Shield tunnel full-life structure deformation monitoring and early warning method based on digital twinning

Abstract

The invention discloses a shield tunnel full-life structure deformation monitoring and early warning method based on digital twinning, and the method comprises the following steps: S1, constructing a tunnel-geological complex model fusing BIM and three-dimensional point cloud technology according to the point cloud data of a tunnel completion section; s2, FBG and incremental sensing are adopted, and full-life-cycle deformation real-time monitoring is conducted on the tunnel; s3, constructing an LGBM proxy model, and updating the deformation prediction of the tunnel structure in the construction period; s4, constructing a GAT-LSTM model, and predicting structural deformation of the tunnel in the service period; s5, constructing a tunnel structure deformation grading early warning method based on a strength reduction and sharp point mutation theory; and S6, constructing a tunnel monitoring and early warning system based on digital twinning. The tunnel-geological complex structure reconstruction is realized by fusing stratum information and high-precision spatial data on the basis of a digital twinning technology, the tunnel construction and service period real-time deformation monitoring and early warning method and the twinning system are provided, and the service safety and failure risk pre-warning of the tunnel structure in the whole life cycle are guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel engineering, and in particular to a full-life structural deformation monitoring and early warning method for a shield tunnel based on digital twins. Background Art

[0002] With the continuous development of transportation construction, the number of shield tunnel construction and capital investment have increased significantly. As a key node connecting the modern transportation network, shield tunnels undertake important tasks such as vehicle traffic and personnel safety. Their safety issues throughout their life cycle need urgent attention. However, their closed and special environmental characteristics bring huge challenges to real-time deformation monitoring of tunnels. Therefore, it is necessary to conduct real-time monitoring and risk warning of the deformation status of shield tunnels throughout their construction and service life cycle.

[0003] Currently, there is a lack of effective real-time monitoring and risk warning solutions for shield tunnel deformation. Traditional manual monitoring and experience-based warning methods are difficult to achieve real-time monitoring and are prone to misjudgment.

[0004] Chinese invention patent: Announcement number is "CN116128290A", and the name is "Digital Twin System for Intelligent Assessment of Environmental Risks of Shield Construction in Urban Built-up Areas", which discloses a digital twin system for intelligent assessment of environmental risks of shield construction in urban built-up areas, including: environmental information perception module, construction information perception module, construction impact information perception module, digital twin simulation module, risk loss evaluation module, and risk assessment and countermeasures module. The present invention can improve the scientific nature of environmental risk assessment, realize the visualization and quantification of risk assessment, and make real-time decisions for different risk levels, reducing the safety risk problems caused by damage to existing buildings and structures during construction. This technical solution is an assessment of the risk level of the impact of the early construction of the shield tunnel on the surrounding environment, and does not consider the assessment of the deformation of the shield tunnel throughout its life cycle after the construction is completed. Summary of the invention

[0005] In order to solve the problem that there is a lack of effective real-time monitoring and risk warning schemes for the deformation of shield tunnels in the above-mentioned prior art, and that traditional manual monitoring and empirical warning methods are difficult to achieve real-time monitoring and are prone to misjudgment, the present invention proposes a full-life structural deformation monitoring and early warning method for shield tunnels based on digital twins. By constructing a tunnel-geological complex structural model, the deformation data and construction status data of the tunnel throughout its life cycle are collected and transmitted in real time, and the full-life space deformation prediction model and hierarchical early warning mechanism of the shield tunnel throughout its life cycle are studied.

[0006] The present invention is achieved by the following technical solution: comprising the following steps:

[0007] S1. Based on the point cloud data of the completed tunnel section, a tunnel-geology complex model integrating BIM and 3D point cloud technology is constructed;

[0008] S2, using FBG and incremental sensing to monitor tunnel deformation in real time throughout its life cycle;

[0009] S3, construct the LGBM proxy model to update the tunnel structure deformation prediction during the construction period;

[0010] S4, construct the GAT-LSTM model to predict the structural deformation of the tunnel during service;

[0011] S5. Construct a tunnel structure deformation classification early warning method based on strength reduction and cusp catastrophe theory;

[0012] S6. Build a tunnel monitoring and early warning system based on digital twins.

[0013] As a further preferred embodiment, the specific steps of step S1 are as follows:

[0014] S11. Obtaining point cloud data of the completed section of the tunnel according to the vehicle-mounted laser detection system;

[0015] S111, a laser radar scanner is set above the inspection vehicle, and a local coordinate system with the center of the laser radar scanner as the origin is established based on the laser radar scanner above the inspection vehicle to obtain the three-dimensional point cloud data P of the tunnel in the local coordinate system local ;

[0016] S112, installing a GPS on the inspection vehicle, and obtaining the position coordinates T of the vehicle body in the global coordinate system according to the GPS on the inspection vehicle;

[0017] S113, installing an inertial navigation system on the inspection vehicle to keep time synchronization with the data of the laser radar scanner;

[0018] S114, setting an inclinometer inside the inspection vehicle to obtain accurate three-axis attitude angle data and construct a rotation matrix R;

[0019] S115, based on the three-dimensional point cloud data P of the tunnel in the local coordinate system obtained in step S111 local , the position coordinates T of the vehicle body obtained in step S112 in the global coordinate system and the rotation matrix R obtained in step S114, to obtain accurate three-dimensional point cloud coordinate information, that is, the three-dimensional point cloud data P of the tunnel in the global coordinate system global , the formula is as follows:

[0020] P global =R·P local +T

[0021] S12. Reconstruct the shield tunnel structure according to the point cloud data of the completed tunnel section obtained in step S11.

[0022] S121, point cloud data preprocessing;

[0023] S122, constructing a triangulation network of flat tunnel point cloud data based on the Delaunay triangulation algorithm, and reconstructing the main contour of the tunnel by combining the three-dimensional point cloud data of the tunnel;

[0024] S123, projecting the point cloud data onto the yox and zox planes respectively, constructing points on the central axis by using boundary functions obtained by fitting boundary points on both sides of the projected point cloud data, and using a cubic polynomial as a fitting function to obtain fitting equations of the tunnel axis in each plane;

[0025] S124, unfolding the point cloud data along the tunnel axis and determining the pixel value of the point based on its geometric features, creating a binary image, using the Laplace operator to detect the boundary of the binary image, dividing it into a single ring lining along the ring joint based on morphological processing of the boundary image, setting the same pixel value area to the same value through connected component analysis and dividing it into different segments, obtaining the center point and radius parameters by fitting the segment point cloud data, developing the as-built model of the lining ring using the CATIA platform, importing the fitting parameters of the center point and radius of different segments, the tunnel axis, and the material properties as input parameters into the platform, establishing a geometric size constraint relationship, creating a parameterized lining ring model, and thus reconstructing the shield tunnel structure;

[0026] S13, adaptively aligning the geological environment modeling and the shield tunnel structure reconstructed in step S12 to construct a tunnel-geology complex model.

[0027] As a further preferred embodiment, the specific steps of step S2 are as follows:

[0028] S21. Fiber Bragg grating multi-point displacement meters are arranged longitudinally on each deformation monitoring section of the shield tunnel to obtain the longitudinal deformation U of the tunnel during construction. l_c and the longitudinal deformation during service life U l_s and transmit it to the receiving instrument;

[0029] S22, on each deformation monitoring section of the shield tunnel, a pulley, a plurality of data acquisition modules with incremental encoders and a weight are sequentially arranged along the tunnel circumferential direction, wherein the pulley, the plurality of data acquisition modules with incremental encoders and the weight are connected with flexible wires to form a non-closed polygonal convergence monitoring system;

[0030] S221, respectively arrange data acquisition modules with incremental encoders at the arch crown, the arch waists on both sides and the middle position of the tunnel deformation monitoring section, and the arrangement points of the data acquisition modules with incremental encoders are named P1, P2, ..., Pn , P1 point is fixed with a pulley, P n Weights are attached at points to maintain wire tension, with the pulley acting as a ball-and-socket joint;

[0031] S222, use the total station to obtain the precise initial position coordinates of point P1, record the initial calibration values ​​of the angle and wire length at each point, and use trigonometric functions to obtain P2, P3, ..., P n The initial position coordinates of the point;

[0032] S223, a hollow shaft magnetic sensor A is set on the left side of each group of data acquisition modules with incremental encoders, and a hollow shaft magnetic sensor B is set on the right side. The hollow shaft magnetic sensor A is used to monitor the angle increment of the left side wire of each point between time t and time t-1, and the hollow shaft magnetic sensor B is used to monitor the angle increment of the right side wire of each point between time t and time t-1. Combined with the initial calibration value obtained in step S222, P1, P2, ..., P n The angle value at the point at time t and The formula is as follows:

[0033]

[0034]

[0035]

[0036] Where: i is the sequence number of the data acquisition module layout point, ranging from 1 to n; P is the time t i The angle value of the conductors on both sides of the point; P is the time t i The angle between the conductor on the right side of the point and the vertical direction; P is measured by the hollow shaft magnetic sensor A i The angle increment of the wire on the left side of the point between time t and time t-1; P is measured by the hollow shaft magnetic sensor B i The angle increment of the wire on the right side of the point between time t and time t-1;

[0037] S224, a wire support pulley and a solid shaft sensor C are set at the connection between two flexible wires inside each group of data acquisition modules with incremental encoders, and the angle increment Δδ of the wire support pulley is obtained by using the solid shaft sensor C. i-1 Point and P i Initial length of the wire between points Get the length of each wire segment at time t The formula is as follows:

[0038]

[0039] Where: P is the time t i Point and P i-1 The length of the wire between the points; is the initial time P i Point and P i-1 The length of the wire between the points; r is the radius of the wire support pulley; and At time t, P i-1 Point and P i The angle increment of the wire support pulley in the point position;

[0040] S225: The angle value at time t obtained in step S223 and the wire length value at time t obtained in step S224 Using trigonometric functions, we can get P2, P3, ..., P n The position coordinates of the point at time t, and finally the tunnel convergence U at different times during the construction period are obtained according to the changes in the coordinate information of each point. c_c 、Vault sinking s_c and the tunnel convergence U at different times during the service life c_s 、Vault sinking s_s .

[0041] As a further preferred embodiment, the specific steps of step S3 are as follows:

[0042] S31. Several pressure monitoring sections are evenly arranged in the longitudinal direction of the shield machine. Several pressure monitoring points are arranged in each pressure monitoring section. Micro weighing sensors are installed outside the shield machine protective cover at each pressure monitoring point. The micro weighing sensors are connected to the data recorder using a shielded data cable to measure the surrounding rock contact pressure F acting on each pressure monitoring point of the shield machine at time t. mn ; Wireless acceleration sensors and strain gauge torque sensors are deployed on different tool installation radii on the back of the cutterhead, and data is transmitted to the server through a wireless gateway near the cutterhead to obtain the average cutterhead speed N and average cutterhead torque T; a linear displacement sensor is installed on the shield machine shell to obtain the shield advancement speed V, thereby obtaining the distance x between different deformation monitoring surfaces and the excavation surface; a pressure sensor and a flow meter are installed in the shield machine grouting system to obtain the grouting pressure P and grouting volume Q;

[0043] S32, the surrounding rock contact pressure F obtained in step S31 mn, average cutterhead speed N, average cutterhead torque T, shield advancement speed V, tunnel burial depth H, distance x between deformation monitoring surface and excavation surface, stratum internal friction angle φ and cohesion c are used as the input of LGBM model. The stratum parameters need to be weighted considering the position and thickness of the overlying stratum of the tunnel. The longitudinal deformation U of the tunnel during construction period is used as the input of LGBM model. l_c 、Peripheral Convergence U c_c and the vault sinks s_c As the output of the LGBM model, the LGBM proxy model is established;

[0044] S33, preprocess the input and output data of the LGBM proxy model and divide them into test set and training set. The formula is as follows:

[0045]

[0046] Where: x i is the original data; max x i and min x i are the minimum and maximum values ​​in the data points respectively; by default, x p and x q 1 and -1 respectively;

[0047] S34, according to the Bayesian optimization method based on Gaussian process, the longitudinal deformation U corresponding to the tunnel construction period is obtained respectively. l_c 、Peripheral Convergence U c_c U with the vault sinking s_c The hyperparameters for the best prediction effect, the optimization process formula is as follows:

[0048]

[0049] R 1:t ={(x1,y1),(x2,y2)…(x t ,y t )}

[0050] y t =f(x t )+ε t

[0051] f(x t )~GP(m(x),k(x,x′))

[0052] Where: f is the objective function; y t is the observed value at the tth step; x t is the hyperparameter of the tth step; ε t is the observation error; R 1:t is the aggregate of observations from the first t steps; p(R 1:t |f) is R under a given objective function1:t The likelihood distribution of f; p(f) is the prior distribution of f; p(f|R 1:t ) is the observed R 1:t The posterior distribution of f; m(x) is the mean function; k(x,x') is the covariance function;

[0053] S35. Use the hyperparameters with the best prediction effect obtained in step S34 and the test set and training set divided in step S33 to train and evaluate the LGBM proxy model established in step S32. Use the trained model to predict the deformation results of the unknown sections of the tunnel under this operating state based on the burial depth, stratum parameters and distance from the excavation face of the unknown sections of the tunnel; collect dynamic monitoring data during the shield construction process, perform incremental training on the model and update the prediction of the tunnel deformation state.

[0054] As a further preferred embodiment, the specific steps of step S4 are as follows:

[0055] S41, the service life longitudinal deformation U of the s sections before and after the target section collected and transmitted in step S2 within l continuous time period l_s 、Peripheral Convergence U c_s and the vault sinks s_s The spatiotemporal data is preprocessed and reconstructed into a window structure as the model input data, expressed as [2s+1, l]. The preprocessing formula is as follows:

[0056]

[0057] Where: x i is the original data; max x i and min x i are the minimum and maximum values ​​in the data points respectively; by default, x p and x q 1 and -1 respectively;

[0058] S42, representing the tunnel segments centered on the deformation monitoring section as different nodes and establishing edges between the nodes of adjacent tunnel segments to construct graphical structure data, taking the spatial data at the same time as a graph, and constructing the input data into a graph sequence;

[0059] S43, inputting the graphics at different times into the corresponding GAT units in the GAT model layer, extracting the deformation space features at each time step, arranging and splicing them in chronological order;

[0060] S44, extracting features of the data output by the GAT layer through a convolutional layer;

[0061] S45. Use the LSTM model layer to extract the time characteristics of the deformation data and obtain the predicted value.

[0062] As a further preferred embodiment, the specific steps of step S5 are as follows:

[0063] S51. A three-dimensional geological model is established in Abaqus software. The position and shape of the tunnel and shield shell are defined according to the tunnel burial depth, excavation radius, and three-dimensional spatial coordinate information. At the same time, material properties are assigned to the tunnel, shield shell, and stratum according to material properties and stratum parameters. During the construction period, dynamic adjustments should be made based on the actual construction status of the shield machine and the construction process should be simulated in real time. Different strength reduction coefficients K = {K1, K2, K3, K4...} are used at equal intervals for specific sections of the tunnel at different stages for simulation analysis, and the corresponding surrounding rock plastic deformation energy increment ΔU is obtained respectively. P and mutation characteristic value Δ, plot and fit ΔU P The relationship curve between ΔU P The state when a sudden change occurs and Δ changes from positive to negative is determined as the limit state, and the corresponding tunnel longitudinal deformation U l_c / U l_s 、Peripheral Convergence U c_c / U l_s and the vault sinking s_c / U l_s That is, the deformation limit value U of the section under the corresponding state o = {U l_o , U c_o , U s_o}, based on the average three-division point of the deformation limit value, four warning levels are divided into red, orange, yellow and green, and the range of each warning level is [U o , +∞), and The corresponding surrounding rock plastic deformation energy increment ΔU is obtained P The formula for the mutation characteristic value Δ is as follows:

[0064]

[0065] Where: m is the number of reductions; n is the number of plastic units; σ ij is the stress of plastic element i at the jth strength reduction; Δε ij is the strain increment of the jth strength reduction of plastic element i; a1, a2, a3, and a4 are the derivatives of ΔU P The unknown coefficients of the quartic polynomial obtained by fitting the reduction coefficient K with the least square method;

[0066] S52, tunnel burial depth H, stratum internal friction angle φ, cohesion c, surrounding rock contact pressure F at different construction stages mn, average cutterhead speed N, average cutterhead torque T, shield advancement speed V, distance x between deformation monitoring surface and excavation surface, and deformation limit value U of the cross section l_o , U c_o and U s_o The data is constructed into a data set, a BP model is established and trained, and the trained BP model is transferred to a new data set of other sections of the tunnel for training, the first few layers of the BP model are frozen, and the parameters of the last few layers of the BP model are trained or new model layers are added;

[0067] S53, obtaining the graded warning range of each monitoring section during the construction period from step S51 and step S52, and performing risk warning during the tunnel construction period in combination with the tunnel section deformation monitoring value or predicted value;

[0068] S54, delete the parameters of the construction period in step S51 and step S52, re-execute steps S51 and S52, obtain the graded warning range of each monitoring section during the service period, and perform real-time risk warning of the tunnel service period in combination with the tunnel section deformation monitoring value or predicted value. The parameters of the construction period include: surrounding rock contact pressure F mn , average cutterhead speed N, average cutterhead torque T, shield advancement speed V, and the distance x between the deformation monitoring surface and the excavation surface;

[0069] S55. Perform real-time risk warning for the entire life cycle of the tunnel according to step S53 and step S54.

[0070] As a further preferred embodiment, the specific steps of step S6 are as follows:

[0071] S61, collecting tunnel 3D point cloud data, full life cycle structural deformation information and shield construction status through sensors, standardizing and encoding the acquired data into JSON format, and transmitting the data to the server via a communication transmission module, wherein the communication transmission module is composed of an interface transmission base station, power supply equipment, GPRS module, Zigbee module and ARM9 processor arranged in the tunnel and providing Zigbee, CAN bus and TCP / IP protocols;

[0072] S62. The digital twin management platform for shield tunnels includes a shield tunnel design and modeling module, a structural deformation prediction and early warning module, and a user interface and interaction module. The digital twin management platform for shield tunnels accesses the database in the server and processes various data to realize the construction of a virtual model of the tunnel-geological complex. Based on the structural deformation prediction model and graded early warning method during the tunnel construction period and service period, the full-space deformation prediction and risk early warning of the shield tunnel throughout its life cycle are realized. A three-dimensional visualization interface is provided on the digital twin management platform for tunnels, with the BIM model as the data carrier, to transmit, update and synchronously render the tunnel deformation status, shield construction status and early warning status evaluation in real time.

[0073] As a further preferred embodiment: the number of data acquisition modules with incremental encoders in step S22 is 5.

[0074] As a further preferred embodiment, the number of pressure monitoring sections in step S31 is 5, and the number of pressure monitoring points arranged on each pressure monitoring section is 5.

[0075] As a further preferred embodiment, the ratio of dividing the test set and the training set in step S33 is 1:4.

[0076] Compared with the prior art, the present invention has the following beneficial effects:

[0077] 1. Based on the shield tunnel-geological complex model that integrates stratigraphic information and high-precision spatial data, the present invention proposes a digital twin-based shield tunnel full-life structural deformation monitoring and early warning method.

[0078] 2. The present invention is based on the constructed tunnel-geological complex model, dynamically adjusts the model in real time according to the monitoring value, forms a digital twin model system, studies the real-time early warning mechanism and method of tunnel deformation, and realizes tunnel deformation monitoring and graded early warning.

[0079] 3. The present invention obtains real-time structural deformation data through a sensor network composed of FBG and incremental encoders, and dynamically adjusts the prediction model, considers the instability state of the tunnel from an energy perspective, and determines the deformation warning threshold of the tunnel monitoring section during the construction period and service period based on strength reduction and cusp mutation theory, thereby realizing deformation monitoring, prediction and warning of the shield tunnel throughout its life cycle. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 The figure is an overall flow chart of the early warning method of the present invention.

[0081] Figure 2 It is a schematic diagram of the vehicle-mounted laser detection system in the present invention.

[0082] Figure 3 This is a layout diagram of the tunnel longitudinal deformation monitoring sensors in the present invention.

[0083] Figure 4 This is a layout diagram of the tunnel circumferential deformation monitoring sensors in the present invention.

[0084] Figure 5 It is a schematic diagram of the installation of the circumferential deformation monitoring sensor in the present invention.

[0085] Figure 6 It is the overall framework diagram of the early warning system of the present invention.

[0086] Indicated in the figure:

[0087] 1. Inspection vehicle; 2. LiDAR scanner; 3. Inertial navigation system; 4. Inclinometer; 5. Computer; 6. GPS; 7. Deformation monitoring section; 8. Fiber Bragg grating multi-point displacement meter; 9. Hollow shaft magnetic sensor A; 10. Hollow shaft magnetic sensor B; 11. Solid shaft sensor C; 12. Wire support pulley; 13. Feeler; 14. Flexible wire; 15. Pulley; 16. Heavy object. DETAILED DESCRIPTION

[0088] Advantages and features of the present invention will be illustrated and explained by the following non-limiting description of preferred embodiments thereof, which are given by way of example only with reference to the accompanying drawings.

[0089] like Figure 1-6 As shown, the present invention proposes a method for monitoring and early warning of structural deformation of shield tunnels throughout their life cycle based on digital twins. This method takes digital twins as the core, integrates stratum information and high-precision spatial data to achieve structural reconstruction of the tunnel-geological complex, and proposes a real-time deformation monitoring, early warning method and twin system during tunnel construction and service to ensure the service safety of the tunnel structure throughout its life cycle and the advance warning of failure risks. Specifically, it includes the following steps:

[0090] S1. Based on the point cloud data of the completed tunnel section, a tunnel-geology complex model integrating BIM and 3D point cloud technology is constructed.

[0091] S11. Obtaining point cloud data of the completed section of the tunnel based on the vehicle-mounted laser detection system.

[0092] S111, a laser radar scanner 2 is arranged above the inspection vehicle 1, and a local coordinate system with the center of the laser radar scanner 2 as the origin is established according to the laser radar scanner 2 above the inspection vehicle 1, and the three-dimensional point cloud data P of the tunnel in the local coordinate system is obtained. local ;

[0093] S112, installing GPS6 on the inspection vehicle 1, and obtaining the position coordinates T of the vehicle body in the global coordinate system according to the GPS6 on the inspection vehicle 1;

[0094] S113, installing an inertial navigation system 3 on the inspection vehicle 1 to keep time synchronization with the data of the laser radar scanner 2;

[0095] S114, an inclinometer 4 is arranged inside the inspection vehicle 1 to obtain accurate three-axis attitude angle data and construct a rotation matrix R;

[0096] S115, based on the three-dimensional point cloud data P of the tunnel in the local coordinate system obtained in step S111 local, the position coordinates T of the vehicle body obtained in step S112 in the global coordinate system and the rotation matrix R obtained in step S114, to obtain accurate three-dimensional point cloud coordinate information, that is, the three-dimensional point cloud data P of the tunnel in the global coordinate system global , the formula is as follows:

[0097] P global =R·P local +T

[0098] like Figure 2 As shown in the figure, a vehicle-mounted laser detection system is used to obtain the point cloud data of the completed section of the tunnel. A laser radar scanner 2 is installed above the detection vehicle 1. The detection vehicle 1 moves forward along the axis of the tunnel to cooperate with the rotation of the laser radar scanner 2. A local coordinate system with the center of the laser radar scanner 2 as the origin is established to obtain the three-dimensional point cloud data P of the tunnel in the local coordinate system. local ; Install GPS 6 under the laser radar scanner 2 of the inspection vehicle 1 to obtain the position coordinates T of the vehicle body in the global coordinate system; Arrange an inertial navigation system 3 on the right side of the laser radar scanner 2 of the inspection vehicle 1 to keep time synchronization with the data of the laser radar scanner 2; Arrange an inclinometer 4 inside the inspection vehicle 1 to assist in correcting the attitude angle of the inertial navigation system, record accurate three-axis attitude angle data and construct a rotation matrix R; Use formula P global =R·P local +T converts the point cloud data in the local coordinate system to the global coordinate system to obtain accurate three-dimensional point cloud coordinate information.

[0099] The construction of the tunnel-geology complex model includes the following steps:

[0100] S12. Reconstruct the shield tunnel structure according to the point cloud data of the completed tunnel section obtained in step S11.

[0101] S121, point cloud data preprocessing;

[0102] The open source point cloud processing platform CloudCompare was used to manually remove obvious abnormal point cloud data, and then the trajectory noise and linear noise were eliminated based on the KNN algorithm. The conditional random field method was used to smooth the point cloud data, and the voxel context was used to compress the octree structured data to achieve data compression processing.

[0103] S122, constructing a triangulation network of flat tunnel point cloud data based on the Delaunay triangulation algorithm, and reconstructing the main contour of the tunnel by combining the three-dimensional point cloud data of the tunnel;

[0104] This step is to extract the main contour of the tunnel.

[0105] S123, projecting the point cloud data onto the yox and zox planes respectively, constructing points on the central axis by using boundary functions obtained by fitting boundary points on both sides of the projected point cloud data, and using a cubic polynomial as a fitting function to obtain fitting equations of the tunnel axis in each plane;

[0106] This step is to fit the tunnel axis, and the fitting equations of the tunnel axis in each plane are x, y(x) and z(x).

[0107] S124. Expand the point cloud data along the tunnel axis and determine the pixel value of the point based on its geometric features, create a binary image, use the Laplace operator to detect the boundary of the binary image, divide it into single annular linings along the annular joints based on morphological processing of the boundary image, set the same pixel value area to the same value through connected component analysis and divide it into different segments, obtain the center point and radius parameters by fitting the segment point cloud data, develop the as-built model of the lining ring using the CATIA platform, import the fitting parameters of the center point and radius of different segments, the tunnel axis, and material properties as input parameters into the platform, establish geometric size constraint relationships, create a parametric lining ring model, and thus reconstruct the shield tunnel structure.

[0108] This step is to reconstruct the shield tunnel segment: First, expand the point cloud along the tunnel axis, use the least squares method to fit the expanded point cloud and calculate the plane normal vector Determine the height and width of the binary image based on the point cloud range, and then calculate the angle between the normal vector of each point in the unfolded point cloud and the plane normal vector when When the pixel value is set to 0, otherwise it is set to 255, so as to obtain a binary image of the point cloud; the Laplace operator is used to detect the boundary of the binary image, and the boundary image is divided into a single annular lining along the annular joint based on morphological processing. Finally, the areas with the same pixel value are set to the same value through connected component analysis and divided into different segments. The geometric parameters such as the center point and radius are obtained by fitting the segment point cloud. Finally, the as-built model of the lining ring is developed using the CATIA platform, and the fitting parameters of the center point and radius of different segments, the tunnel axis, and material properties are imported into the platform as input parameters. The geometric size constraint relationship is established, and a parametric lining ring model is created to reconstruct the shield tunnel structure.

[0109] S13, adaptively aligning the geological environment modeling and the shield tunnel structure reconstructed in step S12 to construct a tunnel-geology complex model.

[0110] The geotechnical engineering information of the tunnel project, such as geometry, topology and attributes, is integrated and stored in the GIS model. GIS data is introduced by extending the IFC standard and the geological entity relationships and attributes are described in detail. The conditional random field method is used to model the spatial variation of geological parameters. The axis data of the completed section structure model of the shield tunnel and key nodes, such as the starting point, end point, turning point, etc., are aligned with the corresponding positions in the geological environment model to realize the construction of the tunnel-geology complex model.

[0111] S2. Use FBG and incremental sensing to monitor the deformation of the tunnel in real time throughout its life cycle.

[0112] S21, arranging fiber Bragg grating multi-point displacement meters 8 along the longitudinal direction on each deformation monitoring section 7 of the shield tunnel, and obtaining the longitudinal deformation U of the tunnel during construction l_c and the longitudinal deformation during service life U l_c and transmit it to the receiving instrument;

[0113] S22, on each deformation monitoring section 7 of the shield tunnel, a pulley 15, a plurality of data acquisition modules with incremental encoders and a weight 16 are sequentially arranged along the tunnel circumferential direction, and the pulley 15, the plurality of data acquisition modules with incremental encoders and the weight 16 are connected by a flexible wire 14 to form a non-closed polygonal convergence monitoring system;

[0114] S221, respectively arrange data acquisition modules with incremental encoders at the arch crown, the arch waists on both sides and the middle position of the tunnel deformation monitoring section, and the arrangement points of the data acquisition modules with incremental encoders are named P1, P2, ..., P n , P1 point is fixed with pulley 15, P n A weight is connected at a point to maintain the wire tension, and the pulley 15 acts as a ball joint;

[0115] The purpose of this step is to arrange a non-closed polygon convergence monitoring system. The number of the data acquisition modules with incremental encoders is preferably five.

[0116] S222, use the total station to obtain the precise initial position coordinates of point P1, record the initial calibration values ​​of the angle and wire length at each point, and use trigonometric functions to obtain P2, P3, ..., P n The initial position coordinates of the point;

[0117] S223, a hollow shaft magnetic sensor A9 is set on the left side of each group of data acquisition modules with incremental encoders, and a hollow shaft magnetic sensor B10 is set on the right side. The hollow shaft magnetic sensor A9 is used to monitor the angle increment of the left side wire of each point between time t and time t-1, and the hollow shaft magnetic sensor B10 is used to monitor the angle increment of the right side wire of each point between time t and time t-1. Combined with the initial calibration value obtained in step S222, P1, P2, ..., P n The angle value at the point at time t and The formula is as follows:

[0118]

[0119]

[0120]

[0121] Where: i is the sequence number of the data acquisition module layout point, ranging from 1 to n; P is the time t i The angle value of the conductors on both sides of the point; P is the time t i The angle between the conductor on the right side of the point and the vertical direction; P is measured by the hollow shaft magnetic sensor A i The angle increment of the wire on the left side of the point between time t and time t-1; P is measured by the hollow shaft magnetic sensor B i The angle increment of the wire on the right side of the point between time t and time t-1.

[0122] S224, a wire support pulley 12 and a solid shaft sensor C11 are set at the connection between two flexible wires 14 inside each group of data acquisition modules with incremental encoders, and the angle increment Δδ of the wire support pulley 12 is obtained by using the solid shaft sensor C11. i-1 Point and P i Initial length of the wire between points Get the length of each wire segment at time t The formula is as follows:

[0123]

[0124] Where: P is the time t i Point and P i-1 The length of the wire between the points; is the initial time P i Point and P i-1 The length of the wire between the points; r is the radius of the wire support pulley 12; and At time t, P i-1 Point and P i The angle increment of the wire support pulley 12 in the point position.

[0125] S225: The angle value at time t obtained in step S223 and the wire length value at time t obtained in step S224 Using trigonometric functions, we can get P2, P3, ..., P n The position coordinates of the point at time t, and finally the tunnel convergence U at different times during the construction period are obtained according to the changes in the coordinate information of each point. c_c 、Vault sinking s_c and the tunnel convergence U at different times during the service life c_s 、Vault sinking s_s .

[0126] Deformation monitoring sections 7 are set up every 50m in the entire shield tunnel, as well as at tunnel entrances and exits, joints, and locations with complex geological conditions.

[0127] like Figure 3 As shown in the figure, fiber Bragg grating multi-point displacement meters 8 are arranged longitudinally at the left and right arch waists of each deformation monitoring section 7 of the shield tunnel, and the longitudinal deformation average value of the left and right arch waists is used as the longitudinal deformation value of the deformation monitoring section. The longitudinal deformation U of the tunnel during construction period is calculated by connecting the instrument to the network. l_c and the longitudinal deformation during service life U l_s Transmitted to receiving instrument over long distance;

[0128] like Figure 4 and Figure 5 As shown, a non-closed polygonal convergence monitoring system consisting of a flexible wire 14, a pulley 15 and a plurality of data acquisition modules with incremental encoders is arranged along the tunnel ring on each deformation monitoring section 7 of the shield tunnel. The number of the data acquisition modules of the incremental encoder is preferably 5, that is, the data acquisition modules of the incremental encoder are arranged at the arch crown, the arch waists on both sides and the middle position of the tunnel deformation monitoring section 7, and the arrangement points of the data acquisition modules of the 5 incremental encoders are named P1, P2, P3, P4, and P5 from left to right, where P The 1 point is fixed with a pulley 15 acting as a ball joint, the P5 point is connected with a weight 16 to maintain the tension of the flexible wire 14, a hollow shaft magnetic sensor A9 is arranged on the left side of each data acquisition module with an incremental encoder, and a hollow shaft magnetic sensor B10 is arranged on the right side, the antennae 13 of the hollow shaft magnetic sensor A9 and the hollow shaft magnetic sensor B10 are respectively overlapped on the left and right flexible wires 14, and a wire support pulley 12 and a solid shaft sensor C11 are arranged at the connection between the two flexible wires 14 inside each data acquisition module with an incremental encoder.

[0129] First, the total station is used to obtain the precise initial position coordinates of point P1, and the initial calibration values ​​of the angles and wire lengths of each point are recorded. The initial position coordinates of points P2, P3, P4, and P5 are calculated in sequence using trigonometric functions. During the deformation monitoring process, hollow shaft magnetic sensors A9 and B10 are used to monitor the angle increments of the left and right wires of each point between time t and time t-1. Based on the initial calibration values ​​recorded by the system and the angle increment values ​​at time t, the angle values ​​at points P1, P2, P3, P4, and P5 at time t are obtained. and At the same time, the solid shaft sensor C11 is used to obtain the angle increment Δδ of the wire support pulley 12, combined with P i-1 Point and P i Initial length of the wire between points Get the length of the wire segment The angle value at time t obtained based on the above calculation and wire length values The position coordinates of points P2, P3, P4, and P5 at time t are obtained in turn using trigonometric functions. Finally, based on the changes in the coordinate information of each point, the tunnel convergence U at different times during the construction period is obtained. c_c 、Vault sinking s_c and the tunnel convergence U at different times during the service life c_s 、Vault sinking s_s .

[0130] S3. Construct the LGBM proxy model to update the tunnel structure deformation prediction during the construction period.

[0131] S31. Several pressure monitoring sections are evenly arranged in the longitudinal direction of the shield machine, and several pressure monitoring points are arranged in each pressure monitoring section. Preferably, there are 5 pressure monitoring sections, and 5 pressure monitoring points are arranged in each pressure monitoring section. A micro weighing sensor is installed outside the shield machine protective cover at each pressure monitoring point. The micro weighing sensor is connected to the data recorder using a shielded data cable to measure the surrounding rock contact pressure F acting on each pressure monitoring point of the shield machine at time t. mn ; Wireless acceleration sensors and strain gauge torque sensors are deployed on different tool installation radii on the back of the cutterhead, and data is transmitted to the server through a wireless gateway near the cutterhead to obtain the average cutterhead speed N and average cutterhead torque T; a linear displacement sensor is installed on the shield machine shell to obtain the shield advancement speed V, thereby obtaining the distance x between different deformation monitoring surfaces and the excavation surface; a pressure sensor and a flow meter are installed in the shield machine grouting system to obtain the grouting pressure P and grouting volume Q;

[0132] S32, the surrounding rock contact pressure F obtained in step S31 mn, average cutterhead speed N, average cutterhead torque T, shield advancement speed V, tunnel burial depth H, distance x between deformation monitoring surface and excavation surface, stratum internal friction angle φ and cohesion c are used as the input of LGBM model. The stratum parameters need to be weighted considering the position and thickness of the overlying stratum of the tunnel. The longitudinal deformation U of the tunnel during construction period is used as the input of LGBM model. l_c 、Peripheral Convergence U c_c and the vault sinks s_c As the output of the LGBM model, a LGBM proxy model is established.

[0133] S33, preprocess the input and output data of the LGBM proxy model and divide them into test set and training set. The formula is as follows:

[0134]

[0135] Where: x i is the original data; max x i and min x i are the minimum and maximum values ​​in the data points respectively; by default, x p and x q The ratio of the test set to the training set is preferably 1:4.

[0136] S34, according to the Bayesian optimization method based on Gaussian process, the longitudinal deformation U corresponding to the tunnel construction period is obtained respectively. l_c 、Peripheral Convergence U c_c U with the vault sinking s_c The hyperparameters for the best prediction effect, the optimization process formula is as follows:

[0137]

[0138] R 1:t ={(x1,y1),(x2,y2)…(x t ,y t )}

[0139] y t =f(x t )+ε t

[0140] f(x t )~GP(m(x),k(x,x′))

[0141] Where: f is the objective function; y t is the observed value at the tth step; x t is the hyperparameter of the tth step; ε t is the observation error; R 1:t is the aggregate of observations from the first t steps; p(R 1:t|f) is R under a given objective function 1:t The likelihood distribution of f; p(f) is the prior distribution of f; p(f|R 1:t ) is the observed R 1:t The posterior distribution of f; m(x) is the mean function; k(x,x') is the covariance function.

[0142] S35. Use the hyperparameters with the best prediction effect obtained in step S34 and the test set and training set divided in step S33 to train and evaluate the LGBM proxy model established in step S32. Use the trained model to predict the deformation results of the unknown sections of the tunnel under this operating state based on the burial depth, stratum parameters and distance from the excavation face of the unknown sections of the tunnel; collect dynamic monitoring data during the shield construction process, perform incremental training on the model and update the prediction of the tunnel deformation state.

[0143] S4. Construct a GAT-LSTM model to predict the structural deformation of the tunnel during its service life.

[0144] S41, the service life longitudinal deformation U of the s sections before and after the target section collected and transmitted in step S2 within l continuous time period l_s 、Peripheral Convergence U c_s and the vault sinks s_s The spatiotemporal data is preprocessed and reconstructed into a window structure as the model input data, expressed as [2s+1, l]. The preprocessing formula is as follows:

[0145]

[0146] Where: x i is the original data; max x i and min x i are the minimum and maximum values ​​in the data points respectively; by default, x p and x q 1 and -1 respectively.

[0147] S42, representing the tunnel segments centered on the deformation monitoring section as different nodes and establishing edges between the nodes of adjacent tunnel segments to construct graphical structure data, taking the spatial data at the same time as a graph, and constructing the input data into a graph sequence;

[0148] S43, inputting the graphics at different times into the corresponding GAT units in the GAT model layer, extracting the deformation space features at each time step, arranging and splicing them in chronological order;

[0149] S44, extracting features of the data output by the GAT layer through a convolutional layer;

[0150] S45. Use the LSTM model layer to extract the time characteristics of the deformation data and obtain the predicted value.

[0151] In this step, the deformation characteristics of the target deformation monitoring section and the two deformation monitoring sections before and after it in five consecutive time periods are used to predict the future deformation of the target monitoring section. The longitudinal deformation U of the tunnel during service collected and transmitted in step S2 is used as the l_s 、Peripheral Convergence U c_s and the vault sinks s_s According to the formula Preprocessing is performed and reconstructed into window structures respectively. The model input data dimension is {5, 5, 3}. The tunnel segments centered on the deformation monitoring section are represented as different nodes, and edges are established between the nodes of adjacent tunnel segments to construct graphical structure data. The spatial data at the same time is used as a graph with a data dimension of {5, 3}.

[0152] The graphics at five different moments are input into the GAT model layer, and different GAT units perform node attention weighted processing, adaptively extract the deformation space features at each time step, arrange and splice them in chronological order; then, the convolution layer is used to extract features of the data output by the GAT layer, and finally the LSTM model layer is used to extract the temporal features of the deformation data and obtain the predicted value.

[0153] S5. Construct a tunnel structure deformation classification early warning method based on strength reduction and cusp catastrophe theory.

[0154] S51. A three-dimensional geological model is established in Abaqus software. The position and shape of the tunnel and shield shell are defined according to the tunnel burial depth, excavation radius, and three-dimensional spatial coordinate information. At the same time, material properties are assigned to the tunnel, shield shell, and stratum according to material properties and stratum parameters. During the construction period, dynamic adjustments should be made based on the actual construction status of the shield machine and the construction process should be simulated in real time. Different strength reduction coefficients K = {K1, K2, K3, K4...} are used at equal intervals for specific sections of the tunnel at different stages for simulation analysis, and the corresponding surrounding rock plastic deformation energy increment ΔU is obtained respectively. P and mutation characteristic value Δ, plot and fit ΔU P The relationship curve between ΔU P The state when a sudden change occurs and Δ changes from positive to negative is determined as the limit state, and the corresponding tunnel longitudinal deformation U l_c / U l_s 、Peripheral Convergence U c_c / U l_s and the vault sinking s_c / U l_s That is, the deformation limit value U of the section under the corresponding state o = {U l_o , Uc_o , U s_o}, based on the average three-division point of the deformation limit value, four warning levels are divided into red, orange, yellow and green, and the range of each warning level is [U o , +∞), and The corresponding surrounding rock plastic deformation energy increment ΔU is obtained P The formula for the mutation characteristic value Δ is as follows:

[0155]

[0156] Where: m is the number of reductions; n is the number of plastic units; σ ij is the stress of plastic element i at the jth strength reduction; Δε ij is the strain increment of the jth strength reduction of plastic element i; a1, a2, a3, and a4 are the derivatives of ΔU P The unknown coefficients of the quartic polynomial obtained by fitting the reduction factor K using the least squares method.

[0157] S52, tunnel burial depth H, stratum internal friction angle φ, cohesion c, surrounding rock contact pressure F at different construction stages mn , average cutterhead speed N, average cutterhead torque T, shield advancement speed V, distance x between deformation monitoring surface and excavation surface, and deformation limit value U of the cross section l_o , U c_o and U s_o The data is constructed as a data set, a BP model is established and trained, and the trained BP model is migrated to new data sets of other sections of the tunnel for training. The first few layers of the BP model are frozen, and the parameters of the last few layers of the BP model are trained or new model layers are added.

[0158] The BP model in this step is used to predict the deformation limit of the tunnel section, freeze the first few layers of the BP model, train the parameters of the last few layers of the BP model or add new model layers to achieve dynamic update of the deformation warning value of the entire tunnel space and improve calculation efficiency.

[0159] S53, obtaining the graded warning range of each monitoring section during the construction period from step S51 and step S52, and performing risk warning during the tunnel construction period in combination with the tunnel section deformation monitoring value or predicted value;

[0160] S54, delete the parameters of the construction period in step S51 and step S52, re-execute steps S51 and S52, obtain the graded warning range of each monitoring section during the service period, and perform real-time risk warning of the tunnel service period in combination with the tunnel section deformation monitoring value or predicted value. The parameters of the construction period include: surrounding rock contact pressure F mn , average cutterhead speed N, average cutterhead torque T, shield advancement speed V, and the distance x between the deformation monitoring surface and the excavation surface;

[0161] S55. Perform real-time risk warning for the entire life cycle of the tunnel according to step S53 and step S54.

[0162] like Figure 6 As shown, the present invention constructs a tunnel monitoring and early warning system based on digital twins, which includes the following steps:

[0163] S6. Build a tunnel monitoring and early warning system based on digital twins.

[0164] S61. Collect tunnel 3D point cloud data, full life cycle structural deformation information and shield construction status through sensors, standardize the acquired data and encode it into JSON format, and transmit it to the server via a communication transmission module. The communication transmission module is composed of an interface transmission base station, power supply equipment, GPRS module, Zigbee module and ARM9 processor arranged in the tunnel and providing Zigbee, CAN bus and TCP / IP protocols.

[0165] S62. The shield tunnel digital twin management platform includes a shield tunnel design and modeling module, a structural deformation prediction and early warning module, and a user interface and interaction module. Based on the shield tunnel digital twin management platform, the database in the server is accessed and various data are processed to realize the construction of a virtual model of the tunnel-geological complex. Based on the structural deformation prediction model and graded early warning method during the tunnel construction period and service period, the full-space deformation prediction and risk early warning of the shield tunnel throughout its life cycle are realized; a three-dimensional visualization interface is provided on the tunnel digital twin management platform, with the BIM model as the data carrier, to transmit, update and synchronously render the tunnel deformation status, shield construction status and early warning status evaluation in real time.

[0166] In addition to the above embodiments, the present invention may also have other implementation modes. Any technical solutions formed by equivalent replacement or equivalent transformation shall fall within the protection scope required by the present invention.

Claims

1. A digital twin-based shield tunnel full-life structural deformation monitoring and early warning method, characterized by: The following steps are included: S1. Based on the point cloud data of the completed tunnel section, a tunnel-geology complex model integrating BIM and 3D point cloud technology is constructed; S2, using FBG and incremental sensing to monitor tunnel deformation in real time throughout its life cycle; S3, construct the LGBM proxy model to update the tunnel structure deformation prediction during the construction period; S4, construct the GAT-LSTM model to predict the structural deformation of the tunnel during service; S5. Construct a tunnel structure deformation classification early warning method based on strength reduction and cusp catastrophe theory; S6. Build a tunnel monitoring and early warning system based on digital twins.

2. The digital twin-based shield tunnel full-life structural deformation monitoring and early warning method according to claim 1 is characterized by: The specific steps of step S1 are as follows: S11. Obtaining point cloud data of the completed section of the tunnel according to the vehicle-mounted laser detection system; S111, a laser radar scanner is set above the inspection vehicle, and a local coordinate system with the center of the laser radar scanner as the origin is established based on the laser radar scanner above the inspection vehicle to obtain the three-dimensional point cloud data P of the tunnel in the local coordinate system local ; S112, installing a GPS on the inspection vehicle, and obtaining the position coordinates T of the vehicle body in the global coordinate system according to the GPS on the inspection vehicle; S113, installing an inertial navigation system on the inspection vehicle to keep time synchronization with the data of the laser radar scanner; S114, setting an inclinometer inside the inspection vehicle to obtain accurate three-axis attitude angle data and construct a rotation matrix R; S115, based on the three-dimensional point cloud data P of the tunnel in the local coordinate system obtained in step S111 local , the position coordinates T of the vehicle body obtained in step S112 in the global coordinate system and the rotation matrix R obtained in step S114, to obtain accurate three-dimensional point cloud coordinate information, that is, the three-dimensional point cloud data P of the tunnel in the global coordinate system global , the formula is as follows: P global =R·P local +T S12. Reconstruct the shield tunnel structure according to the point cloud data of the completed tunnel section obtained in step S11. S121, point cloud data preprocessing; S122, constructing a triangulation network of flat tunnel point cloud data based on the Delaunay triangulation algorithm, and reconstructing the main contour of the tunnel by combining the three-dimensional point cloud data of the tunnel; S123, projecting the point cloud data onto the yox and zox planes respectively, constructing points on the central axis by using boundary functions obtained by fitting boundary points on both sides of the projected point cloud data, and using a cubic polynomial as a fitting function to obtain fitting equations of the tunnel axis in each plane; S124, unfolding the point cloud data along the tunnel axis and determining the pixel value of the point based on its geometric features, creating a binary image, using the Laplace operator to detect the boundary of the binary image, dividing it into a single ring lining along the ring joint based on morphological processing of the boundary image, setting the same pixel value area to the same value through connected component analysis and dividing it into different segments, obtaining the center point and radius parameters by fitting the segment point cloud data, developing the as-built model of the lining ring using the CATIA platform, importing the fitting parameters of the center point and radius of different segments, the tunnel axis, and the material properties as input parameters into the platform, establishing a geometric size constraint relationship, creating a parameterized lining ring model, and thus reconstructing the shield tunnel structure; S13, adaptively aligning the geological environment modeling and the shield tunnel structure reconstructed in step S12 to construct a tunnel-geology complex model.

3. The digital twin-based shield tunnel full-life structural deformation monitoring and early warning method according to claim 2 is characterized by: The specific steps of step S2 are as follows: S21. Fiber Bragg grating multi-point displacement meters are arranged longitudinally on each deformation monitoring section of the shield tunnel to obtain the longitudinal deformation U of the tunnel during construction. l_c and the longitudinal deformation during service life U l_s and transmit it to the receiving instrument; S22, on each deformation monitoring section of the shield tunnel, a pulley, a plurality of data acquisition modules with incremental encoders and a weight are sequentially arranged along the tunnel circumferential direction, wherein the pulley, the plurality of data acquisition modules with incremental encoders and the weight are connected with flexible wires to form a non-closed polygonal convergence monitoring system; S221, respectively arrange data acquisition modules with incremental encoders at the arch crown, the arch waists on both sides and the middle position of the tunnel deformation monitoring section, and the arrangement points of the data acquisition modules with incremental encoders are named P1, P2, ..., P n , P1 point is fixed with a pulley, P n Weights are attached at points to maintain wire tension, with the pulley acting as a ball-and-socket joint; S222, use the total station to obtain the precise initial position coordinates of point P1, record the initial calibration values ​​of the angle and wire length at each point, and use trigonometric functions to obtain P2, P3, ..., P n The initial position coordinates of the point; S223, a hollow shaft magnetic sensor A is set on the left side of each group of data acquisition modules with incremental encoders, and a hollow shaft magnetic sensor B is set on the right side. The hollow shaft magnetic sensor A is used to monitor the angle increment of the left side wire of each point between time t and time t-1, and the hollow shaft magnetic sensor B is used to monitor the angle increment of the right side wire of each point between time t and time t-1. Combined with the initial calibration value obtained in step S222, P1, P2, ..., P n The angle value at the point at time t and The formula is as follows: Where: i is the sequence number of the data acquisition module layout point, ranging from 1 to n; P is the time t i The angle value of the conductors on both sides of the point; P is the time t i The angle between the conductor on the right side of the point and the vertical direction; P is measured by the hollow shaft magnetic sensor A i The angle increment of the wire on the left side of the point between time t and time t-1; P is measured by the hollow shaft magnetic sensor B i The angle increment of the wire on the right side of the point between time t and time t-1; S224, a wire support pulley and a solid shaft sensor C are set at the connection between two flexible wires inside each group of data acquisition modules with incremental encoders, and the angle increment Δδ of the wire support pulley is obtained by using the solid shaft sensor C. i-1 Point and P i Initial length of the wire between points Get the length of each wire segment at time t The formula is as follows: Where: P is the time t i Point and P i-1 The length of the wire between the points; is the initial time P i Point and P i-1 The length of the wire between the points; r is the radius of the wire support pulley; and At time t, P i-1 Point and P i The angle increment of the wire support pulley in the point position; S225: The angle value at time t obtained in step S223 and the wire length value at time t obtained in step S224 Using trigonometric functions, we can get P2, P3, ..., P n The position coordinates of the point at time t, and finally the tunnel convergence U at different times during the construction period are obtained according to the changes in the coordinate information of each point. c_c 、Vault sinking s_c and the tunnel convergence U at different times during the service life c_s 、Vault sinking s_s .

4. The digital twin-based shield tunnel full-life structural deformation monitoring and early warning method according to claim 3 is characterized by: The specific steps of step S3 are as follows: S31. Several pressure monitoring sections are evenly arranged in the longitudinal direction of the shield machine. Several pressure monitoring points are arranged in each pressure monitoring section. Micro weighing sensors are installed outside the shield machine protective cover at each pressure monitoring point. The micro weighing sensors are connected to the data recorder using a shielded data cable to measure the surrounding rock contact pressure F acting on each pressure monitoring point of the shield machine at time t. mn ; Wireless acceleration sensors and strain gauge torque sensors are deployed on different tool installation radii on the back of the cutterhead, and data is transmitted to the server through a wireless gateway near the cutterhead to obtain the average cutterhead speed N and average cutterhead torque T; a linear displacement sensor is installed on the shield machine shell to obtain the shield advancement speed V, thereby obtaining the distance x between different deformation monitoring surfaces and the excavation surface; a pressure sensor and a flow meter are installed in the shield machine grouting system to obtain the grouting pressure P and grouting volume Q; S32: The surrounding rock contact pressure F obtained in step S31 is mn , average cutterhead speed N, average cutterhead torque T, shield advancement speed V, tunnel burial depth H, distance x between deformation monitoring surface and excavation surface, stratum internal friction angle φ and cohesion c are used as the input of LGBM model. The stratum parameters need to be weighted considering the position and thickness of the overlying stratum of the tunnel. The longitudinal deformation U of the tunnel during construction period is used as the input of LGBM model. l_c 、Peripheral Convergence U c_c and the vault sinks s_c As the output of the LGBM model, the LGBM proxy model is established; S33, preprocess the input and output data of the LGBM proxy model and divide them into test set and training set. The formula is as follows: Where: x i is the original data; max x i and min x i are the minimum and maximum values ​​in the data points respectively; by default, x p and x q 1 and -1 respectively; S34, according to the Bayesian optimization method based on Gaussian process, the longitudinal deformation U corresponding to the tunnel construction period is obtained respectively. l_c 、Peripheral Convergence U c_c U with the vault sinking s_c The hyperparameters for the best prediction effect, the optimization process formula is as follows: R 1:t ={(x1,y1) , (x2,y2)…(x t ,y t )} y t =f(x t )+ε t f(x t )~GP(m(x),k(x,x′)) Where: f is the objective function; y t is the observed value at the tth step; x t is the hyperparameter of the tth step; ε t is the observation error; R 1:t is the aggregate of observations from the first t steps; p(R 1:t |f) is R under a given objective function 1:t The likelihood distribution of f; p(f) is the prior distribution of f; p(f|R 1:t ) is the observed R 1:t The posterior distribution of f; m(x) is the mean function; k(x,x') is the covariance function; S35, using the hyperparameters of the best prediction effect obtained in step S34 and the test set and training set divided in step S33, the LGBM proxy model established in step S32 is trained and evaluated, and the deformation result of the shield machine at the unknown section of the tunnel under the running state is predicted based on the buried depth, stratum parameters and distance from the excavation face at the unknown section of the tunnel by using the trained model; Dynamic monitoring data during shield construction is collected to perform incremental training on the model and update the prediction of tunnel deformation status.

5. The digital twin-based shield tunnel full-life structural deformation monitoring and early warning method according to claim 4 is characterized in that: The specific steps of step S4 are as follows: S41, the service life longitudinal deformation U of the s sections before and after the target section collected and transmitted in step S2 within l continuous time period l_s 、Peripheral Convergence U c_s and the vault sinks s_s The spatiotemporal data is preprocessed and reconstructed into a window structure as the model input data, expressed as [2s+1, l]. The preprocessing formula is as follows: Where: x i is the original data; max x i and min x i are the minimum and maximum values ​​in the data points respectively; by default, x p and x q 1 and -1 respectively; S42, representing the tunnel segments centered on the deformation monitoring section as different nodes and establishing edges between the nodes of adjacent tunnel segments to construct graphical structure data, taking the spatial data at the same time as a graph, and constructing the input data into a graph sequence; S43, inputting the graphics at different times into the corresponding GAT units in the GAT model layer, extracting the deformation space features at each time step, arranging and splicing them in chronological order; S44, extracting features of the data output by the GAT layer through a convolutional layer; S45. Use the LSTM model layer to extract the time characteristics of the deformation data and obtain the predicted value.

6. The digital twin-based shield tunnel full-life structural deformation monitoring and early warning method according to claim 5 is characterized by: The specific steps of step S5 are as follows: S51. A three-dimensional geological model is established in Abaqus software. The position and shape of the tunnel and shield shell are defined according to the tunnel burial depth, excavation radius, and three-dimensional spatial coordinate information. At the same time, material properties are assigned to the tunnel, shield shell, and stratum according to material properties and stratum parameters. During the construction period, dynamic adjustments should be made based on the actual construction status of the shield machine and the construction process should be simulated in real time. Different strength reduction coefficients K = {K1, K2, K3, K4...} are used at equal intervals for specific sections of the tunnel at different stages for simulation analysis, and the corresponding surrounding rock plastic deformation energy increment ΔU is obtained respectively. P and mutation characteristic value Δ, plot and fit ΔU P The relationship curve between ΔU P The state when a sudden change occurs and Δ changes from positive to negative is determined as the limit state, and the corresponding tunnel longitudinal deformation U l_c / U l_s 、Peripheral Convergence U c_c / U l_s and the vault sinking s_c / U l_s That is, the deformation limit value U of the section under the corresponding state o = {U l_o , U c_o , U s_o }, based on the average three-division point of the deformation limit value, four warning levels are divided into red, orange, yellow and green, and the range of each warning level is [U o , +∞), and The corresponding surrounding rock plastic deformation energy increment ΔU is obtained P The formula for the mutation characteristic value Δ is as follows: Where: m is the number of reductions; n is the number of plastic units; σ ij is the stress of plastic element i at the jth strength reduction; Δε ij is the strain increment of the jth strength reduction of plastic element i; a1, a2, a3, and a4 are the derivatives of ΔU P The coefficients of the quartic polynomial obtained by fitting the reduction factor K with the least square method; S52, tunnel burial depth H, stratum internal friction angle φ, cohesion c, surrounding rock contact pressure F at different construction stages mn , average cutterhead speed N, average cutterhead torque T, shield advancement speed V, distance x between deformation monitoring surface and excavation surface, and deformation limit value U of the cross section l_o , U c_o and U s_o The data is constructed into a data set, a BP model is established and trained, and the trained BP model is transferred to a new data set of other sections of the tunnel for training, the first few layers of the BP model are frozen, and the parameters of the last few layers of the BP model are trained or new model layers are added; S53, obtaining the graded warning range of each monitoring section during the construction period from step S51 and step S52, and performing risk warning during the tunnel construction period in combination with the tunnel section deformation monitoring value or predicted value; S54, delete the parameters of the construction period in step S51 and step S52, re-execute steps S51 and S52, obtain the graded warning range of each monitoring section during the service period, and perform real-time risk warning of the tunnel service period in combination with the tunnel section deformation monitoring value or predicted value. The parameters of the construction period include: surrounding rock contact pressure F mn , average cutterhead speed N, average cutterhead torque T, shield advancement speed V, and the distance x between the deformation monitoring surface and the excavation surface; S55. Perform real-time risk warning for the entire life cycle of the tunnel according to step S53 and step S54.

7. The digital twin-based shield tunnel full-life structural deformation monitoring and early warning method according to claim 6 is characterized by: The specific steps of step S6 are as follows: S61, collecting tunnel 3D point cloud data, full life cycle structural deformation information and shield construction status through sensors, standardizing and encoding the acquired data into JSON format, and transmitting the data to the server via a communication transmission module, wherein the communication transmission module is composed of an interface transmission base station, power supply equipment, GPRS module, Zigbee module and ARM9 processor arranged in the tunnel and providing Zigbee, CAN bus and TCP / IP protocols; S62. The shield tunnel digital twin management platform includes a shield tunnel design and modeling module, a structural deformation prediction and early warning module, and a user interface and interaction module. The shield tunnel digital twin management platform accesses the database in the server and processes various data to realize the construction of a virtual model of the tunnel-geological complex. Based on the structural deformation prediction model and graded early warning method during the tunnel construction period and service period, the full-space deformation prediction and risk early warning of the shield tunnel throughout its life cycle are realized. A three-dimensional visualization interface is provided on the tunnel digital twin management platform, which uses the BIM model as the data carrier to transmit, update and synchronously render the tunnel deformation status, shield construction status and early warning status assessment in real time.

8. The method for monitoring and early warning of structural deformation of a shield tunnel throughout its life cycle based on digital twin according to claim 3 is characterized in that: The number of data acquisition modules with incremental encoders in step S22 is 5.

9. The method for monitoring and early warning of structural deformation of a shield tunnel throughout its life cycle based on digital twin according to claim 4 is characterized in that: In the step S31, the number of pressure monitoring sections is 5, and the number of pressure monitoring points arranged on each pressure monitoring section is 5.

10. The method for monitoring and early warning of structural deformation of a shield tunnel throughout its life cycle based on digital twin according to claim 4 is characterized in that: The ratio of dividing the test set and the training set in step S33 is 1:4.

Images