CN-120806630-B - Method for analyzing influence of railway tunnel construction on urban underground water

Abstract

The invention discloses an influence analysis method of railway tunnel construction on urban groundwater, which belongs to the field of railway tunnel construction and comprises the following steps of S1, multi-source data acquisition and pretreatment, S2, construction of geological-hydrologic-construction-remote sensing multidimensional feature vectors, extraction of groundwater disturbance indexes, S3, establishment of a stress-seepage-heat conduction three-field coupling model, S4, analysis of the three-field coupling effect based on the three-field coupling model, output of comprehensive risk level of tunnel construction on groundwater influence, S5, fusion of the comprehensive risk level by utilizing a GIS space analysis function, realization of space visualization and grading early warning of groundwater influence by combining real-time data, S6, updating of geological parameters of the three-field coupling model in real time, and searching of optimal construction parameter combinations. By adopting the method for analyzing the influence of the railway tunnel construction on the urban underground water, the scientificity and the engineering practicability of the influence analysis of the railway tunnel construction on the urban underground water are obviously improved.

Inventors

• XU WANGHAO
• CUI JIAN
• TANG CHAO

Assignees

• 甬舟铁路有限公司

Dates

Publication Date

20260508

Application Date

20250704

Claims (6)

1. 1. The method for analyzing the influence of railway tunnel construction on urban groundwater is characterized by comprising the following steps of: s1, multi-source data acquisition and pretreatment are carried out, and a multi-dimensional space-time data set containing ground surface cracks, geological hydrology and construction disturbance is constructed; S2, integrating the preprocessed multi-source data through a multi-dimensional feature fusion algorithm, constructing a geological-hydrologic-construction-remote sensing multi-dimensional feature vector, identifying the face state by utilizing an improved RNPC-net model, and extracting an underground water disturbance index; s3, based on an elastoplastic damage theory and seepage mechanics, establishing a stress-seepage-heat conduction three-field coupling model, and setting boundary conditions of the stress-seepage-heat conduction three-field coupling model based on underground water disturbance indexes; s4, analyzing a three-field coupling effect based on the three-field coupling model, and outputting a comprehensive risk level of the tunnel construction on the influence of the underground water based on the comprehensive risk assessment model; S5, fusing comprehensive risk levels by utilizing a GIS space analysis function, and realizing space visualization and grading early warning of groundwater influence by combining real-time data; S6, using an extended Kalman filtering algorithm to update geological parameters of the three-field coupling model in real time, combining a digital twin technology to construct a virtual mapping of the construction process, searching for an optimal construction parameter combination, and forming monitoring-analysis-optimization closed-loop control; the step S2 specifically comprises the following steps: S21, constructing a geological-hydrologic-construction-remote sensing multidimensional feature vector: S211, based on the corrected vertical displacement sequence Diameter of tunnel And depth of burial Calculating sedimentation characteristic parameters including volume loss rate Sum groove width ; Calculation of volume loss rate based on Gaussian sedimentation model : ; In the formula, Representing the volume of the settling tank, and , Representing the vertical displacement distribution function of the earth surface subsider, and , Represents the maximum settlement of the earth surface, , Indicating the inflection point distance of the settling tank, Representing transverse coordinates; represents the cross-sectional area of the tunnel, an , Represents the tunnel diameter; the groove width was calculated using the following formula : ; In the formula, Representing inflection point distance; Representing the tunnel burial depth; s212, constructing a feature vector, namely integrating geological hydrologic data, construction data, remote sensing data and ground fracture distribution data to obtain a feature matrix , Representing the earth's surface porosity; representing anisotropic permeability coefficient tensors of each layer; Representing a loss factor; representing the propulsion speed; Representing cutter torque; s213, mapping each eigenvalue in the eigenvalue matrix to each eigenvalue by adopting a minimum-maximum standardization method Obtaining normalized feature matrix in interval ; S22, recognizing the face, namely constructing an improved RNPC-net model, and utilizing the preprocessed face image, the three-dimensional coordinates and the feature matrix Training and improving RNPC-net model until convergence, and outputting weathering grade probability distribution And groundwater state probability distribution : The improved RNPC-net model adopts a three-branch mixed feature extraction module, wherein the three-branch mixed feature extraction module comprises a two-dimensional image convolution branch taking ResNet as a main body, a three-dimensional point cloud processing branch and a construction parameter full-connection branch, and the multi-source data input is fused through a self-adaptive weight auxiliary classifier; The cross entropy loss function is set as follows: ; in the formula, Representing a loss value; Representing a label vector; Representing a distribution probability; representing a total category; representing a category index; S23, extracting underground water disturbance indexes, wherein the underground water disturbance indexes comprise weathering comprehensive indexes And groundwater activity intensity index ; Wherein based on a weathering grade probability distribution Calculating the comprehensive index of weathering : ; In the formula, Representing the weight; Represent the first The predicted probability of the class of weathering grade, Respectively represent non-efflorescence, micro-efflorescence, medium efflorescence and strong efflorescence; and distributing groundwater state probability Mapping to numerical index, constructing groundwater activity intensity index : ; In the formula, 、 、 And The water inflow probability of the tunnel face, the water dripping line forming probability of the tunnel face and the drying probability of the tunnel face are respectively represented; The step S3 specifically comprises the following steps: s31, modeling an elastoplastic damage stress field: s311, inverting damage area parameters including blasting damage radius and anisotropic material parameter matrix The blasting damage radius is estimated by adopting the following empirical formula, and is calibrated by combining the sound wave test result to determine the damage area range: ; in the formula, Representing geometric parameters of the damaged area; representing the geological correction coefficient; Anisotropic material parameter matrix Modulus of elasticity in damaged region in (a) The calculation formula is as follows: ; in the formula, Representing the modulus of elasticity of the virgin rock; Poisson's ratio The calculation formula is as follows: ; in the formula, Poisson's ratio representing the original rock; S312, radial effective stress of elastic region And tangential effective stress : ; ; In the formula, Representing the stress of the original rock; representing the inner wall of a tunnel; represents the inner radius of the tunnel; Deriving radial effective stress of plastic region based on Mohr-Coulomb criterion And tangential effective stress : ; ; In the formula, Represents the cohesion of the plastic region and , Represents the plastic region cohesion; Representing the soil pressure coefficient in the plastic region, an ; Representing the radius of the plastic region; represents the internal friction angle of the plastic region, and , Representing the damage reduction coefficient; s32, constructing a seepage equation considering the temperature effect, and setting boundary conditions to obtain a pore water pressure distribution cloud chart and seepage flow Wherein the percolation equation is expressed as follows: ; in the formula, Represents a seepage velocity vector; represents the dynamic viscosity of water; Represents pore water pressure; Represents water density; representing gravitational acceleration; representing the head height; representing the coefficient of thermal expansion; representing a temperature field distribution; the boundary conditions are as follows Pore water pressure , Indicating the water pressure in tunnel and far field Time of day , Indicating the radius of influence of the seepage flow, Representing the initial pore water pressure; Seepage flow The calculation formula is as follows: ; in the formula, Represents the aquifer thickness; representing the initially estimated permeability coefficient.
2. 2. The method for analyzing the influence of railway tunnel construction on urban groundwater according to claim 1, wherein the step S1 comprises the following steps: S11, acquiring multisource data, namely acquiring Sentinel-1A/B satellite C wave band remote sensing data through an EGMS (enhanced gas/liquid separation system) to acquire ground surface fracture data, simultaneously acquiring drilling data, geological investigation report, groundwater level monitoring data, permeability coefficient test results and pore water pressure distribution data to form geological hydrologic data, and acquiring construction process parameters and a face image to form construction disturbance data; s12, preprocessing multi-source data: For remote sensing data, firstly, according to the orbit geometric parameters of a Sentinel-1 satellite, LOS displacement sequences of the orbit ascending and descending data of the remote sensing data are projected to the vertical direction, and the radar incident angle is eliminated The deviation is filtered through a coherence threshold value to keep effective monitoring points, and a corrected vertical displacement sequence is output : ; In the formula, Representing an original LOS displacement time sequence; then effective monitoring points are screened through coherence threshold, construction disturbance components and natural sedimentation components in the displacement signals are separated by adopting a principal component analysis-independent component analysis method decomposition technology, and a construction disturbance displacement sequence is extracted ; For drilling data, testing particle size by sieving and sedimentation to obtain agglomerate content, and primarily estimating permeability coefficient based on agglomerate content , Representing a reference osmotic coefficient related to the content of the cosmid, and finally calibrating the initially estimated osmotic coefficient by a pressurized water test Obtaining anisotropic permeability coefficient tensor of each layer : Meanwhile, analyzing the crack development degree according to the CT image of the drill hole and defining the damage factor , And Respectively representing the length of the surface crack and the length of the rock core, and generating a three-dimensional damage factor field by adopting common Kriging interpolation 。
3. 3. The method for analyzing influence of railway tunnel construction on urban groundwater according to claim 2, wherein in step S11, drilling data is obtained by arranging drilling holes every 50m-100m along a tunnel axis, penetrating the depth of the drilling holes 10m-20m below a tunnel bottom plate, and performing lithology classification on collected core samples to obtain lithology distribution data; At the same time, the standard penetration test is adopted to obtain the number of hits penetrating 30cm Estimating effective internal friction angle of rock mass by combining De Mello empirical formula : ; The construction process parameters comprise TBM tunneling parameters and blasting construction parameters, and the TBM tunneling parameters comprise cutter torque Speed of propulsion And shield pressure The blasting construction parameters comprise single-hole loading capacity Distance between explosive centers Blasting time interval and surface vibration velocity 。
4. 4. The method for analyzing the influence of railway tunnel construction on urban groundwater according to claim 3, wherein the step S4 comprises the following steps: s41, calculating total stress by coupling the stress field and the seepage field: ; in the formula, And Respectively representing radial total stress and tangential total stress; And Respectively represent radial stress and tangential stress of stress field, and , ; S42, introducing a construction heat source, calculating temperature distribution through a heat conduction equation, and correcting seepage parameters: ; in the formula, Represents the dynamic viscosity of the water after temperature correction, and , Represents the dynamic viscosity of the reference water, The temperature coefficient is represented by a temperature coefficient, And Respectively representing the real-time temperature and the initial temperature; S43, setting risk indexes as water burst risk, sedimentation risk and extrusion risk, wherein the water burst risk Risk of sedimentation Risk of extrusion , And Are all set up to be a threshold value, Representing rock quality index, obtaining single risk factor vector ; S44, constructing a comprehensive risk index : ; In the formula, 、 And All represent weights, and ; S45, based on comprehensive risk index The risk is classified into low risk, medium risk and high risk, and a risk-measure mapping relationship is established.
5. 5. The method for analyzing the influence of railway tunnel construction on urban groundwater according to claim 4, wherein the step S5 comprises the following steps: s51, uniformly converting the geological parameters, the analysis results of the three-field coupling model and the real-time monitoring data into ESRI SHAPEFILE format, and projecting the result to a UTM coordinate system; s52, creating a thematic layer of a geological layer, a stress field and a seepage field in the ArcGIS Pro; s53, interpolating the discrete risk points by adopting an inverse distance weighting method to generate a continuous risk curved surface: ; in the formula, Representing three-dimensional space coordinates Continuous risk value at; representing the predicted point and the first The spatial distance of the discrete risk monitoring points; Represent the first A comprehensive risk index of the discrete risk monitoring points; Representing the total number of discrete risk monitoring points; S54, constructing a BIM tunnel model based on the analysis results of the tunnel geometric parameters, the construction process parameters and the three-field coupling model, and importing the BIM tunnel model and the comprehensive risk level data into a Unity 3D platform to realize dynamic association display of construction progress and risk evolution; s55, setting a grading early warning rule, namely a first grade early warning condition, namely a comprehensive risk index And the earth surface sedimentation rate is more than 1 mm/day, triggering red early warning and automatically starting an emergency grouting plan; Secondary early warning condition: or the sedimentation rate is more than 0.5 mm/day, triggering yellow early warning and prompting the strengthening monitoring frequency.
6. 6. The method for analyzing the influence of railway tunnel construction on urban groundwater according to claim 5, wherein the step S6 comprises the following steps: S61, updating geological parameters of the three-field coupling model in real time by using the extended Kalman filter, and outputting updated damage factors Coefficient of permeability ; Defining a state vector Observation vector , Represents the actual measurement value of the vertical displacement of the ground surface, Expressing the actual measurement value of seepage flow, and establishing the following state equation and observation equation: ; ; in the formula, And Respectively represent State vector predictor sum of time instants A state vector estimate of time; representing a state transfer function; representing state noise; Representation of An observation vector at a moment; representing an observation function; representing observed noise; Iteratively calculating a state estimation value by using an extended Kalman filtering algorithm: ; in the formula, And Respectively represent Optimal estimation sum of time state vectors based on Time of day estimation A time state prediction value; represents the Kalman gain, an , The representation is based on Time of day estimation Covariance matrix of time-of-day state prediction, A covariance matrix representing observed noise; s62, updating the updated damage factor Coefficient of permeability Substituting the three-field coupling model in the step S3 to recalculate stress field, seepage field and risk index; S63, comparing the recalibration result with the actual measurement data, calculating the root mean square error to evaluate the accuracy of the three-field coupling model, and returning to the step S61 if the root mean square error is more than 5%, and performing a secondary calibration flow; s64, optimizing construction parameters; s641, constructing a digital twin model, namely developing a tunnel construction digital twin body based on a Unity platform, integrating a BIM tunnel model and real-time monitoring data, and realizing 1:1 virtual mapping in the construction process; S642, adopting a near-end strategy optimization algorithm to minimize the comprehensive risk index To aim, search for optimal parameter combinations , And Respectively representing the optimal support pressure and the optimal excavation rate.

Description

Method for analyzing influence of railway tunnel construction on urban underground water Technical Field The invention relates to the technical field of railway tunnel construction, in particular to a method for analyzing influence of railway tunnel construction on urban groundwater. Background Along with the acceleration of the urban process, railway tunnel engineering increasingly goes deep into urban dense areas, and the influence of construction disturbance on urban groundwater systems has become a key problem of engineering safety and environmental protection. The traditional analysis method has the following technical bottlenecks: 1. The prior art mainly relies on single data sources (such as drilling investigation or InSAR (synthetic aperture radar interferometry) monitoring, so that multidimensional influences of tunnel construction on underground water are difficult to comprehensively reflect, wherein the traditional drilling sampling density is low (one drilling is usually arranged every 50-100 m), space variation of rock mass damage and permeability coefficients is difficult to capture, an estimated error of the permeability coefficients based on a Standard Penetration Test (SPT) can reach +/-30%, and permeability mutation caused by blasting damage cannot be reflected. 2. The tunnel face state identification has the defect of intellectualization, namely the Weathering Degree (WD) and the groundwater state (GC) of the tunnel face are key factors influencing the seepage field. For this, early reliance on manual geological logging is time consuming and is subject to considerable experience. Later, a structural surface identification method based on point cloud clustering or region growth is developed, but parameters are required to be set manually, and robustness is poor. For example, DBSCAN-based clustering algorithms have a false positive rate of over 15% in the fracture-dense region. 3. The modeling of the multi-physical field coupling is incomplete, the existing model does not fully consider the multi-physical field interaction effect of construction disturbance, such as stress-seepage coupling simplification, and the elastoplastic model based on Mohr-Coulomb criterion ignores the influence of damage on permeability. In certain railway tunnel engineering, permeability attenuation (actually reduced to 1/5 of the original rock) of a blasting damaged area is not considered, so that the seepage prediction error exceeds 50%. Temperature effects are absent-pore water pressure changes caused by tunnel construction heat sources (e.g., mechanical frictional heating) are not included in the analysis. Theoretical studies show that a 5 ℃ temperature rise can increase the seepage rate by 12% -18%, but the effect is generally ignored by the existing model. The dynamic damage evolution ignores that the time-space evolution process of the damage factor (D) is not coupled with the seepage field in real time, and the dynamic change of the permeability caused by support hysteresis cannot be reflected. 4. The existing risk assessment system has the following defects that (1) a static index system cannot reflect multi-factor coupling effect based on the risk classification of a single parameter (such as a seepage threshold). In certain tunnel engineering, seepage flow does not exceed a threshold value, but sudden water burst is caused by stress concentration, so that the limitation of the traditional method is exposed. (2) The space analysis capability is weak, and the integration of GIS (geographic information system) and BIM (building information model) is only stopped at the superposition of static image layers, and the dynamic risk field simulation is lacked. For example, in a subway project, risk stack effects of fault intersection areas are not recognized in time, resulting in emergency plan hysteresis. (3) The real-time performance is insufficient, the update period of the monitoring data and the model is long (usually in days), and the instantaneous risk change in the construction process cannot be captured. The seepage flow is increased by 3 times within 1 hour after blasting of a certain tunnel, but the existing system can not early warn in time. 5. The traditional construction parameter adjustment depends on experience, does not form a closed loop of 'monitoring-analysis-optimization', and has the following defects that (1) the parameter sensitivity analysis lacks, and the optimization of the support pressure and the excavation rate lacks quantitative basis. In certain tunnel engineering, blindly improving the supporting pressure leads to 20% increase in cost, but less than 10% decrease in risk. (2) The dynamic adaptability is poor, and the parameter automatic adjustment cannot be triggered by the geological condition change (such as sudden fault) in the construction process. Because the blasting parameters of a certain railway tunnel are not adjusted in time, the overbreak amount is increased by 30%, and