Cross-activity fracture tunnel full life cycle service state dynamic analysis method and system

Abstract

The invention discloses a cross-activity fracture tunnel full life cycle service state dynamic analysis method and system, relates to the technical field of numerical simulation, and solves the technical problems that the dynamic change of data in a tunnel full life cycle cannot be reflected and the service state analysis is inaccurate in the prior art. The method comprises the following steps: in a design period, establishing a tunnel-surrounding rock-fault three-dimensional refined model considering an internal structure of a fault and creep / stick-slip fault deformation characteristics, inputting collected initial parameters into a model database for numerical simulation, and carrying out preliminary design of seismic resistance and error resistance of the structure on the basis of a numerical simulation result; during the construction period, updating the database, inputting the model for re-simulation, changing the anti-seismic and anti-error design of the structure based on a new numerical simulation result, and preliminarily evaluating the service state of the tunnel; and in an operation period, predicting a dynamic data development trend by using the prediction model, inputting the dynamic data development trend into the numerical model for numerical simulation, and correcting a tunnel service state evaluation and emergency processing scheme according to a numerical simulation result.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of numerical simulation, in particular to a dynamic analysis method and system for the service state of a tunnel across an active fault in the whole life cycle. BACKGROUND

[0002] At present, there is a lack of fine, intelligent and dynamic analysis method considering the whole life cycle for the dynamic analysis of the service state of a tunnel across a fault. The traditional method is based on the geological data in the survey stage and the preliminary design data of the tunnel lining. The data provided by the above-mentioned data are static and inaccurate, and the dynamic updating of the data is not considered. On the other hand, the tunnel-stratum-fault numerical calculation model only considers the fault dislocation in general, and cannot truly simulate the deformation characteristics of creep and stick-slip. The above limitations further lead to a large difference between the analysis result of the service state of the tunnel across an active fault and the real service state of the tunnel, and cannot reflect the dynamic change of the data in the whole life cycle of the tunnel, and the service state analysis is not accurate. SUMMARY

[0003] In order to solve the above-mentioned problems in the prior art, the present application provides a dynamic analysis method and system for the service state of a tunnel across an active fault in the whole life cycle, which solves the technical problem that the prior art cannot reflect the dynamic change of the data in the whole life cycle of the tunnel, and the service state analysis is not accurate.

[0004] A dynamic analysis method for the service state of a tunnel across an active fault in the whole life cycle, comprising: dividing the whole life cycle into a design period, a construction period and an operation period;

[0005] During the design period, data are collected and an input database is constructed, a three-dimensional fine model of a tunnel-surrounding rock-fault of a creep fault or a three-dimensional fine model of a tunnel-surrounding rock-fault of a stick-slip fault considering the internal structure of the fault is established, and fine numerical simulation is carried out based on the input database, and finally preliminary design of structure seismic resistance and fault resistance is carried out based on the fine numerical simulation result;

[0006] During the construction period, the stratum parameters and the lining parameters are reviewed and the input database is updated, the monitoring section is laid out, the monitoring data are obtained, the updated input database is input into the model for re-simulation, the data of the displacement, deformation and damage of the tunnel lining are extracted, the seismic resistance and fault resistance design is changed, and the service state of the tunnel is preliminarily evaluated in combination with the numerical simulation result;

[0007] During the operation period, annual inspection, long-term monitoring and emergency detection are carried out, and the input database is updated based on the monitoring and detection data; a prediction model database is established based on the dynamic data in the updated input database, a prediction model is constructed, and the prediction model is optimized based on the preprocessed prediction model database; the optimized prediction model is used to predict the development trend of the dynamic data, the input numerical value model is input with the predicted future service period dynamic data, and the numerical simulation is carried out; according to the numerical simulation result and the monitoring data, the tunnel service state evaluation and emergency treatment scheme are corrected.

[0008] Further, the collected data types include stratum aspect data, structure aspect data, and fault dislocation and seismic aspect data, and the collected data is divided into static database, dynamic database and uncertain database to form the input database of the model.

[0009] Further, the characteristics of the model include:

[0010] (1) Determination method of fault core width: T = 0.01-0.05W, wherein T is the core width and W is the influence zone width;

[0011] (2) Simulation of creep fault and stick-slip fault deformation characteristics: different slip surface friction coefficients are set for creep faults and stick-slip faults; different slip rates are set for creep faults and stick-slip faults;

[0012] (3) Only fault dislocation load is applied to creep fault; fault dislocation load and seismic load are applied to stick-slip fault.

[0013] Further, the refined numerical simulation includes:

[0014] (1) Numerical simulation of creep fault

[0015] Step 1: Initial geostress balance, set pre-defined geostress field and boundary normal constraint, and ensure that the vertical displacement of the stratum is less than a threshold value;

[0016] Step 2: Tunnel excavation simulation, through the birth and death element method, "kill" the tunnel excavation part of the soil, "activate" the initial support and secondary lining, simulate the tunnel excavation process, and simulate the construction process of the initial support and secondary lining based on the convergence constraint method;

[0017] Step 3: Application of creep fault dislocation, the quasi-static method is used to simulate the long-term cumulative shear displacement of the creep fault, that is, the annual creep rate is applied through the displacement boundary condition;

[0018] (2) Numerical simulation of stick-slip fault

[0019] The difference between the load for the numerical simulation of the creep fault and the load for the numerical simulation of the stick-slip fault is that:

[0020] Step3: Stick-slip fault dislocation- seismic motion is applied, displacement time history curve is used to simulate the rapid slip process for stick-slip fault, and viscoelastic boundary is set at the bottom and around the model to absorb reflected waves, and design earthquake acceleration time history is input to realize vibration coupling.

[0021] Further, the preliminary design of structure seismic and fault resistance based on the refined numerical simulation results includes: seismic and fault resistance length, seismic and fault resistance section length, seismic and fault resistance surrounding rock expansion size, and seismic and fault resistance surrounding rock grouting range;

[0022] The seismic and fault resistance length is the part along the tunnel axial angle greater than or equal to 0, the angle θ = dx / dL;

[0023] The seismic and fault resistance section length is:

[0024]

[0025] V s is the shear wave velocity of surrounding rock, T d is the frequency period of the seismic motion to be prevented, generally 0.2-0.5s

[0026] The seismic and fault resistance surrounding rock expansion size is:

[0027] R ex = kδ max

[0028] δ max is the maximum fault displacement, κ is the flexibility factor, which is adjusted according to the surrounding rock category

[0029] The seismic and fault resistance surrounding rock grouting range (radial + longitudinal)

[0030] Radial value:

[0031]

[0032] E c , E r , E e are the elastic modulus of fault core, original rock and surrounding rock in fault influence zone respectively, α is the stiffness factor of fault core surrounding rock, which is valued according to the completeness of the core surrounding rock, β is the stiffness factor of fault influence zone, which is valued according to the completeness of the surrounding rock in fault influence zone;

[0033] Longitudinal value:

[0034] L g = L s + η (L e + L c )

[0035] Ls is the anti-seismic and error-proof length, L e , L c are the fault influence zone and fault core length respectively, and η is the longitudinal amplification coefficient, which is determined according to the earthquake prevention level.

[0036] Further, the re-checking of the stratum parameters and lining parameters and the updating of the input database include:

[0037] When the tunnel construction reaches the fault core, the influence zone and the original rock, core sampling is performed on the surrounding rock of the working face, triaxial compression and ultrasonic wave velocity tests are used to obtain the elastic modulus, Poisson's ratio, cohesion and internal friction angle of the surrounding rock, and the strength and thickness of the tunnel lining of the constructed part are determined at the same time. At this time, the data of the constructed part

[0038] The dynamic data is changed to static data (such as the physical and mechanical parameters of the surrounding rock), and the static database and the dynamic database need to be updated. Further, the monitoring section is laid out, and the monitoring data are obtained, which include:

[0039] Monitoring sections are laid out every 5-40 m along the tunnel axis, and the density is increased to 5-10 m near the fault core and the influence zone; the monitoring includes: strain gauge monitoring of the lining circumferential and axial strain; multi-point displacement meter monitoring of the surrounding rock convergence deformation; crack monitor real-time recording of the crack width, depth and expansion rate; three-way sinusoidal stress meter monitoring of the ground stress change, etc. The monitoring data are uploaded to the monitoring platform in real time through optical fiber, providing a basis for the verification of the subsequent numerical model calculation results.

[0040] Further, the annual inspection, long-term monitoring and sudden situation detection include:

[0041] Annual inspection in operation period, comprehensive detection is carried out once a year, including lining appearance quality detection, lining material performance detection and transverse and longitudinal geometric shape detection.

[0042] Long-term monitoring, the development of surrounding rock deformation, lining stress, ground stress, etc. is monitored through the monitoring section.

[0043] Sudden situation detection, crack expansion, lining damage degree, lining material performance detection, transverse and longitudinal geometric shape detection are checked;

[0044] Based on the annual inspection, long-term monitoring and sudden situation detection data, the dynamic database is updated synchronously, providing data support for the intelligent prediction of dynamic data and fine numerical simulation.

[0045] Further, the establishment of the prediction model database, the construction of the prediction model, and the parameter optimization of the prediction model based on the pre-processed prediction model database include:

[0046] (1) The database of the prediction model is established, which includes ground stress, lining strength, lining thickness, fault creep rate, etc.

[0047] (2) Data preprocessing and grouping, including mapping parameters with large dimensional differences to the [0, 1] interval through normalization processing, cleaning abnormal data, and grouping data according to time sequence.

[0048] (3) Model construction and parameter optimization based on LightGBM algorithm, in which multi-step prediction is used to avoid cumulative errors. The LightGBM algorithm improves the overall performance of the model by sequentially increasing decision trees, and the specific formula of this process is as follows:

[0049] F m (x)=F m-1 (x)+γ m h m (x)

[0050] F m (x) is the model after the mth iteration; h m (x) is the mth tree; γ m is the learning rate of the mth tree

[0051] (4) Bayesian optimization

[0052] Optimize key parameters through Bayesian algorithm and time series cross-validation to improve model performance.

[0053] (5) Model verification, selecting E RMRS , E MAPE as evaluation indicators to verify the accuracy of the model.

[0054] (6) Model application, using the prediction model to predict the development trend of dynamic data.

[0055] Fine numerical simulation

[0056] Input the predicted dynamic data in the future service period into the numerical model for simulation, and extract the results of longitudinal displacement, transverse deformation, stress, strain, tensile damage, compressive damage, and crack distribution of the tunnel lining again.

[0057] Service state evaluation and dynamic correction of emergency treatment scheme

[0058] The above numerical simulation results and monitoring data can be applied to tunnel service state evaluation and correction of emergency treatment scheme.

[0059] A dynamic analysis system for the whole life cycle service state of a cross-activity fracture tunnel, comprising: a data acquisition module, an intelligent prediction module, a modeling simulation module, and a dynamic decision module.

[0060] Data acquisition module: used for acquiring static, dynamic, and uncertain data of the cross-activity fracture tunnel;

[0061] Intelligent prediction module: for dynamic data including ground stress, lining strength, lining thickness, fault creep rate, etc.

[0062] Modeling simulation module: for establishing a three-dimensional refined model of tunnel-surrounding rock-fault considering the internal structure of creep fault and stick-slip fault and its refined numerical simulation;

[0063] Dynamic decision module: for seismic and error prevention reinforcement changes, tunnel service state evaluation and emergency treatment scheme correction, etc.

[0064] The beneficial effects of the present application include:

[0065] 1. The present patent establishes a dynamic analysis method throughout the whole life cycle of the tunnel by combining the refined numerical model of ABAQUS and the prediction model of LightGBM.

[0066] 2. Creep and stick-slip faults are modeled independently, improving the pertinence of structural response analysis. In actual engineering, fault displacement mode directly affects structural stress. The present application simulates creep and stick-slip with different friction coefficients and slip rates, which is different from the traditional "unified slip surface" treatment method, and can reflect the deformation characteristics of creep and stick-slip faults.

[0067] 3. The input parameters of the traditional numerical model are fixed, and the changes of geological parameters, lining parameters and seismic parameters with time are not considered. The input parameters of the numerical model are divided into static, dynamic and uncertain databases in the present application, and the data is dynamically updated during the design-construction-operation process, so that the numerical model can reflect the actual geological environment, lining structural response and seismic action, etc. BRIEF DESCRIPTION OF DRAWINGS

[0068] Figure 1 The flowchart of the whole life cycle service state dynamic analysis method of the cross-active fault tunnel involved in the embodiments of the present application.

[0069] Figure 2 The three-dimensional refined model of tunnel-surrounding rock-fault considering the internal structure of creep fault and stick-slip fault involved in the embodiments of the present application.

[0070] Figure 3 The schematic diagram of the reinforcement length involved in the embodiments of the present application.

[0071] Figure 4 The deformation characteristics diagram of the stratum in the creep and stick-slip fault involved in the embodiments of the present application.

[0072] Figure 5 The deformation characteristics diagram of the tunnel in the creep and stick-slip fault involved in the embodiments of the present application.

[0073] Figure 6 A structural diagram of a cross-activity fault tunnel full life cycle service state dynamic analysis system related to the embodiments of the present application. DETAILED DESCRIPTION

[0074] In order to make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of the present application.

[0075] Embodiment 1

[0076] The following will be described in detail with reference to the accompanying drawings Figure 1 The specific embodiments of the present application will be described in detail;

[0077] A cross-activity fault tunnel full life cycle service state dynamic analysis method, comprising: dividing the full life cycle into a design period, a construction period and an operation period;

[0078] During the design period, data is collected and an input database is constructed, a tunnel-surrounding rock-fault three-dimensional refined model of a creep fault or a tunnel-surrounding rock-fault three-dimensional refined model of a stick-slip fault considering the internal structure of the fault is established, and based on the input database, refined numerical simulation is carried out, and finally based on the refined numerical simulation results, preliminary design of structure seismic resistance and error resistance is carried out;

[0079] During the construction period, the stratum parameters and lining parameters are reviewed and the input database is updated, the monitoring section is laid out, the monitoring data is obtained, the updated input database is input into the model for re-simulation, the tunnel lining displacement, deformation and damage data are extracted, the seismic resistance and error resistance design is changed in combination with the numerical simulation results, and the service state of the tunnel is preliminarily evaluated;

[0080] During the operation period, the input database is updated synchronously based on the monitoring data; a prediction model database is established based on the dynamic data in the updated input database, a prediction model is constructed, and the prediction model is optimized based on the preprocessed prediction model database; the optimized prediction model is used to predict the development trend of the dynamic data, the future service period dynamic data predicted by the prediction model is input into the numerical model for numerical simulation, and according to the numerical simulation results and the monitoring data, the tunnel service state evaluation and emergency treatment scheme are corrected.

[0081] In another embodiment, the collected data types include stratum aspect data, structure aspect data, and fault dislocation and ground motion aspect data, and the collected data is divided into static database, dynamic database, and uncertain database, which together constitute the input database of the model.

[0082] Stratum aspect data is collected from the engineering geological exploration report, including the geometry size, lithology, distribution range of each rock mass, physical parameters of rock mass (natural density, dry density, saturated density, etc.), mechanical parameters of rock mass (internal friction angle, dilation angle, deformation modulus, Poisson's ratio, etc.), and underground water level of the protolith, fault influence zone, and fault core.

[0083] Structure aspect data is collected from the engineering feasibility study and pre-feasibility study report, including the tunnel length, tunnel burial depth, tunnel cross-section type, tunnel cross-section inner contour size, lining strength, and lining thickness.

[0084] Fault dislocation and ground motion aspect data is collected from the engineering site environmental assessment report, including the creep rate, design ground motion peak acceleration, response spectrum characteristic period, ground motion time history, and maximum coseismic dislocation.

[0085] Among the above collected data, the tunnel length and tunnel burial depth are static, forming a static database.

[0086] The tunnel cross-section inner contour size, lining strength, lining thickness, creep rate, and ground stress are dynamic, forming a dynamic database.

[0087] The ground motion peak acceleration, response spectrum characteristic period, and ground motion time history are uncertain, forming an uncertain database.

[0088] The static database, dynamic database, and uncertain database together constitute the input database of the numerical model.

[0089] In another embodiment, a three-dimensional refined tunnel-surrounding rock-fault model considering the internal structure of the fault and the creep / stick-slip fault deformation characteristics is constructed. ABAQUS software is used to establish a three-dimensional refined tunnel-surrounding rock-fault model of creep fault or stick-slip fault considering the internal structure of the fault (such as Figure 2 ).

[0090] The refined model mainly includes the following features:

[0091] The establishment of the stratum in the numerical model considers the internal structure of the fault, specifically including:

[0092] (1) Determination method of fault core width: T = 0.01-0.05W, where T is the core width and W is the influence zone width

[0093] (2) Simulation of creep fault and stick-slip fault deformation characteristics:

[0094] a. Different slip surface friction coefficients are set for the creeping fault and the stick-slip fault: the value of the creeping fault is 0.6-0.8, and the value of the stick-slip fault is 0.2-0.4.

[0095] b. Different slip rates are set for the creeping fault and the stick-slip fault: the value of the creeping fault is 0.1-10 mm / year, and the value of the stick-slip fault is 0.1-10 m / s.

[0096] (3) Only fault displacement load is applied in the creeping fault; and both fault displacement load and seismic load are applied in the stick-slip fault.

[0097] In another embodiment, the refined numerical simulation comprises:

[0098] A refined numerical model is established by using ABAQUS finite element software. In order to reduce the influence of boundary effect, the height and width of the model are greater than 5 times the tunnel diameter.

[0099] The soil body and the supporting structure adopt C3D8R solid elements. In the numerical model, binding constraints are adopted between the initial support and the soil body, the original rock and the fault influence zone, and the fault influence zone and the fault core. Face-to-face contact is set between the initial support and the secondary lining. The normal direction adopts hard contact, and the tangential direction adopts friction contact with a friction coefficient of 0.4. Face-to-face contact is set on the slip surface of the fault core, the normal direction adopts hard contact, and the tangential direction adopts friction contact. The friction coefficient of the slip surface of the creeping fault is 0.6-0.8, and the friction coefficient of the slip surface of the stick-slip fault is 0.2-0.4.

[0100] The model soil body adopts the Mohr-Coulomb constitutive model, the initial support adopts the elastic-plastic constitutive model, and the secondary lining adopts the concrete damage plastic constitutive model.

[0101] The corresponding static data, dynamic data and uncertain data are input into the tunnel-surrounding rock-fault three-dimensional refined model of the creeping fault considering the internal structure of the fault and the tunnel-surrounding rock-fault three-dimensional refined model of the stick-slip fault, and numerical simulation of the creeping fault and the stick-slip fault is carried out respectively.

[0102] (1) Numerical simulation of the creeping fault

[0103] Step 1: Initial geostress equilibrium

[0104] By setting the pre-defined geostress field and the normal constraint of the boundary, the error between the initial geostress gradient and the measured geostress data is small, and the vertical displacement of the stratum is less than 0.00001 m, so as to control the initial deformation.

[0105] Step 2: Tunnel excavation simulation

[0106] Through the birth and death unit method, the soil body of the tunnel excavation part is "killed", the primary support and the secondary lining are "activated", the tunnel excavation process is simulated, the construction process of the primary support and the secondary lining is simulated based on the convergence constraint method, specifically, 20-30% of the tunnel peripheral surrounding rock stress is released in the primary support stage, 80-90% of the tunnel peripheral surrounding rock stress is released in the secondary lining stage, the dynamic balance of the surrounding rock-support system and the redistribution of the surrounding rock stress are realized.

[0107] Step3: Application of creep fault dislocation

[0108] For the creep fault, the pseudo-static method is used to simulate the long-term cumulative shear displacement, that is, the annual creep rate is applied through the displacement boundary condition;

[0109] (2) Numerical simulation of stick-slip fault

[0110] The initial stress balance of the stick-slip fault is the same as that of the creep fault and the tunnel excavation simulation

[0111] Step3: Application of stick-slip fault dislocation-seismic motion

[0112] For the stick-slip fault, the displacement time history curve is used to simulate the rapid slip process, and viscoelastic boundaries are set at the bottom and around the model to absorb reflected waves, and the designed seismic acceleration time history is input to realize vibration coupling.

[0113] In another embodiment, preliminary design of structure seismic and fault resistance is carried out based on the refined numerical simulation results, specifically including: seismic and fault resistance defense length, seismic and fault resistance section length, seismic and fault resistance surrounding rock expansion size and seismic and fault resistance surrounding rock grouting range, etc.

[0114] Seismic and fault resistance defense length:

[0115] The seismic and fault resistance defense length Ls is taken as the part along the tunnel axial angle greater than or equal to 0, the angle θ = dx / dL, the defense length is shown in the figure Figure 3 .

[0116] Seismic and fault resistance section length:

[0117]

[0118] Vs is the shear wave velocity of the surrounding rock, Td is the frequency period of the defense earthquake motion, generally taken as 0.2-0.5s;

[0119] Seismic and fault resistance expansion size:

[0120] R ex = κδ max

[0121] δ max is the maximum fault displacement, κ is the flexibility factor, which is adjusted according to the surrounding rock category

[0122] Anti-seismic and anti-fault grouting range (radial + longitudinal)

[0123] Radial value:

[0124]

[0125] Ec, Er, Ee are the elastic modulus of fault core, original rock, and fault-affected zone surrounding rock, respectively, α is the stiffness factor of the fault core surrounding rock, which is valued according to the completeness of the core surrounding rock, and β is the stiffness factor of the fault-affected zone, which is valued according to the completeness of the fault-affected zone surrounding rock.

[0126] Longitudinal value:

[0127] L g = L s + η (L e + L c )

[0128] Ls is the anti-seismic and anti-fault length, Le and Lc are the lengths of the fault-affected zone and the fault core, respectively, and η is the longitudinal amplification factor, which is determined according to the earthquake prevention level.

[0129] In another embodiment, the complex review of stratum parameters, lining parameters, and updating of the input database includes: when the tunnel construction reaches the fault core, affected zone, and original rock, core sampling is performed on the surrounding rock at the working face, respectively, and triaxial compression and ultrasonic wave velocity tests are used to obtain key parameters such as the elastic modulus, Poisson's ratio, cohesion, and internal friction angle of the surrounding rock, while the strength and thickness of the tunnel lining of the constructed part are also measured. At this time, this part of the data is changed from dynamic to static (such as the physical and mechanical parameters of the surrounding rock), and the static database and dynamic database need to be updated.

[0130] In another embodiment, the layout of the monitoring section and the acquisition of monitoring data include:

[0131] Monitoring sections are laid out every 5-40m along the tunnel axis, with a density of 5-10m near the fault core and affected zone. Monitoring includes:

[0132] Strain gauges monitor the lining's circumferential and axial strain; multi-point displacement meters monitor the convergence deformation of the surrounding rock; crack monitors record the crack width, depth, and expansion rate in real time; three-way sinusoidal stress meters monitor changes in ground stress, etc. The monitoring data is uploaded to the monitoring platform in real time through optical fibers, providing a basis for the verification of subsequent numerical model calculation results.

[0133] Data updating and refined numerical simulation

[0134] The updated static data and dynamic data are input into a numerical model, and the tunnel lining displacement, deformation and damage data are extracted by re-simulating based on the birth-death element method and convergence constraint method through ABAQUS software, to provide data support for design changes.

[0135] Construction period seismic and error resistant design changes and service state preliminary evaluation

[0136] Combined with the numerical simulation results, the seismic and error resistant design is changed, and the service state of the tunnel is preliminarily evaluated.

[0137] In another embodiment, the annual inspection, long-term monitoring and emergency detection include:

[0138] Annual inspection during operation, comprehensive detection is carried out once a year, including lining appearance quality detection (geological radar scanning detection of lining cavity), lining material performance detection (rebound detector detects lining strength) and transverse and longitudinal geometric shape detection (three-dimensional laser scanning detects structure displacement and deformation) and the like.

[0139] Long-term monitoring, through the sensors arranged on the above-mentioned typical sections, the development of surrounding rock deformation, lining stress, ground stress and the like is long-term monitored.

[0140] Emergency detection, mainly for starting emergency detection within 24 hours after the earthquake, focusing on checking crack expansion, lining damage degree, lining material performance detection (rebound detector detects lining strength), transverse and longitudinal geometric shape detection (three-dimensional laser scanning detects structure displacement and deformation) and the like.

[0141] Based on the annual inspection, long-term monitoring and emergency detection data, the dynamic database is updated synchronously, to provide data support for intelligent prediction of dynamic data and fine numerical simulation.

[0142] In another embodiment, the establishing a prediction model database and constructing a prediction model, and performing parameter optimization on the prediction model based on the preprocessed prediction model database include:

[0143] (1) Establishing a database of prediction models, the database including ground stress, lining strength, lining thickness, fault creep rate and the like.

[0144] (2) Data preprocessing and grouping, including mapping parameters with large dimensional differences to the [0, 1] interval through normalization processing, abnormal data cleaning and grouping data according to time sequence.

[0145] (3) Model construction and parameter optimization based on the LightGBM algorithm, in the construction of the model, multi-step prediction is adopted to avoid cumulative error. The algorithm of LightGBM is to improve the overall performance of the model by sequentially increasing the decision tree, and the specific formula of this process is as follows:

[0146] Fm (x) = F m-1 (x) + γ m h m (x)

[0147] F m (x) is the model after the mth iteration; h m (x) is the mth tree; γ m is the learning rate of the mth tree;

[0148] (4) Bayesian optimization

[0149] Optimize key parameters through Bayesian algorithms and time series cross-validation to improve the performance of the model.

[0150] (5) Model validation, select E RMRS , E MAPE as evaluation indicators to verify the accuracy of the model.

[0151] (6) Application of the model, use the model to predict the development trend of dynamic data such as ground stress.

[0152] Fine numerical simulation:

[0153] Input the predicted dynamic data in the future service period into the numerical model for simulation, and extract the results of tunnel lining longitudinal displacement, transverse deformation, stress, strain, tensile damage, compressive damage, and crack distribution again.

[0154] Service state assessment and emergency treatment scheme dynamic correction:

[0155] The above numerical simulation results and monitoring data can be applied to the assessment of tunnel service state and the correction of emergency treatment scheme.

[0156] This embodiment simulates creep and stick-slip using different friction coefficients and slip rates, which is different from the traditional "unified slip surface" treatment method, and can reflect the deformation characteristics of creep and stick-slip faults. The deformation characteristics of creep and stick-slip faults are shown in Figure 4 and Figure 5 : Figure 4 (a) Reflects the distribution of stratum deformation along the Y direction under different displacement amounts (b) Reflects the percentage of fault slip amount in total displacement amount under different displacement amounts (c) Reflects the percentage of fault core displacement in total displacement amount under different displacement amounts; Figure 5 Reflects the deformation distribution of the tunnel along the tunnel axis direction under two different displacement types of creep fault and stick-slip fault, and compares the deformation response under different displacement amounts (δ = 0.2m, 0.4m, 0.6m). The deformation characteristics are as follows:

[0157] 1. Creep fault: the percentage of fault slip in total displacement is low, the percentage of fault core displacement in total displacement is high, the stratum deformation range is large, the tunnel displacement distribution is uniform, the curve presents S-shaped smooth transition, the transition zone is wide, the influence range on tunnel structure is large, and the damage degree is small.

[0158] 2. Stick-slip fault: the percentage of fault slip in total displacement is high, the percentage of fault core displacement in total displacement is high, the stratum deformation range is small, the fault slip is the main feature of creep fault displacement, the tunnel displacement changes dramatically, there is obvious steep rise in the fault core, the transition zone is narrow, the influence on tunnel structure is concentrated, and the damage degree is large.

[0159] In another embodiment, it relates to a dynamic analysis system for the service state of a tunnel across an active fault throughout its life cycle, as shown in the figure, the system mainly includes: a data acquisition module, an intelligent prediction module, a modeling simulation module and a dynamic decision module. Figure 6

[0160] Data acquisition module: used for acquiring static, dynamic and uncertain data of a tunnel across an active fault;

[0161] Intelligent prediction module: used for predicting dynamic data such as ground stress, lining strength, lining thickness, fault creep rate, etc.

[0162] Modeling simulation module: used for establishing a three-dimensional refined model of tunnel-surrounding rock-fault considering the internal structure of creep fault and stick-slip fault and its refined numerical simulation;

[0163] Dynamic decision module: used for seismic and fault resistance fortification change, tunnel service state evaluation and emergency treatment scheme correction, etc.

[0164] The above-described embodiments only express the specific implementation of the present application, and the description is more specific and detailed, but it cannot be understood as a limitation on the protection scope of the present application. It should be noted that for ordinary skilled persons in the art, without departing from the technical concept of the present application, a number of modifications and improvements can be made, which are within the protection scope of the present application.​

Claims

1. A method for dynamic analysis of service state of a tunnel across an active fault fracture throughout a life cycle, characterized in that, The method comprises the following steps: The whole life cycle is divided into design period, construction period and operation period; In the design period, data is collected and an input database is constructed, a three-dimensional fine model of tunnel-surrounding rock-fault of a creep fault or a three-dimensional fine model of tunnel-surrounding rock-fault of a stick-slip fault considering the internal structure of the fault is established, and fine numerical simulation is carried out based on the input database, and finally, preliminary design of structure seismic resistance and error resistance is carried out based on the fine numerical simulation results; In the construction period, stratum parameters and lining parameters are reviewed and the input database is updated, a monitoring section is laid out, monitoring data is obtained, the updated input database is input into the model for re-simulation, tunnel lining displacement, deformation and damage data are extracted, combined with the numerical simulation results, the seismic resistance and error resistance design is changed, and the service state of the tunnel is preliminarily evaluated; In the operation period, annual inspection, long-term monitoring and sudden situation detection are carried out, and the input database is updated based on the monitoring and detection data; A prediction model database is established based on the dynamic data in the updated input database, a prediction model is constructed, the prediction model is optimized based on the preprocessed prediction model database, the optimized prediction model is used to predict the development trend of the dynamic data, the future service period dynamic data obtained by prediction is input into the numerical model for numerical simulation, and the service state of the tunnel is evaluated and the emergency treatment scheme is corrected according to the numerical simulation results and the monitoring data.

2. The method of claim 1, wherein, The method comprises the following steps: collecting data types including stratum aspect data, structure aspect data and fault dislocation and seismic motion aspect data, dividing the collected data into static database, dynamic database and uncertain database, and combining to form an input database of the model.

3. The method of claim 1, wherein, The model comprises the following characteristics: (1) a determination method of the fault core width: T=0.01-0.05W, wherein T is the core width and W is the influence zone width; (2) simulation of deformation characteristics of creep faults and stick-slip faults: different slip surface friction coefficients are set for the creep faults and the stick-slip faults; different slip rates are set for the creep faults and the stick-slip faults; (3) only fault dislocation load is applied to the creep fault; fault dislocation load and seismic load are simultaneously applied to the stick-slip fault.

4. The method of claim 1, wherein, The fine numerical simulation comprises: (1) numerical simulation of the creep fault Step 1: initial ground stress balance, pre-defined ground stress field and boundary normal constraint are set, and the vertical displacement of the stratum is ensured to be less than a threshold value; Step 2: tunnel excavation simulation, through the birth and death element method, the soil body of the tunnel excavation part is "killed", the initial support and the secondary lining are "activated", the tunnel excavation process is simulated, and the construction process of the initial support and the secondary lining is simulated based on the convergence constraint method; Step 3: application of creep fault dislocation, the pseudo-static method is used to simulate the long-term cumulative shear displacement of the creep fault, that is, the annual creep rate is applied through the displacement boundary condition; (2) numerical simulation of the stick-slip fault The difference between the load of the numerical simulation of the creep fault and the load of the numerical simulation of the stick-slip fault is that: Step3: Dislocation of stick-slip fault - application of ground motion, displacement time history curve is used to simulate the rapid slip process for stick-slip fault, and the bottom and four sides of the model are set as viscoelastic boundary to absorb reflected waves, and the input design earthquake acceleration time history is used to realize vibration coupling.

5. The method of claim 1, wherein, Based on the results of fine numerical simulation, the preliminary design of structure seismic and fault resistance includes: seismic and fault resistance length, seismic and fault resistance section length, seismic and fault resistance surrounding rock expansion size and seismic and fault resistance surrounding rock grouting range; The seismic and fault resistance length is taken along the part of the tunnel axial angle greater than or equal to 0, and the angle θ = dx / dL; The seismic and fault resistance section length is: V s is the shear wave velocity of the surrounding rock, T d is the frequency period of the protected seismic motion, generally taken as 0.2-0.5s; The seismic and fault resistance surrounding rock expansion size is: R ex = Kδ max delta max is the maximum fault displacement, and kappa is a flexibility factor, adjusted according to the surrounding rock class; The seismic and fault resistance surrounding rock grouting range (radial + longitudinal) The radial value is: E c 、E r 、E e Elastic modulus of fault core, original rock, and surrounding rock in the fault influence zone, respectively, α is the stiffness factor of the surrounding rock in the fault core, which is valued according to the completeness of the surrounding rock in the fault core, and β is the stiffness factor of the fault influence zone, which is valued according to the completeness of the surrounding rock in the fault influence zone. The longitudinal value is: L g = L s + η(L e + L c ) L s is the anti-seismic and error prevention length, L e , L c are the fault influence zone and fault core length, respectively, and η is the longitudinal amplification factor, which is determined according to the earthquake prevention level.

6. The method of claim 1, wherein, The said review of stratum parameters, lining parameters and updating of input database includes: When the tunnel construction reaches the fault core, influence zone and original rock, drill core sampling is carried out on the surrounding rock of the working face, and triaxial compression and ultrasonic wave velocity test is used to obtain the elastic modulus, Poisson's ratio, cohesion and internal friction angle of the surrounding rock, and the strength and thickness of the tunnel lining of the constructed part are also measured. At this time, part of the data is changed from dynamic to static, which needs to update the static database and dynamic database.

7. The method of claim 1, wherein, The said layout of monitoring section and acquisition of monitoring data includes: Monitoring section is laid along the tunnel axis every 5-40m, and the density is increased to 5-10m near the fault core and influence zone; Monitoring includes: strain gauge monitoring of lining circumferential and axial strain; multi-point displacement meter monitoring of surrounding rock convergence deformation; crack monitor real-time recording of crack width, depth and expansion rate; three-way sinusoidal stress meter monitoring of ground stress change, and monitoring data is uploaded to the monitoring platform in real time through optical fiber, which provides basis for verification of subsequent numerical model calculation results.

8. The method of claim 1, wherein, The said annual inspection, long-term monitoring and sudden situation detection includes: Annual inspection during operation, which is carried out once a year, including lining appearance quality detection, lining material physical and mechanical property detection and horizontal and longitudinal geometric shape detection; Long-term monitoring, which monitors the development of surrounding rock deformation, lining stress and ground stress through long-term monitoring of monitoring section; Sudden situation detection, which checks crack expansion, lining damage degree, lining material performance detection and horizontal and longitudinal geometric shape detection; Based on the data of annual inspection, long-term monitoring and sudden situation detection, dynamic database is updated synchronously, which provides data support for intelligent prediction of dynamic data and fine numerical simulation.

9. The method of claim 1, wherein, The said establishment of prediction model database, construction of prediction model and parameter optimization of prediction model based on preprocessed prediction model database includes: (1) Establishment of prediction model database, which includes ground stress, lining strength, lining thickness and fault creep rate; (2) Data preprocessing and grouping, including mapping of parameters with large dimensional difference to [0, 1] interval through normalization processing, abnormal data cleaning and grouping of data according to time sequence; (3) Construction and parameter optimization of model based on LightGBM algorithm, which adopts multi-step prediction to avoid cumulative error in model construction; (4) Bayesian optimization, which optimizes key parameters through Bayesian algorithm and time series cross-validation; (5) Model validation, select E RMRS , E MAPE as an evaluation index to verify the accuracy of the model; (6) Application of model, which predicts the development trend of dynamic data using prediction model.

10. A system for dynamic analysis of service state of a tunnel across an active fault fracture throughout its life cycle, characterized in that, ​ Data acquisition module, intelligent prediction module, modeling simulation module and dynamic decision module; Data acquisition module: used for collecting static, dynamic and uncertain data of cross-activity fault tunnel; Intelligent prediction module: used for the prediction of dynamic data including ground stress, lining strength, lining thickness and fault creep rate; Modeling simulation module: used for establishing three-dimensional refined model of tunnel-surrounding rock-fault considering internal structure of fault and its refined numerical simulation of creep fault and stick-slip fault; Dynamic decision module: used for seismic and error prevention alteration, tunnel service state evaluation and emergency treatment scheme modification.

Images