Shield construction ground settlement automatic monitoring method based on intelligent sensing

Abstract

This invention discloses an automatic monitoring method for ground settlement during shield tunneling construction based on intelligent sensing, belonging to the field of ground settlement monitoring technology for shield tunneling construction. The method involves deploying a sensor array consisting of multiple intelligent settlement monitoring nodes in the shield tunneling impact zone. Each node incorporates a triaxial accelerometer and a pore water pressure gauge. Initial surface elevation data is acquired to generate a first surface morphology grid. Local surface tilt vectors are determined based on the tilt angle changes of the triaxial accelerometers, and the effective soil stress is corrected using the water pressure changes from the pore water pressure gauges to obtain a corrected tilt vector. Based on all corrected tilt vectors, the first surface morphology grid is reconstructed node-by-node curvature to generate a second surface morphology grid. The second surface morphology grid is spatially coupled with shield tunneling parameters, and the morphological deviation is extracted as the ground settlement monitoring result. This method can reconstruct continuous surface morphology from discrete sensor data and accurately extract settlement changes at the tunneling location.

Description

Technical Field

[0001] This invention relates to the field of ground settlement monitoring technology in shield tunneling construction, specifically to an automatic monitoring method for ground settlement in shield tunneling construction based on intelligent sensing. Background Technology

[0002] During tunnel boring machine (TBM) construction, timely acquisition of the spatial distribution of surface settlement is crucial for protecting the safety of surrounding buildings and structures. Traditional ground settlement monitoring often uses leveling or total stations to periodically collect discrete point elevations, which are then interpolated to generate regional settlement cloud maps. This discrete monitoring method suffers from limitations in monitoring frequency, high human intervention, and the inability to perceive continuous changes in surface tilt and curvature in real time. Some automated solutions utilize hydrostatic leveling systems or GNSS for continuous elevation acquisition, but the measurement points are sparse, making it difficult to reconstruct the complex undulations of the surface morphology under construction disturbance. When uneven settlement occurs in a localized area, the true curvature between discrete points is easily masked by smoothing, leading to an underestimation of the risk of differential settlement.

[0003] Existing technologies have attempted to capture changes in surface tilt using tilt sensors, but relying solely on tilt data cannot distinguish between soil skeleton deformation and the buoyancy tilting effect caused by fluctuations in pore water pressure within the soil. During the process of changes in excess pore water pressure caused by tunnel boring machine (TBM) excavation, changes in the effective stress of the surface soil will generate additional lateral deformation components. If the tilt measurement value is directly used as the surface tilt amount, it will include false tilts not caused by settlement, affecting the accuracy of settlement assessment. Furthermore, existing monitoring methods do not spatially register the dynamic changes in surface morphology with the TBM excavation parameters in real time, making it difficult to accurately separate the surface deformation induced at the current excavation position from the overall surface morphology, resulting in lag and spatial ambiguity in the assessment of the stratum response. Therefore, how to eliminate the interference of pore water pressure changes on surface tilt detection, reconstruct a surface grid morphology that approximates the actual bending state from discrete sensor data, and accurately couple the reconstructed morphology with the excavation position during TBM advancement to extract the corresponding morphological changes are urgent problems to be solved. Summary of the Invention

[0004] This invention provides an automatic monitoring method for ground settlement during shield tunneling based on intelligent sensing. The purpose is to reconstruct a surface curvature grid corrected by the effective stress of the soil from discrete dip angle and water pressure sensing data, and to couple this grid with the shield tunneling location to automatically extract the morphological deviation within the tunneling influence area to reflect the spatial distribution of ground settlement.

[0005] To achieve the above objectives, the present invention provides the following technical solution: The present invention provides an automatic monitoring method for ground settlement during shield tunneling construction based on intelligent sensing, comprising:

[0006] The surface of the tunnel boring machine's affected area is equipped with a sensor array consisting of multiple intelligent settlement monitoring nodes, with each node having a built-in triaxial accelerometer and a pore water pressure gauge.

[0007] Acquire the initial surface elevation data output by the sensor array and generate the first surface morphology grid corresponding to the affected area.

[0008] The local surface tilt vector at each monitoring node is determined based on the tilt angle change output by the triaxial accelerometer, and the effective soil stress is corrected by the water pressure change value output by the pore water pressure gauge to obtain the corrected tilt vector.

[0009] Based on the corrected tilt vectors of all monitoring nodes, the curvature of the first landform grid is reconstructed node by node to obtain the second landform grid.

[0010] The second surface morphology grid is spatially coupled with the preset shield tunneling parameters. The morphological deviation of the second surface morphology grid relative to the first surface morphology grid is extracted at the coupling point, and the morphological deviation is used as the ground settlement monitoring result.

[0011] As a technical solution of this invention, when generating the first landform grid, the planar coordinates of each monitoring node are used as interpolation base points, and radial basis functions are used to spatially interpolate the initial land surface elevation data to generate a regular grid covering the influence area. For the elevation value of each grid point in the regular grid, buoyancy correction is performed based on the pore water pressure gauge readings of its adjacent monitoring nodes, and the buoyancy-corrected elevation value is used as the node elevation of the first landform grid. This processing method enables the initial landform to more realistically reflect the actual elevation distribution under soil buoyancy, providing a reliable reference surface for subsequent settlement analysis.

[0012] Preferably, when determining the local surface tilt vector, the three components of gravitational acceleration from the triaxial accelerometer at each monitoring node during the stationary period are read, the pitch and roll angles of that node relative to the initial horizontal plane are calculated, and then the pitch and roll angles are converted into a three-dimensional unit tilt vector at that node as the local surface tilt vector. The tilt vector obtained in this way accurately characterizes minute attitude changes of the surface.

[0013] In the soil effective stress correction stage, the pore water pressure increment at the monitoring node is calculated based on the difference between the current water pressure value output by the pore water pressure gauge and the initial water pressure value. The decrease in effective soil stress at the node is determined by the pore water pressure increment, and the lateral deformation coefficient of the surface soil is calculated accordingly. The corrected tilt vector is obtained by multiplying the lateral deformation coefficient by the local surface tilt vector. By introducing the pore water pressure change to correct the tilt vector, the surface buoyancy tilt caused by excess pore water pressure is effectively eliminated, making the tilt characterization closer to the actual deformation direction of the soil skeleton.

[0014] During curvature reconstruction, each grid point in the first surface morphology grid is sequentially selected, and a local neighborhood window centered on that grid point is determined. The corrected tilt vectors of all monitoring nodes within the window are then obtained. The surface curvature tensor at that grid point is calculated based on the rate of change of the corrected tilt vectors within the window. The elevation value of that grid point is then updated using second-order surface fitting based on the surface curvature tensor. This process is repeated for all grid points to obtain the second surface morphology grid. This reconstruction method fully utilizes the spatial continuity of multi-node tilt information, enabling precise characterization of local surface bending deformation caused by shield tunneling and improving the accuracy of settling trough morphology capture.

[0015] Furthermore, the local neighborhood window is determined as follows: a circular area is delineated with the planar position of the grid point as the center and a radius that is an integer multiple of the preset shield tunnel diameter. This circular area is used as the local neighborhood window. Within the circular area, all deployed monitoring nodes are searched, and the corresponding corrected tilt vectors are extracted. When calculating the surface curvature tensor, the corrected tilt vector of each monitoring node within the window is decomposed into a first component along the radial direction and a second component along the tangential direction. Spatial difference is performed on the first component to obtain the radial curvature, and spatial difference is performed on the second component to obtain the tangential curvature. The components are combined to obtain the surface curvature tensor. This decomposition and difference operation independently extracts the radial and tangential bending features, making the reconstructed grid more accurately reflect the asymmetric deformation caused by shield tunneling.

[0016] During the spatial coupling phase, the real-time 3D coordinates and tunneling direction vector of the current shield machine cutterhead center are acquired. The tunneling influence strip area is delineated in the second surface morphology grid, and all grid points within this area are marked as coupling grid points. The planar coordinates of each coupling grid point are spatially registered with the real-time 3D coordinates of the shield machine cutterhead center. For each coupling grid point, its updated elevation value in the second surface morphology grid and its initial elevation value in the first surface morphology grid are read. The difference between the two is calculated as the single-point deviation at that coupling grid point. The single-point deviations of all coupling grid points are combined according to their spatial distribution to form the morphological deviation. This coupling method correlates surface morphology changes with the shield tunneling position in real time, quantitatively separating the range and degree of influence of each ring of tunneling on the surface.

[0017] When using morphological deviation as the monitoring result, the sign of the deviation distinguishes between surface uplift and subsidence, and the absolute value distinguishes the settlement level. Based on the preset shield tunneling ring number, the morphological deviation corresponding to each tunneling ring is stored as a ring-sequence deviation sequence in chronological order, forming a complete ground settlement monitoring result. This method realizes a fully automated processing chain from sensor acquisition, tilt correction, curvature reconstruction to tunneling coupling, continuously outputting the temporal and spatial distribution of surface deformation during tunneling without manual intervention, effectively improving the real-time performance and accuracy of ground settlement monitoring during shield tunneling construction.

[0018] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0019] After acquiring the local surface tilt vector at the monitoring node using a triaxial accelerometer, the effective stress of the soil is corrected based on the water pressure change value measured by a pore water pressure gauge. When the shield tunneling causes an increase in excess pore water pressure, the increase in pore water pressure reduces the effective stress of the soil, leading to lateral unloading deformation of the soil skeleton. This deformation is not a direct reflection of vertical settlement. By calculating the reduction in effective soil stress based on the increase in pore water pressure, the lateral deformation coefficient of the surface soil is determined. Multiplying this coefficient by the local surface tilt vector allows the removal of the spurious tilt component caused by pore water pressure fluctuations, resulting in a corrected tilt vector that only reflects the settlement characteristics of the soil skeleton. Using the corrected tilt vector instead of the original dip angle data for subsequent curvature reconstruction ensures that the generated surface morphology mesh truly reflects the bending caused by soil consolidation settlement, avoiding misjudgments or omissions of settlement due to changes in pore pressure in water-rich strata.

[0020] After obtaining the corrected tilt vectors at all monitoring nodes, the curvature of the first surface morphology grid is reconstructed node by node. A local neighborhood window is defined centered on each grid point, and the corrected tilt vectors of all monitoring nodes within the window are extracted. The rate of change of the tilt vector direction is calculated to obtain the surface curvature tensor at that grid point. The grid point elevation is updated by performing second-order surface fitting on this curvature tensor. After traversing all grid points, a second surface morphology grid with continuous curvature changes is obtained. This processing method does not rely on global interpolation smoothing, but directly restores the surface curvature state through the rate of change of the local tilt direction. It can reproduce the curvature abrupt change characteristics of the surface at differential settlement with high fidelity, making the locations of surface depressions, uplifts, and inflection points clearer. In the spatial location coupling stage, the real-time three-dimensional coordinates and tunneling direction vector of the current shield machine cutterhead center are obtained. The tunneling influence strip area is defined in the second surface morphology grid. The updated elevation values ​​of all coupled grid points in this area are subtracted from the initial elevation values ​​in the first surface morphology grid to obtain the morphological deviation at that tunneling location. This method precisely links the overall changes in the surface morphology with the location of the tunnel boring machine (TBM), and can directly separate the changes in the surface morphology induced by the current tunneling ring. It eliminates the interference of historical cumulative settlement or surface undulations caused by environmental factors far from the tunnel face, and obtains ground settlement monitoring results that are synchronized with the TBM's advancement. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0022] Figure 1 This is a flowchart of an automatic monitoring method for ground settlement during shield tunneling construction based on intelligent sensing.

[0023] Figure 2 This is a schematic diagram showing the change in the combined acceleration amplitude of the three-axis accelerometer and the stationary period.

[0024] Figure 3 It is an elevation comparison curve between the first and second landform grids;

[0025] Figure 4 This is a schematic diagram showing the spatial distribution of ground settlement levels in the area affected by shield tunneling. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] See Figure 1 This invention provides an automatic monitoring method for ground settlement during shield tunneling construction based on intelligent sensing, comprising: determining that a sensor array consisting of multiple sets of intelligent settlement monitoring nodes is deployed on the ground surface within the shield tunneling impact zone, each set of nodes having a built-in triaxial accelerometer and a pore water pressure gauge; acquiring initial ground elevation data output by the sensor array, and generating a first ground morphology grid corresponding to the impact zone based on the initial ground elevation data; determining the local ground tilt vector at each monitoring node based on the tilt angle change output by the triaxial accelerometer, and correcting the local ground tilt vector for effective soil stress using the water pressure change value output by the pore water pressure gauge to obtain a corrected tilt vector; reconstructing the curvature of the first ground morphology grid node by node based on the corrected tilt vectors of all monitoring nodes to obtain a second ground morphology grid; spatially coupling the second ground morphology grid with preset shield tunneling parameters, extracting the morphological deviation of the second ground morphology grid relative to the first ground morphology grid at the coupling position, and using the morphological deviation as the ground settlement monitoring result.

[0028] Example 1

[0029] In practical implementation, each intelligent settlement monitoring node in the sensor array has a corresponding planar coordinate and an initial elevation value given by the initial surface elevation data. The set of planar coordinates of all intelligent settlement monitoring nodes is obtained, denoted as... and the corresponding set of initial elevation values. ,in , This indicates the total number of intelligent settlement monitoring nodes.

[0030] Radial basis functions are used to spatially interpolate the initial surface elevation data, generating a regular grid covering the influence zone. The radial basis functions are chosen in a polyquadratic form, with the expression: In the formula This represents the output value of the radial basis function. Indicates spatial distance. Represents shape parameters. Shape parameters The specific value is set to 1.5 times the average spacing of all intelligent settlement monitoring nodes. The average spacing of intelligent settlement monitoring nodes is determined as follows: For each intelligent settlement monitoring node, the straight-line distance between that node and its nearest neighbor is calculated. The arithmetic mean of the nearest neighbor distances for all intelligent settlement monitoring nodes is then calculated, and the resulting average value is the average spacing of the intelligent settlement monitoring nodes. The shape parameter... The setting is 1.5 times the average spacing because the support range of the radial basis function at this multiple can completely cover the area between adjacent smart settlement monitoring nodes, so that the interpolation surface can smoothly transition between nodes, while suppressing the excessive diffusion effect of distant nodes on the interpolation results and maintaining the interpolation surface's ability to restore the local surface morphology.

[0031] Using the planar coordinates of each intelligent settlement monitoring node As the interpolation base point, for any grid point in the regular grid, its planar coordinates are marked as... The interpolated elevation of the grid point Determined by the following formula:

[0032]

[0033] in, Indicates the relationship with the first The weighting coefficients corresponding to each intelligent settlement monitoring node; Radial basis functions; This represents the total number of intelligent settlement monitoring nodes. For the first Planar coordinates of each intelligent settlement monitoring node. Weighting coefficients. The solution process is as follows: ensure that the above interpolation formula holds strictly at all intelligent settlement monitoring nodes, that is, for each intelligent settlement monitoring node... ,have This leads to the construction of a system of linear equations, the first of which is the first of the linear equations. The equations are ,in and They represent the first The and the first The planar position vector of each intelligent settlement monitoring node. Let represent the Euclidean distance between the position vectors of the two planes. Solving the above system of linear equations yields all the weight coefficients. Substitute the planar coordinates of all grid points into the formula to generate the complete regular grid interpolation elevation.

[0034] For the interpolated elevation of each grid point in the regular grid, buoyancy correction is performed based on the pore water pressure gauge readings of its adjacent smart settlement monitoring nodes. In specific implementation, for a given grid point, the set of smart settlement monitoring nodes adjacent to that grid point is first determined. The method for determining adjacent smart settlement monitoring nodes is as follows: a circular neighborhood is delineated with the grid point's planar coordinates as the center and twice the average spacing between smart settlement monitoring nodes as the radius. All smart settlement monitoring nodes falling within this circular neighborhood are recorded as adjacent smart settlement monitoring nodes.

[0035] The current pore water pressure value output by the pore water pressure gauge built into each adjacent intelligent settlement monitoring node is obtained, and the initial pore water pressure value recorded by the same pore water pressure gauge at the initial moment is also read. The current pore water pressure value is subtracted from the initial pore water pressure value to obtain the pore water pressure increment of that adjacent intelligent settlement monitoring node. The pore water pressure increment is divided by the unit weight of water to obtain the equivalent head change. The equivalent head change is multiplied by the soil buoyancy correction factor to obtain the buoyancy correction for a single node. The soil buoyancy correction factor is the reciprocal of the ratio of the effective unit weight of soil to the unit weight of water. The effective unit weight of soil is determined based on the soil physical and mechanical parameters given in the geotechnical engineering investigation report of the construction impact area, and the unit weight of water uses the standard value.

[0036] When multiple adjacent smart settlement monitoring nodes exist around a grid point, a single-node buoyancy correction is calculated for each adjacent smart settlement monitoring node. A distance-inverse weighted method is used to fuse all single-node buoyancy corrections to obtain a comprehensive buoyancy correction: the reciprocal of the Euclidean distance from the grid point's planar coordinates to the planar coordinates of each adjacent smart settlement monitoring node is used as the fusion weight. All single-node buoyancy corrections are weighted and averaged according to this weight, and the result is the comprehensive buoyancy correction. The comprehensive buoyancy correction is added to the original interpolated elevation of the grid point to obtain the buoyancy-corrected elevation value. This process is repeated for all grid points in the regular grid. After each grid point undergoes the above buoyancy correction operation, the buoyancy-corrected elevation values ​​of all grid points together constitute the node elevation of the first surface morphology grid.

[0037] Example 2

[0038] In practical implementation, each intelligent settlement monitoring node's built-in triaxial accelerometer continuously outputs acceleration measurements in three mutually orthogonal directions at fixed time intervals. The three sensing axes of the triaxial accelerometer are denoted as the sensor x-axis, sensor y-axis, and sensor z-axis, respectively, and the corresponding output acceleration components are denoted as... , and .

[0039] Stationary periods are identified in the acceleration time series output by the triaxial accelerometer. The change in the resultant acceleration amplitude at each sampling moment is calculated, and the absolute value of the difference between the resultant acceleration amplitude at the current sampling moment and the previous sampling moment is taken as the change in resultant acceleration amplitude. The resultant acceleration amplitude is determined by... Calculation. When the change in the combined acceleration amplitude at multiple consecutive sampling times is less than a preset amplitude change threshold, and the duration reaches a preset static duration threshold, the triaxial accelerometer is determined to be in a static period during that time interval. The amplitude change threshold is set to... , This represents the standard value of gravitational acceleration. The threshold for the duration of stillness is set to... The basis for setting the above amplitude change threshold is that, in the absence of large-scale disturbances, the change in acceleration amplitude caused by environmental micro-vibrations in the construction impact area is usually lower than [the threshold value]. Set the static duration threshold to ; The basis for the seconds is that the typical sampling frequency of a triaxial accelerometer can ensure that enough sample points for averaging are collected within that duration. The time span of seconds can filter out the interference of instantaneous impact vibrations on tilt angle estimation while ensuring real-time performance.

[0040] Within the determined static time period, the three components of gravitational acceleration from all sampling points output by the triaxial accelerometer are extracted. The average values ​​of the sensor's x-axis acceleration component set, y-axis acceleration component set, and z-axis acceleration component set are calculated respectively. These three average values ​​are denoted as follows: , and , which serves as the representative value of the three components of gravitational acceleration under the current attitude of the monitoring node.

[0041] Based on the representative values ​​of the three components of gravitational acceleration, the pitch and roll angles of the monitoring node relative to the initial horizontal plane are calculated. The pitch angle is defined as the angle by which the monitoring node rotates off the initial horizontal plane around the sensor's y-axis. and The arctangent of the ratio as the pitch angle The expression is The roll angle is defined as the angle by which the monitoring node rotates off the initial horizontal plane around the sensor's x-axis. and The arctangent of the ratio is used as the roll angle. The expression is The initial horizontal plane is determined by the horizontal reference plane initialized by each intelligent settlement monitoring node after installation and before it is disturbed by the tunnel boring machine. This initial horizontal plane corresponds to the state where the gravity direction coincides with the z-axis of the sensor.

[0042] From pitch angle and roll angle A three-dimensional unit tilt vector is constructed and used as the local surface tilt vector at the monitoring node. Determine by the following formula:

[0043]

[0044] in, Represents the three-dimensional unit tilt vector at the monitoring node; This indicates the current pitch angle of the monitoring node relative to the initial horizontal plane; This indicates the current roll angle of the monitoring node relative to the initial horizontal plane; and These represent the sine and cosine functions, respectively. This represents the transpose of a vector. The vectors mentioned above... The three components correspond to the direction cosines of the local surface tangent plane normal vector of the monitoring node in the spatial rectangular coordinate system, and the magnitude of this vector is constant. This represents the direction and degree of inclination of the ground surface. Each monitoring node independently calculates its corresponding local ground surface inclination vector, and all local ground surface inclination vectors constitute the input data for subsequent soil effective stress correction.

[0045] See Figure 2 In the graph, the horizontal axis represents time in seconds, and the vertical axis represents the change in the amplitude of the resultant acceleration from the three-axis accelerometer, in g (the standard value of gravitational acceleration). The curves represent the change in the amplitude of the resultant acceleration at the sampling time, reflecting the dynamic characteristics of the current acceleration change at the monitoring node. The red shaded areas in the graph indicate stationary periods. A stationary period is determined when the change in the amplitude of the resultant acceleration at multiple consecutive sampling times is less than a preset amplitude change threshold of 0.01g, and the duration reaches a stationary duration threshold of 0.5 seconds. The threshold of 0.01g is determined by the amplitude characteristics of environmental micro-vibrations, and the 0.5-second duration takes into account both the sampling frequency and the need to filter out instantaneous disturbances.

[0046] The curve exhibits multiple fluctuations. During the red-shaded static period, the change in the magnitude of the resultant acceleration remains consistently below the 0.01g threshold, indicating that the monitoring node is in a dynamically stable state and is suitable as an effective time window for extracting the gravitational acceleration component during static periods. During non-static periods, the change in the magnitude of the resultant acceleration fluctuates significantly, with the maximum value exceeding the 0.01g threshold, indicating the presence of transient disturbances or micro-vibrations on the ground. Between each static period, the curve shows several brief jumps in the amplitude peak value, reflecting the impact of the construction environment or surface changes on the sensor.

[0047] Example 3

[0048] In practice, for each intelligent settlement monitoring node, the water pressure measurement value output by the built-in pore water pressure gauge at the current moment is read and recorded as the current water pressure value. Simultaneously, the initial water pressure value of the same pore water pressure gauge is read when the sensor array is initially deployed and before the shield tunneling begins to affect it, and recorded as the initial water pressure value. Initial water pressure value The initial water pressure value is obtained during a stable period after the intelligent settlement monitoring node has been installed and debugged, but before the tunnel boring machine cutterhead has entered the plane projection range of the monitoring node. During this stable period, the pore water pressure gauge reading tends to be constant, and the average value of the time series output by the pore water pressure gauge during this period is taken as the initial water pressure value. .

[0049] Calculate the current water pressure value Compared with the initial water pressure value The difference is recorded as the increment of pore water pressure. The expression is Increment in pore water pressure A positive value indicates that the pore water pressure at the monitoring node has increased compared to the initial state, while a negative value indicates that the pore water pressure has decreased compared to the initial state.

[0050] Based on the increase in pore water pressure Determine the decrease in effective soil stress at the monitoring node. In the consolidation theory of saturated soil, under the condition that the total stress remains constant, the increase in pore water pressure equals the decrease in effective soil stress. Therefore, the increase in pore water pressure... As the effective stress reduction of soil ,satisfy .

[0051] Based on the reduction in effective soil stress Calculate the lateral deformation coefficient of the surface soil at the monitoring node. Lateral deformation coefficient Determined by the following formula:

[0052]

[0053] in, This represents the lateral deformation coefficient, which is a dimensionless quantity. Poisson's ratio represents the ratio of surface soil. The lateral confined compression modulus of surface soil, expressed in kilopascals (kPa). Poisson's ratio represents the decrease in effective stress in soil, expressed in kilopascals (kPa). The range of values ​​is to The specific value is determined based on the soil type given in the geotechnical engineering investigation report of the surface soil in the construction impact area. A larger value is taken for cohesive soil, and a smaller value is taken for sandy soil. Lateral confined compression modulus of surface soil. Also obtained from the geotechnical engineering investigation report, the recommended value of the lateral confined compression modulus of the surface soil layer is used. Lateral deformation coefficient. The coefficient of lateral deformation reflects the tendency of the surface soil to deform laterally under the action of a unit reduction in effective stress. The larger the Poisson's ratio or the smaller the compression modulus, the larger the coefficient of lateral deformation.

[0054] Lateral deformation coefficient Multiply by the local surface tilt vector This yields the corrected tilt vector. Corrected tilt vector The calculation method is as follows , where vector The local surface tilt vector at the same monitoring node is calculated based on the output of the triaxial accelerometer, with all three components of the vector multiplied by a scalar coefficient. .when A positive number indicates an increase in pore water pressure. A positive number represents the lateral deformation coefficient. A positive value indicates that the direction of the corrected tilt vector is consistent with the direction of the local surface tilt vector, meaning that the reduction in effective stress causes additional lateral deformation of the surface soil along the original tilt direction; when When the value is negative, it means that the pore water pressure is decreasing. Negative values ​​indicate the lateral deformation coefficient. A negative value indicates that the direction of the corrected tilt vector is opposite to that of the local surface tilt vector, signifying that the increase in effective stress causes the surface soil to undergo lateral constraint deformation in the opposite direction. The above calculation process is performed at each monitoring node to obtain the corrected tilt vector corresponding to each monitoring node.

[0055] Example 4

[0056] In practice, the set of all grid points in the first surface morphology grid is taken as the object to be processed, and the curvature reconstruction process is executed sequentially for each grid point.

[0057] For the currently selected grid point, a local neighborhood window centered on that grid point is determined. The local neighborhood window is defined as follows: using the planar coordinates of the selected grid point as the center and an integer multiple of the preset shield tunnel diameter as the radius, a circular area is delineated in the ground plane coordinate system. This circular area serves as the local neighborhood window. The shield tunnel diameter is taken as the design value of the outer diameter of the shield machine cutterhead, which is obtained from the shield construction design drawings. The integer multiple of the radius is set to... Multiples. Integer multiples are set to... The basis for this multiple is that the width of the transverse settlement trough caused by shield tunneling is approximately [a fraction of] the tunnel diameter. to Double, take Using the diameter of the shield tunnel as the radius of the local neighborhood window ensures that the circular area completely covers the influence range of a single settlement trough, guaranteeing that the local neighborhood window contains a sufficient number of monitoring nodes to correct the tilt vectors that characterize the surface curvature change features.

[0058] Search for all deployed smart settlement monitoring nodes within the defined circular area. The search method is as follows: traverse all smart settlement monitoring nodes in the sensor array, calculate the Euclidean distance between the plane coordinates of each smart settlement monitoring node and the plane coordinates of the selected grid points, and determine smart settlement monitoring nodes whose Euclidean distance is less than or equal to the radius of the circular area as nodes falling within the local neighborhood window. Extract the corrected tilt vector corresponding to each searched smart settlement monitoring node. This corrected tilt vector is a three-dimensional vector that has been processed by the effective soil stress correction, denoted as […]. Construct a set of corrected tilt vectors corresponding to each grid point by collecting all the searched corrected tilt vectors.

[0059] For each corrected tilt vector in the set of corrected tilt vectors within the local neighborhood window, radial and tangential components are decomposed. The decomposition process is as follows: For a given corrected tilt vector in the set of corrected tilt vectors, the azimuth vector of the intelligent settlement monitoring node to which the corrected tilt vector belongs is calculated relative to the selected grid point. The direction of the azimuth vector points from the plane coordinates of the grid point to the plane coordinates of the intelligent settlement monitoring node, and the magnitude of the azimuth vector is the Euclidean distance between the two points. The azimuth vector is then normalized to obtain the radial unit vector. In a two-dimensional Earth's surface plane, the radial unit vector... Rotate counterclockwise Degrees, to obtain the tangential unit vector Correcting the tilt vector It is itself a three-dimensional vector, and the corrected skew vector is taken. The projection vector in the horizontal plane, and the radial unit vector are respectively compared with the projection vector. and tangential unit vector Perform the inner product operation to obtain the first component along the radial direction of the circular region. and the second component tangential to the circular region First component The calculation method is based on the projection vector and the radial unit vector. The inner product, the second component The calculation method is based on the projection vector and the tangential unit vector. The inner product. The above decomposition operation is performed on each corrected tilt vector in the corrected tilt vector set to obtain the first component set and the second component set.

[0060] Spatial difference is performed on the first component of all monitoring nodes to obtain the radial curvature. The spatial difference uses a central difference scheme, specifically: within a local neighborhood window, pairs of intelligent settlement monitoring nodes that are radially adjacent are selected. For each pair of nodes, the difference in radial position coordinates between the two nodes and the difference in their corresponding first components are calculated. The difference in the first components is divided by the difference in radial position coordinates to obtain a radial rate of change. The arithmetic mean of the radial rates of change obtained for all eligible node pairs within the local neighborhood window is taken as the radial curvature of the selected grid point, denoted as . .

[0061] Spatial difference is performed on the second component of all monitoring nodes to obtain the tangential curvature. The spatial difference uses a central difference scheme, specifically: within a local neighborhood window, pairs of intelligent settlement monitoring nodes that are mutually adjacent in the tangential direction are selected. For each pair of nodes, the difference in the position coordinates of the two nodes in the tangential direction and the difference in the corresponding second components are calculated. The difference in the second components is divided by the difference in the tangential position coordinates to obtain a tangential direction change rate. The arithmetic mean of the tangential direction change rates obtained for all node pairs meeting the conditions within the local neighborhood window is taken as the tangential curvature of the selected grid point, denoted as . .

[0062] radial curvature and tangential curvature The combination is the surface curvature tensor. Surface curvature tensor Represented in two-dimensional diagonal matrix form:

[0063]

[0064] in, The surface curvature tensor of the selected grid points is a single... OK A matrix of columns; Represents radial curvature, and is the element in the first row and first column of the surface curvature tensor matrix; Represents tangential curvature, which is the element in the second row and second column of the surface curvature tensor matrix.

[0065] According to the surface curvature tensor The elevation values ​​of selected grid points in the first surface morphology grid are updated using second-order surface fitting. The second-order surface fitting update process is as follows: a local planar coordinate system is established with the selected grid points as the origin. The horizontal and vertical axes of the local planar coordinate system are aligned with the radial and tangential directions of the circular region, respectively. In this local planar coordinate system, the updated local surface is assumed to be... It has the following form:

[0066]

[0067] in, and These represent the radial and tangential coordinates of any point in the local planar coordinate system, respectively. This represents the updated surface elevation value at that coordinate point; Indicates the coefficient of the constant term to be determined; This represents the undetermined radial first-order coefficient; These represent the undetermined first-order tangential coefficients. Undetermined coefficients , and Solving using the least squares method, the constraint is: at the planar coordinate position of each intelligent settlement monitoring node in the modified tilt vector set, the surface... Horizontal gradient and corrected tilt vector The horizontal projections remain consistent. After obtaining all the undetermined coefficients, let... and ,at this time The value is the updated elevation value of the selected grid point. The original elevation value of the selected grid point in the first surface morphology grid is replaced with the updated elevation value.

[0068] Traverse all grid points in the first landform grid, and perform all the operations described above for each grid point, including local neighborhood window delineation, modified tilt vector extraction, radial and tangential component decomposition, radial and tangential curvature calculation, land surface curvature tensor combination, and second-order surface fitting update. The grid data after all grid point elevation values ​​have been updated is the second landform grid.

[0069] See Figure 3In the figure, the horizontal axis represents the distance on the ground plane, in meters, ranging from approximately -30 meters to 30 meters, covering the typical spatial scale within the shield tunneling impact zone. The vertical axis represents the ground elevation value at the corresponding location, in millimeters. The two curves in the figure represent the elevation of the first surface morphology grid described in Example 4 (solid line) and the elevation of the second surface morphology grid obtained after curvature reconstruction (dashed line), respectively.

[0070] Judging from the curve trends, the elevation curve of the first landform grid is relatively smooth, showing general surface undulation characteristics, with local elevation values ​​fluctuating between approximately -1 mm and 0.6 mm. After curvature reconstruction based on local neighborhoods, the elevation curve of the second landform grid shows a greater overall elevation fluctuation, especially in the ranges of -20 m to -10 m and 0 m to 10 m, where more obvious elevation depressions and uplifts appear, with the lowest point reaching approximately -1.6 mm, significantly lower than the corresponding elevation value of the first landform grid. This reflects the improved ability of curvature reconstruction to depict the detailed features of surface subsidence.

[0071] The figure clearly illustrates the entire process in Example 4, which involves defining a local neighborhood window with a radius of 6 times the diameter of the shield tunnel, extracting a set of locally corrected tilt vectors, performing radial and tangential decomposition, calculating the surface curvature tensor, and combining it with second-order surface fitting to update the elevation of the first surface morphology grid point by point.

[0072] Example 5

[0073] In practice, the real-time three-dimensional coordinates and tunneling direction vector of the tunnel boring machine (TBM) cutterhead center are obtained. These coordinates are read from the TBM's built-in guidance system, which outputs the three-dimensional spatial coordinates of the cutterhead center in either the total station coordinate system or the construction coordinate system at a fixed frequency. and the unit vector in the tunneling direction Unit vector in the direction of tunneling The direction is that the center of the cutterhead points in the direction of tunneling.

[0074] Based on real-time 3D coordinates and the tunneling direction vector, the tunneling influence strip area is delineated in the second surface morphology grid. The delineation method for the tunneling influence strip area is as follows: the real-time 3D coordinates of the tunnel boring machine cutterhead center are projected onto the ground plane to obtain the coordinates of the projection point of the cutterhead center on the ground plane. Using this projection point as a reference, the vector along the tunneling direction... The projection direction and its reverse direction onto the ground plane are extended forward and backward by a predetermined distance, forming a rectangular strip with its centerline along the tunneling direction. The width of the rectangular strip is perpendicular to the projection direction of the tunneling direction vector onto the ground plane. The width of the rectangular strip is taken as a fraction of the diameter of the shield tunnel. The length of the rectangular strip extending forward is taken as a multiple of the diameter of the shield tunnel. The length extending backward is taken as a multiple of the diameter of the shield tunnel. The diameter of the shield tunnel is taken as the design value of the outer diameter of the shield machine cutterhead, obtained from the shield construction design drawings. The width of the tunneling influence strip area is set as the shield tunnel diameter. The basis for this multiple is that the width of the transverse settlement troughs on the ground surface caused by shield tunneling is usually concentrated on both sides of the tunnel axis. Within a range of twice the tunnel diameter, take A tunnel diameter that is twice the diameter of the shield tunnel can completely cover the boundary effect area beyond the settlement trough. The tunneling influence zone extends forward by the diameter of the shield tunnel. The basis for the multiple is in front of the cutter head. Within a radius equal to the tunnel diameter, the soil is subjected to compression and unloading disturbance during cutterhead excavation, resulting in premature surface heave or subsidence. Extending the shield tunnel diameter backward... The basis for the multiple is the area behind the cutter head. The soil within a radius equal to the tunnel diameter has detached from the shield tail and the grouting has not yet fully solidified, making it the main area for settlement development.

[0075] All grid points within the tunneling influence strip area are marked as coupled grid points. The marking method is as follows: traverse all grid points in the second surface morphology grid, calculate the planar coordinates of each grid point, and determine whether the planar coordinates fall within the rectangular area of ​​the tunneling influence strip. Grid points falling within the rectangular area are marked as coupled grid points. The planar coordinates of each coupled grid point are then spatially registered with the real-time 3D coordinates of the tunnel boring machine cutterhead center. The spatial registration method is as follows: record the planar coordinates of each marked coupled grid point. ,in and These represent the row and column indices of the coupled grid points in the second surface morphology grid, respectively, and simultaneously record the real-time three-dimensional coordinates of the tunnel boring machine cutterhead center at the current moment. In addition to the shield tunneling ring number, establish the correspondence between the spatial index of the coupled grid points and the position of the shield machine cutterhead to complete the spatial position coupling.

[0076] Within the tunneling influence zone, for each coupled grid point, the updated elevation value of the coupled grid point in the second surface morphology grid is read and recorded as follows. Read the initial elevation values ​​of the coupled grid points in the first surface morphology grid, and record them as follows: Calculate the difference between the updated elevation value and the initial elevation value, and use this difference as the single-point deviation at the coupled grid point. Single-point deviation. The formula for calculation is:

[0077]

[0078] in, Indicates row index is Column index is The single-point deviation of the coupled grid points, in millimeters; Indicates row index is Column index is The updated elevation values ​​of coupled grid points in the second surface morphology grid, in millimeters; Indicates row index is Column index is The initial elevation values ​​of the coupled grid points in the first surface morphology grid, in millimeters.

[0079] The individual point deviations of all coupled grid points are combined according to their spatial distribution to form the morphological deviation. The method of combining according to spatial distribution is as follows: the spatial index of each coupled grid point is maintained. and Keep all single-point deviations unchanged. By row index and column indexes The natural order of the elements forms a two-dimensional morphological deviation matrix. Each element of the two-dimensional morphological deviation matrix corresponds to the single-point deviation of a coupled grid point. The row and column directions of the matrix correspond to the grid arrangement order of the tunneling direction and its orthogonal direction in the ground plane, respectively.

[0080] The sign of the morphological deviation is used to distinguish between surface uplift and surface subsidence. The single-point deviation of a certain coupled grid point in the morphological deviation is considered... When the value is positive, it indicates that surface uplift has occurred at that coupled grid point; when the single-point deviation of a coupled grid point is positive... A negative value indicates that ground subsidence has occurred at that coupled grid point.

[0081] Settlement levels are distinguished by the absolute value of the morphological deviation. Settlement levels are divided into three ranges: absolute value less than... The range of millimeters indicates a slight deformation level; absolute values ​​greater than or equal to millimeters and less than The range of millimeters indicates a medium deformation level; absolute values ​​greater than or equal to The millimeter range indicates a severe deformation level. The threshold values ​​for the aforementioned settlement level ranges are based on the general settings for surface settlement monitoring control values ​​in the "Technical Specification for Monitoring of Urban Rail Transit Engineering," which specifies cumulative settlement control values ​​for shield tunneling construction. Millimeters, the warning value is taken from the control value. Right now millimeters, taken here millimeters and Millimeters serve as the grading threshold, distinguishing between safe, alert, and dangerous states.

[0082] Based on the preset shield tunneling ring numbers, the morphological deviations corresponding to each tunneling ring are stored as a ring-sequence deviation sequence in chronological order. The shield tunneling ring number is the sequence number of the segment ring counted from the start of shield tunneling; each tunneling ring corresponds to a unique ring number. Whenever the shield machine completes the advancement of a tunneling ring, the morphological deviation matrix corresponding to all coupled grid points under the current tunneling ring number is packaged with the tunneling ring number and the recording time into a single data record. The data records corresponding to each tunneling ring number are arranged in ascending order of the tunneling ring number to form the ring-sequence deviation sequence. The ring-sequence deviation sequence is output as the ground settlement monitoring result, which includes the spatial settlement distribution information of each tunneling ring and the settlement development information over time.

[0083] See Figure 4 The figure illustrates the spatial distribution of morphological deviations within the tunneling influence zone obtained from the automatic ground settlement monitoring method for shield tunneling construction in Example 5. The horizontal axis represents the X-coordinate (unit: meters) on the ground plane, and the vertical axis represents the Y-coordinate (unit: meters) on the ground plane, covering a two-dimensional matrix grid range of the shield tunneling area. In the legend, green dots represent areas of slight deformation, corresponding to absolute settlement deviations of less than 10 mm; orange dots represent areas of moderate deformation, corresponding to absolute settlement deviations between 10 and 30 mm; and red dots represent areas of severe deformation, corresponding to absolute settlement deviations greater than or equal to 30 mm.

[0084] As shown in the image, the red severely deformed points are mainly concentrated in the central area, indicating that the surface subsidence in this area has reached the severe deformation level, with a subsidence value exceeding 30 mm. This reflects the surface subsidence caused by the tunnel boring machine in the core area of ​​the strip-shaped influence zone. The orange moderately deformed areas are distributed in a ring around the central red area, with a relatively wide spatial range, indicating that the surface subsidence on the periphery of the tunneling influence zone has reached the moderate deformation level, with a subsidence value between 10 and 30 mm. The green slightly deformed points are mainly distributed around the edges of the image and are relatively scattered, indicating that the surface subsidence is smaller and at the slightly deformed level in areas farther from the core tunneling area, with a subsidence value less than 10 mm.

[0085] This spatial distribution characteristic is consistent with the setting in Example 5, where the width of the tunneling influence strip area is determined to be four times the diameter of the shield tunnel, and the morphological deviation within the strip is used to distinguish settlement levels by absolute values, reflecting the spatial attenuation law of the tunneling influence. The trend of settlement levels gradually decreasing from the center to the outside in the figure indicates that the surface settlement caused by shield tunneling is significantly concentrated in the cutterhead tunneling path and its surrounding area, and the settlement level classification threshold effectively distinguishes the surface response areas with different degrees of deformation.

[0086] The above description is merely a specific embodiment of this application, but the scope of protection of this application 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 this application should be included within the scope of protection of this application.

Claims

1. An automatic monitoring method for ground settlement during shield tunneling construction based on intelligent sensing, characterized in that, include: The surface of the tunnel boring machine's impact zone is equipped with a sensor array consisting of multiple sets of intelligent settlement monitoring nodes. Each set of nodes is equipped with a triaxial accelerometer and a pore water pressure gauge. Acquire the initial surface elevation data output by the sensor array, and generate the first surface morphology grid corresponding to the affected area based on the initial surface elevation data; Based on the tilt angle change output by the triaxial accelerometer, the local surface tilt vector at each monitoring node is determined, and the effective soil stress is corrected by using the water pressure change value output by the pore water pressure gauge to obtain the corrected tilt vector. Based on the corrected tilt vectors of all monitoring nodes, the first landform grid is reconstructed node by node to obtain the second landform grid; The second surface morphology grid is spatially coupled with the preset shield tunneling parameters. The morphological deviation of the second surface morphology grid relative to the first surface morphology grid is extracted at the coupling position, and the morphological deviation is used as the ground settlement monitoring result.

2. The automatic monitoring method for ground settlement during shield tunneling based on intelligent sensing according to claim 1, characterized in that, The generation of the first landform grid corresponding to the affected area based on the initial surface elevation data specifically includes: Using the planar coordinates of each monitoring node as the interpolation base point, the initial surface elevation data is spatially interpolated using radial basis functions to generate a regular grid covering the affected area; For the elevation value of each grid point in the regular grid, buoyancy correction is performed based on the pore water pressure gauge readings of the adjacent monitoring nodes, and the buoyancy-corrected elevation value is used as the node elevation of the first surface morphology grid.

3. The automatic monitoring method for ground settlement during shield tunneling based on intelligent sensing according to claim 1, characterized in that, Based on the tilt angle change output by the triaxial accelerometer, the local surface tilt vector at each monitoring node is determined, specifically including: Read the three components of gravitational acceleration from the triaxial accelerometer at each monitoring node during the stationary period, and calculate the pitch and roll angles of that node relative to the initial horizontal plane based on the three components of gravitational acceleration. The pitch angle and the roll angle are converted into a three-dimensional unit tilt vector at the node, and the three-dimensional unit tilt vector is used as the local surface tilt vector.

4. The automatic monitoring method for ground settlement during shield tunneling based on intelligent sensing according to claim 1, characterized in that, Using the water pressure change value output by the pore water pressure gauge, the effective soil stress is corrected on the local surface tilt vector to obtain the corrected tilt vector, specifically including: The pore water pressure increment at the monitoring node is calculated based on the difference between the current water pressure value output by the pore water pressure gauge and the initial water pressure value. The reduction in effective stress of the soil at the node is determined based on the increase in pore water pressure, and the lateral deformation coefficient of the surface soil at the node is calculated based on the reduction. The corrected tilt vector is obtained by multiplying the lateral deformation coefficient by the local surface tilt vector.

5. The automatic monitoring method for ground settlement during shield tunneling based on intelligent sensing according to claim 1, characterized in that, Based on the corrected tilt vectors of all monitoring nodes, the first landform grid is reconstructed node-by-node curvature to obtain the second landform grid, specifically including: Each grid point in the first landform grid is selected sequentially, a local neighborhood window centered on the grid point is determined, and the corrected tilt vector of all monitoring nodes within the local neighborhood window is obtained. Calculate the surface curvature tensor at the grid point based on the rate of change of direction of all corrected tilt vectors within the local neighborhood window; Based on the surface curvature tensor, the elevation value of the grid point in the first surface morphology grid is updated by second-order surface fitting, and the second surface morphology grid is obtained after traversing all grid points.

6. The automatic monitoring method for ground settlement during shield tunneling based on intelligent sensing according to claim 5, characterized in that, Determine a local neighborhood window centered on the grid point, and obtain the corrected tilt vector of all monitoring nodes within the local neighborhood window, specifically including: Using the planar position of the grid point as the center, a circular area is delineated with a radius that is an integer multiple of the preset shield tunnel diameter, and the circular area is used as the local neighborhood window; Search for all deployed monitoring nodes within the circular area and extract the corrected tilt vector corresponding to each found monitoring node.

7. The automatic monitoring method for ground settlement during shield tunneling based on intelligent sensing according to claim 6, characterized in that, Based on the rate of change of direction of all corrected tilt vectors within the local neighborhood window, the surface curvature tensor at the grid point is calculated, specifically including: The corrected tilt vector of each monitoring node within the local neighborhood window is decomposed into a first component along the radial direction of the circular region and a second component along the tangential direction of the circular region. Spatial difference is performed on the first component of all monitoring nodes to obtain radial curvature, and spatial difference is performed on the second component of all monitoring nodes to obtain tangential curvature. The radial curvature and the tangential curvature are combined to obtain the surface curvature tensor.

8. The automatic monitoring method for ground settlement during shield tunneling based on intelligent sensing according to claim 1, characterized in that, The second surface morphology grid is spatially coupled with preset shield tunneling parameters, specifically including: Obtain the real-time three-dimensional coordinates and tunneling direction vector of the current shield machine cutterhead center, and delineate the tunneling influence strip area in the second surface morphology grid based on the real-time three-dimensional coordinates and the tunneling direction vector; All grid points within the tunneling influence strip area are marked as coupled grid points, and the planar coordinates of each coupled grid point are spatially registered with the real-time three-dimensional coordinates of the shield machine cutterhead center to complete the spatial position coupling.

9. The automatic monitoring method for ground settlement during shield tunneling based on intelligent sensing according to claim 8, characterized in that, Extracting the morphological deviation of the second landform grid relative to the first landform grid at the coupling location, specifically including: Within the tunneling influence strip area, for each of the coupled grid points, the updated elevation value of the grid point in the second surface morphology grid and the initial elevation value of the grid point in the first surface morphology grid are read. The difference between the updated elevation value and the initial elevation value is calculated as the single-point deviation at the coupled grid point. The single-point deviations of all coupled grid points are combined according to their spatial distribution to form the morphological deviation.

10. The automatic monitoring method for ground settlement during shield tunneling based on intelligent sensing according to claim 1, characterized in that, The aforementioned morphological deviation is used as a result of ground subsidence monitoring, specifically including: The sign of the morphological deviation is used to distinguish between surface uplift and surface subsidence, and the absolute value of the morphological deviation is used to distinguish the subsidence level. According to the preset shield tunneling ring number, the morphological deviation amount corresponding to each tunneling ring is stored as a ring sequence deviation sequence in chronological order, and the ring sequence deviation sequence is output as the ground settlement monitoring result.

Images