor eigenvalue solving method to obtain the principal stress values ​​and principal stress direction angles. The formula for calculating the stress release rate is as follows:

[0063] ;

[0064] In the formula, For stress relief rate, This represents the principal stress value at the current moment, in MPa. The initial principal stress values ​​are expressed in MPa. The time interval is expressed in days (d). The stress reference value is taken as 1 MPa. The time reference value is taken as 1 day. The formula for calculating the displacement rate is expressed as follows:

[0065] ;

[0066] In the formula, For displacement rate, The displacement component is in the x-direction, and the unit is mm. This represents the displacement component in the y-direction, in mm. The displacement component is in the z-direction, and the unit is mm. The displacement reference value is taken as 1mm. The time reference value is set to 1 day. Among them, , , These are the three orthogonal displacement components measured by the displacement sensor, which are directly acquired through the displacement sensor.

[0067] The specific implementation of step S03 involves normalizing the stress release rate data of all monitoring sections at each monitoring time. The normalization method involves dividing each stress release rate value by the maximum value of all stress release rate values ​​to construct a spatiotemporal characteristic matrix of surrounding rock stress release. The expression for the spatiotemporal characteristic matrix of surrounding rock stress release is as follows:

[0068] ;

[0069] In the formula, This is the spatiotemporal characteristic matrix of surrounding rock stress release. Let be the normalized stress release rate value of the i-th monitoring section at time j. To monitor the total number of cross sections, To monitor the total number of time points, The value range is 1 to , The value range is 1 to .in, The calculation formula is expressed as follows:

[0070] ;

[0071] In the formula, Let be the stress release rate of the i-th monitoring section at time j. This represents the maximum stress release rate at all times across all monitored sections. Singular value decomposition (SVD) is performed on the spatiotemporal characteristic matrix of surrounding rock stress release, decomposing it into:

[0072] ;

[0073] In the formula, It is a left singular matrix, and its column vectors are left singular vectors. It is a singular-valued diagonal matrix, where its diagonal elements are singular values. Let be a right singular matrix, and its column vectors be right singular vectors. This is the transpose of the right singular matrix. Extract the singular value with the largest value from the singular value diagonal matrix. and its corresponding left singular vector and right singular vectors The left singular vector The kth element The right singular vector reflects the contribution weight of the k-th monitoring section to the dominant stress release mode. The qth element The intensity reflects the evolution of the dominant stress release mode at time q, where The value range is 1 to , The value range is 1 to .

[0074] The specific implementation of step S04 involves calculating the difference in principal stress values ​​between two adjacent monitoring sessions for each monitoring section. The formula for calculating the principal stress release rate is as follows:

[0075] ;

[0076] In the formula, The principal stress release rate, The value is the principal stress from the last monitoring, in MPa. The values ​​are the principal stresses from the previous monitoring, in MPa. The time interval between two monitoring sessions is expressed in days (d). The stress reference value is taken as 1 MPa. Set the time reference value to 1 day. Set the first threshold. 0.15 MPa / d and the second threshold It is 0.05 MPa / d, when The monitored section is marked as being in a rapid release state when... The monitoring section is marked as being in a slow release state. The monitored section is then marked as being in a stable state.

[0077] The specific implementation of step S06 involves using a crack width measuring instrument or high-resolution image processing method to conduct a crack survey on the surface of the lining section, recording the start coordinates, end coordinates, and crack width of each crack, identifying all surface crack intersections and numbering them as graph nodes, defining the surface crack segment between two adjacent graph nodes as a graph edge, assigning the crack width corresponding to the graph edge as the edge weight, constructing a crack network topology graph model based on the graph nodes and graph edges, and using a depth-first search algorithm to traverse all reachable graph nodes starting from any graph node, grouping all interconnected graph nodes into a connected subgraph, and counting the number of graph nodes in all connected subgraphs. The formula for calculating the crack network connectivity index is as follows:

[0078] ;

[0079] In the formula, This represents the connectivity index of the crack network. This represents the number of nodes in the largest connected subgraph. This represents the total number of nodes in the crack network graph. When... If it is determined that the crack system has formed a through-seepage channel, a chemical grouting repair procedure should be immediately initiated to fill and seal the crack.

[0080] The stress component data output by the triaxial stress sensor includes radial stress components. tangential stress components and axial stress components The characteristic equation of the stress tensor formed by the three stress components in the principal stress coordinate system is as follows:

[0081] ;

[0082] In the formula, Let be the stress tensor matrix. These are the eigenvalues, i.e., the principal stress values. It is the identity matrix. This represents determinant operations. Where, The stress tensor matrix is ​​3×3, and its specific expression is:

[0083] ;

[0084] In the formula, This represents the radial tangential shear stress component, in MPa. These are the radial and axial shear stress components, in MPa. The shear stress component is tangential and axial, expressed in MPa. The stress tensor matrix is ​​symmetric and obtained through coupled measurements using a triaxial stress sensor. The empirical value of the shear stress component is 0.1 MPa to 0.5 MPa. Solving the characteristic equation yields the three principal stress values. , and The principal stress values ​​satisfy Select the maximum principal stress value The principal stress values ​​are used in subsequent stress release rate calculations. Principal stress direction angles. The principal stress direction angle is calculated using the cosine of the angle between the eigenvector and the axis of the engineering coordinate system. The formula for calculating the principal stress direction angle is as follows:

[0085] ;

[0086] In the formula, The principal stress direction angle, in degrees. This is the eigenvector corresponding to the maximum principal stress. Let be the radial unit vector in the engineering coordinate system. This represents the vector dot product operation. This represents the L2 norm operation of vectors. and The calculation formula is expressed as follows:

[0087] ;

[0088] In the formula, For any three-dimensional vector, , , For vectors The three components, for a unit vector and , and .

[0089] The implementation process of the depth-first search algorithm is as follows: Initialize all graph nodes to be unvisited; select any unvisited graph node as the starting node; mark the starting node as visited and add it to the current connected subgraph; start from the starting node and visit adjacent unvisited graph nodes along the graph edges; mark the visited graph nodes as visited and add them to the current connected subgraph; recursively visit the adjacent unvisited graph nodes of the newly added graph node until no new graph node can be visited, at which point the construction of the current connected subgraph ends; select the next unvisited graph node as the new starting node and repeat the above process until all graph nodes have been visited; count the number of graph nodes contained in each connected subgraph and determine the largest connected subgraph.

[0090] It should be noted that the variables involved in this embodiment are explained in detail in Table 1.

[0091] Table 1. Variable Explanation Table

[0092]

[0093] To better understand and implement this invention, the following is an example 2 of a specific application scenario: A technical team undertook a monitoring task for a soft rock tunnel project. The access tunnel to the plant is approximately 2300m long, with a shallow burial depth at the tunnel entrance. The thickness of the overlying rock layer gradually increases from 7m at the entrance to 75m at chainage JT0+129. The surrounding rock is mainly phyllite slate, characterized by a loose structure, poor cementation, severe wrinkling, and disordered bedding. It is a typical loose soft rock with a dry compressive strength of less than 15.0MPa and extremely low bearing capacity. After tunnel excavation, the peak seepage rate of the surrounding rock reached 255ml / s. The softening effect of the water flow and the seepage pressure further weakened the stability of the rock mass. Tunnel excavation began in September 2024, and excavation was halted on October 20, 2024, to prepare for secondary lining construction. Secondary lining construction began in January 2025. During the tunnel excavation phase, the area exhibited significant uncontrolled deformation characteristics, such as a maximum daily settlement of 96 mm at the tunnel entrance slope, slope cracks, continuous divergence of convergent deformation within the tunnel, and landslides. To date, the secondary lining of the tunnel has been completed from JT0+0.0 to JT0+129.5. Cracks of varying degrees have appeared in the completed sections, particularly at the tunnel entrance, and these cracks include surface cracks, deep cracks, and through cracks. The technical team decided to use the large deformation stress release monitoring method for soft rock tunnels of this invention to systematically monitor and address the issue in this tunnel.

[0094] The technical team deployed monitoring sections every 10 meters along the tunnel axis behind the tunnel excavation face, totaling 14 monitoring sections from chainage JT0+0.0 to JT0+130.0, numbered D1 to D14. At each monitoring section, a set of triaxial stress sensors and one displacement sensor were installed at six locations: the arch crown, left arch waist, right arch waist, left side wall, right side wall, and invert arch, forming a spatial stress-displacement monitoring network. The triaxial stress sensors were vibrating wire triaxial stress gauges, each containing three mutually perpendicular measuring units corresponding to the radial, tangential, and axial stress components of the tunnel, with a measurement range of 0 to 10 MPa and an accuracy of 0.01 MPa. The displacement sensors were laser rangefinder displacement gauges, with a measurement range of 0 to 500 mm and an accuracy of 0.1 mm. The technical team began collecting stress component data and displacement vector data for each monitoring section on January 15, 2025, collecting data every 4 hours for 180 days until July 13, 2025.

[0095] The technical team performed coordinate transformation on the collected stress component data, converting the stress component data in the engineering coordinate system to the principal stress coordinate system using a coordinate rotation transformation formula to obtain the principal stress direction angle and principal stress value. Taking the crown measuring point of monitoring section D2 as an example, at the initial moment, the radial stress component of this measuring point was 2.85 MPa, the tangential stress component was 1.92 MPa, and the axial stress component was 1.45 MPa. After coordinate transformation, the maximum principal stress value was obtained as 3.12 MPa, the intermediate principal stress value as 1.86 MPa, and the minimum principal stress value as 1.24 MPa. The maximum principal stress direction angle relative to the radial direction was 18.5 degrees. The technical team calculated the stress release rate and displacement rate for each measuring point. The stress release rate is the ratio of the change in the principal stress value of the measuring point within a certain time period to the length of the time period and the initial principal stress value. The displacement rate is the ratio of the cumulative displacement of the measuring point within a certain time period to the length of the time period. The average stress release rate of the monitoring point at the D2 arch crown in the first 30 days was 0.0086 per day, and the average displacement rate was 3.2 mm / d, indicating that the monitoring point was in a rapid release state.

[0096] The technical team constructed a spatiotemporal characteristic matrix of surrounding rock stress release. This matrix has 14 rows corresponding to 14 monitoring sections and 1081 columns corresponding to the number of monitoring time points every 4 hours over 180 days. The matrix elements are the normalized values ​​of the stress release rate of the corresponding monitoring section at the corresponding time. For example... Figure 2As shown, the spatiotemporal characteristic matrix of surrounding rock stress release clearly reveals the variation of stress release rate with time and space at different monitoring sections. The stress release rate at monitoring sections D1 to D6 in the tunnel entrance section is significantly higher than that at monitoring sections D10 to D14 in the deep-buried section. The technical team performed singular value decomposition on the spatiotemporal characteristic matrix of surrounding rock stress release, obtaining 14 singular values. The largest singular value is 58.7, the second largest is 12.3, and the third largest is 6.8. The largest singular value accounts for 76.2% of the total sum of squares of singular values, indicating that the dominant stress release mode contributes the most to the stress release process. The left singular vector of the dominant stress release mode shows that the weight coefficient of the monitoring section in the tunnel entrance section is between 0.28 and 0.32, while the weight coefficient of the monitoring section in the deep-buried section is between 0.08 and 0.12, indicating that the stress release process mainly occurs in the tunnel entrance section. The right singular vector of the dominant stress release mode shows that the weighting coefficients for the first 60 days are between 0.015 and 0.018, and the weighting coefficients for the last 120 days are between 0.004 and 0.007, indicating that the stress release rate is higher in the early stage and gradually decreases in the later stage.

[0097] The technical team established a stress release rate threshold discrimination system. A rapid release state is defined as the principal stress release rate exceeding 0.15 MPa / d at the monitored section; a slow release state is defined as the principal stress release rate is between 0.05 MPa / d and 0.15 MPa / d; and a stable state is defined as the principal stress release rate is below 0.05 MPa / d. The principal stress release rate is calculated by dividing the difference between two adjacent monitoring principal stress values ​​by the time interval. The technical team compiled the state discrimination results for each monitoring section at different times, as shown in Table 2.

[0098] Table 2. Statistical table of state discrimination results of each monitoring section at different times.

[0099]

[0100] As shown in Table 1, monitoring sections D1 to D4 at the tunnel entrance were in a rapid stress release state for the first 30 days, while the stress release rate of monitoring sections D6 to D8 in the deep-buried section was lower, entering a slow release or stable state within the first 30 days. Based on the state assessment results, the technical team adjusted the monitoring frequency and lining construction timing. For monitoring sections in a rapid stress release state, a more frequent monitoring frequency of once every 4 hours was adopted, and lining construction was temporarily suspended. For monitoring sections in a slow stress release state, a regular monitoring frequency of once every 12 hours was adopted, and the feasibility of lining construction was assessed. For monitoring sections in a stable state, a maintenance monitoring frequency of once every 24 hours was adopted, and lining construction was carried out. Figure 3As shown, the principal stress release rate of the monitoring section D2 averaged 0.18 MPa / d in the first 30 days, decreased to 0.09 MPa / d from day 31 to 90, and further decreased to 0.03 MPa / d from day 91 to 180. Accordingly, the technical team adopted a higher monitoring frequency and suspended lining construction from day 1 to 30, adopted a regular monitoring frequency and assessed the feasibility of lining construction from day 31 to 90, and after confirming that the section had entered a stable state on day 91, maintained the monitoring frequency and carried out lining construction.

[0101] For the lining sections already constructed, the technical team conducted crack inspections from March 1, 2025 to June 30, 2025, discovering 73 cracks on the lining surface within the range of chainage JT0+0.0 to JT0+100.0. Of these, 52 cracks (71.2%) were found in the opening section from JT0+0.0 to JT0+60.0. The team measured the location, length, and width of each crack, with widths ranging from 0.2 mm to 2.8 mm and an average width of 1.1 mm. The team constructed a crack network topology model, using the intersections of cracks on the lining surface as nodes, crack segments as edges, and crack widths as edge weights. The crack network topology model contained 96 nodes and 73 edges. The team used a depth-first search algorithm to analyze the connectivity of the crack network, identifying a maximum connected subgraph containing 62 nodes. The technical team calculated the crack network connectivity index to be 62 divided by 96, which equals 0.646. This exceeds the threshold of 0.6, indicating that the crack network has formed a continuous structure. The technical team immediately initiated the chemical grouting repair procedure. Figure 4 As shown, the crack network topology model intuitively displays the spatial distribution and connectivity of cracks. The most connected subgraph is mainly distributed in the range from JT0+0.0 to JT0+60.0 at the entrance of the tunnel. The cracks in this area are interconnected to form a complex network structure.

[0102] The technical team used epoxy resin chemical grout to fill and reinforce the cracks. First, they used a wire brush to clean the surface of the cracks of loose material and debris. Then, they used a blower to remove dust from inside the cracks. Sealing strips were then applied to both sides of the cracks to create a sealed space. A grouting hole was installed every 0.5m along the crack's direction, with a diameter of 12mm and a depth of 2 / 3 of the lining thickness. The team prepared an epoxy resin chemical grout with a viscosity of 280mPa·s and a compressive strength of 85MPa. A low-pressure, slow-speed grouting process was employed, controlling the grouting pressure between 0.3MPa and 0.5MPa and the grouting speed between 2L / min and 5L / min. Grouting began from the lowest grouting hole. When grout overflowed from an adjacent hole, grouting of that hole was stopped, and the team moved to the next hole to continue grouting until the highest grouting hole was completed. After the chemical grout solidified in the cracks, it restored the integrity and impermeability of the lining structure. The technical team conducted a quality inspection 14 days after the chemical grouting repair. Impact testing revealed no voids, and crack width measurements showed that all cracks were completely filled and the seepage was entirely eliminated. Figure 5 As shown, the comparison of crack widths before and after chemical grouting repair shows that the repair effect is significant, with the width of the repaired cracks all being less than 0.1 mm.

[0103] The technological advancements brought about by this invention compared to traditional methods are mainly reflected in the following aspects. Traditional monitoring methods employ a single-point discrete layout, which can only acquire stress and displacement data at local locations and cannot reflect the spatial distribution differences and temporal evolution patterns of stress release in the surrounding rock. In contrast, this invention forms a spatial monitoring network by deploying triaxial stress sensors and displacement sensors at key locations across multiple monitoring sections. This acquires stress and displacement field data covering the tunnel space, constructs a spatiotemporal characteristic matrix of surrounding rock stress release, and extracts the dominant stress release modes through singular value decomposition. This achieves a quantitative characterization of the spatiotemporal evolution characteristics of the stress release process and can accurately identify the overall trend and local differences in stress release. Traditional methods for judging rock stability use fixed stress or displacement thresholds, neglecting the dynamic changes in stress release rate. This can easily lead to inaccurate timing of lining construction. This invention establishes a dynamic threshold discrimination system based on the principal stress release rate, dividing the rock state into three stages: rapid release, slow release, and stable. Corresponding monitoring frequencies and lining construction strategies are matched to different states, ensuring that the timing of lining construction is adapted to the stress release process of the surrounding rock. This avoids situations where premature lining construction leads to excessive structural loads or delayed construction results in schedule delays. Traditional crack assessment methods rely on manual inspection and individual crack parameter measurements, making it difficult to systematically evaluate the overall connectivity of the crack network. This invention, however, constructs a crack network topology model and uses graph connectivity analysis algorithms to calculate the crack network connectivity index, quantifying the connectivity of the crack system as a connectivity index value. This enables a systematic assessment of crack development and a scientific determination of repair timing.

[0104] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for monitoring stress release during large deformation in soft rock tunnels, characterized in that, A monitoring section is set up every 5m to 15m along the tunnel axis behind the tunnel excavation face. Each monitoring section has triaxial stress sensors and displacement sensors installed at the arch crown, arch waist, sidewalls, and invert arch positions to form a spatial stress and displacement monitoring network. Stress component data and displacement vector data of each monitoring section are collected for 180 consecutive days after excavation. The stress component data are processed by coordinate transformation to obtain the principal stress direction angle and principal stress value. The stress release rate and displacement rate of each measuring point are calculated. A spatiotemporal characteristic matrix of surrounding rock stress release is constructed. Singular value decomposition is performed on the spatiotemporal characteristic matrix of surrounding rock stress release to obtain the dominant stress release mode. Establish a threshold discrimination system for stress release rate; adjust the monitoring frequency and lining construction timing based on the discrimination results; construct a crack network topology graph model, and use graph connectivity analysis algorithm to calculate the crack network connectivity index and the size of the maximum connected subgraph.

2. The method according to claim 1, characterized in that, The triaxial stress sensor is a pressure sensor that simultaneously measures stress components in three orthogonal directions. Each triaxial stress sensor contains three mutually perpendicular measurement units, which correspond to the radial, tangential, and axial stress components of the tunnel, respectively.

3. The method according to claim 2, characterized in that, A displacement sensor is a sensing device that measures the three-dimensional displacement vector of a measuring point relative to a fixed reference point. Displacement sensors use laser ranging principle or vibrating wire principle to achieve high-precision displacement monitoring.

4. The method according to claim 3, characterized in that, The spatial stress-displacement monitoring network is a three-dimensional monitoring system composed of triaxial stress sensors and displacement sensors on multiple monitoring sections. The spatial stress-displacement monitoring network covers the stress field and displacement field of the surrounding rock within the influence range of tunnel excavation.

5. The method according to claim 4, characterized in that, Coordinate transformation is a mathematical process that converts stress component data in the engineering coordinate system to the principal stress coordinate system using coordinate rotation transformation formulas. The three coordinate axes of the principal stress coordinate system are aligned with the principal stress directions, and the shear stress component is zero in the principal stress coordinate system.

6. The method according to claim 5, characterized in that, The principal stress direction angle is the rotation angle of the principal stress coordinate system relative to the engineering coordinate system. The principal stress direction angle is obtained by solving the characteristic equation of the stress tensor. The principal stress value is the magnitude of the normal stress in the three orthogonal directions in the principal stress coordinate system after coordinate transformation. The principal stress value is the three characteristic values ​​of the stress tensor.

7. The method according to claim 6, characterized in that, The stress release rate is the ratio of the change in principal stress value at a measuring point within a certain time period to the length of the time period and the initial principal stress value. The stress release rate reflects the rate at which the principal stress value decays over time. The displacement rate is the ratio of the cumulative displacement at a measuring point within a certain time period to the length of the time period.

8. The method according to claim 7, characterized in that, The spatiotemporal characteristic matrix of surrounding rock stress release is a mathematical matrix that describes the evolution of the stress release rate process in the time and space dimensions. The row vectors of the spatiotemporal characteristic matrix of surrounding rock stress release are the time series of stress release rates at different monitoring sections, and the column vectors of the spatiotemporal characteristic matrix of surrounding rock stress release are the spatial distribution of stress release rates at different monitoring sections at the same time.

9. The method according to claim 8, characterized in that, Singular value decomposition (SVD) is a mathematical method that decomposes the spatiotemporal characteristic matrix of surrounding rock stress release into a product of three matrices. The main features of the spatiotemporal characteristic matrix of surrounding rock stress release are extracted through SVD, and the magnitude of the singular values ​​obtained by SVD reflects the importance of the corresponding modes.

10. The method according to claim 9, characterized in that, The dominant stress release mode is the spatiotemporal evolution mode that contributes the most to the stress release process of the surrounding rock. The dominant stress release mode is determined by the left and right singular vectors corresponding to the maximum singular value obtained by singular value decomposition. The left singular vector represents the spatial distribution characteristics of the dominant stress release mode, and the right singular vector represents the temporal evolution characteristics of the dominant stress release mode.

Images