A method for dynamic monitoring and early warning of slope excavation based on digital twins

Abstract

This invention discloses a dynamic monitoring and early warning method for slope excavation based on digital twins, comprising the following steps: acquiring and preprocessing topographic point clouds, soil and rock stratification, and construction progress; establishing three-dimensional geometric and physical twins, employing finite element simulation, and completing boundary backpropagation and pore pressure rebalancing to form a twin baseline; calculating observation residuals and physical conservation residuals, and obtaining a highly consistent twin and assimilated sample set through joint updates; applying a topology-constrained slip surface evolver and implementing adaptive mesh refinement and interface reconstruction in the critical zone; constructing a cone domain for displacement and seepage pressure uncertainty based on the assimilated sample set and the critical zone reinforcement model, and outputting graded early warning results. This invention employs a dual-consistency assimilation and topology-constrained slip surface evolution method to achieve highly consistent simulation and quantitative early warning throughout the entire slope excavation process, possessing advantages such as high accuracy, strong stability, and good traceability.

Description

Technical Field

[0001] This invention relates to the field of dynamic simulation of construction processes, and in particular to a method for dynamic monitoring and early warning of slope excavation based on digital twins. Background Technology

[0002] Current slope monitoring and early warning systems mostly employ multi-source sensing combined with 3D geometric modeling and finite element simulation. Based on material parameter assignment, mesh generation, and boundary condition settings, they obtain the stress-displacement-pore pressure field, often supplemented by strength reduction or limit equilibrium assessments for stability. Parameters are then updated and corrected through cyclic calibration using historical-real-time data or Kalman-type methods, forming a closed-loop process of "monitoring-simulation-early warning." However, this process relies heavily on observational fitting in the data assimilation stage, making it difficult to simultaneously ensure strict compliance with both mechanical and seepage mechanism constraints.

[0003] However, these methods generally focus on observation fitting and lack a dual-consistency assimilation layer that combines observation consistency with physical conservation consistency, making them prone to "pseudo-fitting". Slip zone identification often relies on thresholds or geometric search and does not introduce a topologically constrained slip surface evolver that coordinates connectivity preservation, curvature continuity and interface reconstruction. Early warning is often based on empirical thresholds and lacks quantitative support for uncertainty cone domains constructed by assimilation samples, slip surface feasible regions and critical region scale factors, as well as threshold crossing probabilities, making it difficult to achieve stable and traceable hierarchical early warning.

[0004] Therefore, how to provide a dynamic monitoring and early warning method for slope excavation based on digital twins is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] One objective of this invention is to propose a dynamic monitoring and early warning method for slope excavation based on digital twins. This invention adopts a dual consistency assimilation and topological constraint slip surface evolution method to achieve highly consistent simulation and quantitative early warning of the entire slope excavation process, which has the advantages of high accuracy, strong stability and good traceability.

[0006] A method for dynamic monitoring and early warning of slope excavation based on digital twin according to an embodiment of the present invention includes the following steps:

[0007] Acquire topographic point clouds, soil and rock stratification, and construction progress; complete time alignment and spatial registration; form a standardized multi-source input set and establish a phased work condition library.

[0008] Based on the standardized multi-source input set and stage working condition library, a three-dimensional geometric twin and a physical twin are established. Finite element simulation is used to complete boundary back-pushing and pore pressure rebalancing to form a twin baseline.

[0009] The twin simulation method of double consistency-topological constraint excavation is adopted, the observation residual and the physical conservation residual are calculated in the double consistency assimilation layer, and the material properties, boundary conditions and state variables are jointly updated, to obtain the high consistency twin body after assimilation and the assimilation sample set;

[0010] In the topological constraint slip surface evolution layer, the topological constraint slip surface evolutioner is applied to the high consistency twin body, and the critical zone adaptive grid refinement and interface reconstruction are implemented, to obtain the slip surface feasible region and the critical zone reinforcement model conforming to the topological constraint;

[0011] In the uncertainty warning layer, the displacement and seepage pressure uncertainty cone domain is constructed based on the assimilation sample set and the critical zone reinforcement model, the threshold crossing probability is calculated, and the graded warning result is output.

[0012] Optionally, the generation of the stage working condition library specifically includes:

[0013] The terrain point cloud, rock-soil layering and construction progress are acquired and integrity checking is performed, the progress time axis with clear start and end time and corresponding operation layer is generated for the construction progress, and the initial data set including time mark, spatial coordinates and layering attribute is formed;

[0014] Time alignment is performed on the initial data set, a unified time axis is selected as a global time reference, time offset estimation and interpolation resampling are performed on the original time series of each monitoring channel, and the unified time axis sequence is obtained;

[0015] The terrain point cloud and rock-soil layering are spatially registered under the unified spatial coordinate system, the original point cloud coordinates are mapped to the unified coordinate set through rotation and translation based on three-dimensional rigid body transformation, the registered point cloud and layering interface are consistent with the reference in position, direction and scale, and the spatial registration result is output;

[0016] The unified time axis sequence and the spatial registration result are normalized, and for displacement, inclination, seepage pressure and layering attribute and other multi-source data, de-dimensioning, centralization and variance normalization are respectively performed, to generate the normalized multi-source input set;

[0017] Based on the construction progress time axis and the normalized multi-source input set, each time point on the unified time axis is assigned to a unique stage index according to the stage start and end interval it falls into, the working condition items of the excavation site, the operation layer, the support configuration and the external hydrological conditions of each stage are recorded, the time series data cluster, the spatial element set and the working condition description set aggregated by stage are formed, and the stage working condition library including the stage index, the stage time range, the stage space range and the stage working condition item is output.

[0018] Optionally, the generation of the twin baseline specifically includes:

[0019] Based on the stage space range of the normalized multi-source input set and the stage working condition library, a terrain triangulation network and voxel division of rock-soil layering are generated, independent partition identification of soft interlayer and potential structural plane is performed, and a three-dimensional geometric twin containing grid topology, layering boundary and partition attribute is formed;

[0020] Material properties, contact types of layering interfaces and drainage boundary types are assigned to each partition of the three-dimensional geometric twin, an initial value and constraint set of the physical twin are established, unloading phases and recharge phases in the stage working condition library are respectively mapped to mechanical load sequences and seepage working condition sequences, and physical constraints and working condition inputs are obtained;

[0021] According to the grid and partition list, the unit contribution and boundary conditions of the discrete equation are assembled, the mechanical solver of the stress field and displacement field and the seepage solver of the seepage field and pore pressure field are determined, the solving sequence, convergence criterion and stage advancing strategy are determined, and the finite element simulation scheme is formed;

[0022] The initial field results are obtained by executing the finite element simulation scheme, wherein the mechanical equilibrium satisfies the balance relationship that the overall stiffness formed by the grid assembly is equal to the external load vector after being multiplied by the displacement field on each degree of freedom, and the initial displacement field, initial stress field and initial reaction force set are output by the mechanical solver according to the balance relationship;

[0023] Boundary backstepping is implemented according to the initial displacement field and the observation corresponding relationship in the stage working condition library, the displacement values consistent with the monitoring points are extracted from the initial displacement field through the extraction mapping of the observation position, the displacement values are compared point by point with the observation displacement at the same time to form displacement residuals, the residuals at each position are weighted by using the observation weight, the weighted residuals are back-projected to the model boundary and load degree of freedom through the transposition mode of the extraction mapping, the boundary correction amount is obtained, and the boundary condition and equivalent external load are updated;

[0024] The pore pressure initial value is re-estimated according to the seepage working condition and the updated boundary condition, and the seepage solver is advanced to the stage stable state on the unified time axis, the contributions of the storage coefficient matrix to the change of the pore pressure with time, the contributions of the hydraulic conductivity matrix to the seepage flux, and the influences of the consolidation coupling term on the interaction between displacement and pore pressure are superimposed, and the stage convergence condition is satisfied to obtain the pore pressure field and seepage field after re-equilibrium, which are consistent with the updated displacement field, stress field;

[0025] In the twin baseline generation stage, the finite element simulation is re-executed based on the updated boundary condition and the re-equilibrium pore pressure field and seepage field, the displacement field, stress field, pore pressure field and reaction force set of each stage are summarized according to the time sequence and space range of the stage working condition library, and the twin baseline is formed.

[0026] Optionally, the generation of the high-consistency twin and the assimilation sample set thereof specifically includes:

[0027] Based on the twin consistency-topology constraint excavation twin simulation method, the displacement field, stress field, pore pressure field and reaction force set of the twin baseline are read from the twin consistency assimilation layer, the observation displacement, observation pore pressure and corresponding observation position consistent with the unified time axis are extracted from the stage working condition library, the observation mapping relationship and observation weight are established, and the to-be-assimilated input set containing the model prediction data and measured observation data is formed;

[0028] The observation residual vector is calculated based on the to-be-assimilated input set, and the observation consistency term is constructed, which is a quadratic form measurement with observation residual as independent variable and observation weight as coefficient;

[0029] The physical conservation residual combination vector is calculated according to the mechanical equilibrium condition and seepage continuity condition of the twin baseline, which is composed of mechanical equilibrium residual and mass conservation residual by component parallel, and the physical consistency term is constructed, which is a quadratic form measurement with physical conservation residual as independent variable and physical constraint weight as coefficient;

[0030] The observation consistency term and the physical consistency term are combined to form the dual-consistency assimilation joint target, and the joint update is performed according to the fixed update order, including three steps of state quantity update, boundary condition update and material property update, and the intermediate twin is output;

[0031] The intermediate twin is predicted at the current stage of the unified time axis, and the observation residual and the physical conservation residual are recalculated, and the normalized ratio of the observation residual to the observation data and the normalized ratio of the physical conservation residual to the physical reference scale are judged, and any ratio is higher than the corresponding threshold value, and the joint update is performed according to the fixed update order, and the update results of the state quantity, the boundary condition and the material property in this round are recorded, and when both ratios are not higher than the corresponding threshold value, the iteration in this round is ended, and a plurality of intermediate twin samples are obtained;

[0032] The model state, boundary condition and material property of the plurality of intermediate twin samples are collected and summarized to form an assimilation sample set;

[0033] The statistical mean of displacement, stress and pore pressure is calculated on each degree of freedom of the assimilation sample set, the average state field is generated, the mechanical conservation error, seepage continuity error and slip surface topology connectivity are applied to the average state field, the local results that do not satisfy the constraint conditions are filtered out, the material parameters and boundary conditions corresponding to the high consistency state field obtained by screening are backfilled into the twin main model, and the high consistency twin is reconstructed.

[0034] Optionally, the generation of the slip surface feasible region and the critical region reinforced model specifically includes:

[0035] The displacement field, stress field and pore pressure field of the assimilated high-consistency twin are read in the topological constraint slip surface evolution layer, a joint index field for slip identification is constructed, a candidate high gradient band is formed by taking shear strain, plastic strain energy density and pore pressure gradient as components, and a candidate slip band initial mask is output;

[0036] The connectivity of the candidate slip band initial mask is marked, adjacent fragments are merged according to the connectivity maintenance rule of the topological constraint slip surface evolution, and isolated noise pieces are eliminated, thereby generating a connected component set and a dominant path thereof;

[0037] The curvature distribution and normal consistency distribution of the connected component set are calculated, the high-frequency fluctuation segment is smoothed and cropped according to the curvature threshold and normal consistency threshold, the dominant path meeting the threshold condition is retained, and a curvature-restricted slip path sketch is generated;

[0038] A slip band strip region is generated in the normal band of the curvature-restricted slip path sketch, and the topological correction is performed on the bifurcation and confluence according to the connectivity maintenance rule, thereby outputting a topologically constrained slip band template;

[0039] The critical region adaptive grid refinement is performed in the coverage range of the topologically constrained slip band template, the refinement level is determined according to the error estimation and gradient index, the local encryption is implemented in the band and the band boundary, and the consistent transition with the surrounding grid is maintained, thereby outputting the encrypted grid in the critical region and the mapping relationship;

[0040] Based on the encrypted grid in the critical region, the interface reconstruction is performed on the slip band boundary, the interface type, contact parameter and permeability reduction parameter are established according to the contact boundary processing strategy, and an interface element set is formed;

[0041] On the encrypted grid in the critical region and the interface element set, the displacement field, stress field and pore pressure field of the high-consistency twin are used for consistency correction of the slip band template, the slip occurrence possibility weight of each grid cell is calculated and threshold processing is performed accordingly, and a slip surface feasible region meeting the topological constraint is output;

[0042] The material properties, interface parameters and boundary conditions in the slip surface feasible region are locally strengthened, a critical region strengthening model containing the encrypted grid, interface element and local parameter set is established, and the strengthening model is associated with the high-consistency twin at the structure level.

[0043] Optionally, the generation of the hierarchical early warning result specifically includes:

[0044] The assimilation sample set, the slip surface feasible region and the critical region strengthening model are read in the uncertainty early warning layer, the observed displacement and observed seepage pressure of the corresponding stage are obtained according to the time sequence of the stage working condition library, and the stage input for uncertainty evaluation is formed;

[0045] With the sliding surface feasible region as the spatial weight base, the displacement and seepage pressure samples of the assimilation sample set are statistically weighted in the coverage range to obtain displacement set statistics and seepage pressure set statistics;

[0046] The grid refinement level, interface element distribution and local parameter set are extracted from the critical zone strengthening model, the critical zone scale factor is calculated, the spatial weight and threshold sensitivity in the sliding surface feasible region are adjusted, and the critical zone weight field containing the spatial weight and scale adjustment is output;

[0047] Based on the displacement set statistics, the seepage pressure set statistics and the critical zone weight field, the displacement and seepage pressure uncertainty cone domain is constructed with the set statistical mean value as the center and the set covariance and the critical zone weight field as the measurement boundary, and the joint value range of displacement and seepage pressure under the unified time axis is respectively constrained;

[0048] According to the displacement and seepage pressure uncertainty cone domain, the threshold crossing probability of displacement or seepage pressure exceeding the preset safety threshold in the stage working condition library in the cone domain boundary is calculated by Monte Carlo traversal of the assimilation sample set in the sliding surface feasible region;

[0049] The threshold crossing probability is compared with the hierarchical threshold interval of the stage working condition library to generate a hierarchical early warning result, and the corresponding sliding surface feasible region spatial range, trigger sub-region and time window are marked in the early warning result.

[0050] The beneficial effects of the present application are:

[0051] The present application introduces a double-consistency assimilation layer, a topologically constrained sliding surface evolution device and an uncertainty cone domain construction mechanism in a digital twin slope simulation system, realizes the unity of virtual-real fusion, topological continuity and warning quantization, and forms a closed loop. The double-consistency assimilation layer establishes a parallel structure between observation consistency and physical conservation consistency, and jointly updates state variables, boundary conditions and material properties in a fixed order, so that the twin body converges synchronously in the observation space and the mechanical space, and forms a statistical assimilation sample set. The topologically constrained sliding surface evolution device takes connectivity preservation rules, curvature threshold and interface reconstruction trigger criteria as the core, cooperates sliding path evolution, critical zone adaptive grid refinement and interface element reconstruction, generates a physically continuous sliding surface feasible region and a critical zone strengthening model. The uncertainty warning layer constructs a displacement and seepage pressure uncertainty cone domain based on the assimilation sample set, the sliding surface feasible region and the critical zone scale factor, quantitatively calculates the threshold crossing probability and outputs the hierarchical early warning result. The method effectively avoids the problems of observation fitting distortion, sliding surface fracture and early warning ambiguity in the prior art, and realizes high-consistency simulation and interpretable dynamic early warning of the slope excavation process. BRIEF DESCRIPTION OF DRAWINGS

[0052] The accompanying drawings are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and are used to explain the application without restricting it. In the drawings:

[0053] Fig. 1 A flow chart of a slope excavation dynamic monitoring and early warning method based on digital twinning proposed by the application;

[0054] Fig. 2 A high-consistency twinning body generation schematic diagram of a slope excavation dynamic monitoring and early warning method based on digital twinning proposed by the application. DETAILED DESCRIPTION

[0055] The application will now be described in further detail with reference to the drawings. These drawings are simplified schematic diagrams, and only illustrate the basic structure of the application in a schematic manner, and thus only show the components related to the application.

[0056] REFERENCE Figs. 1-2 A slope excavation dynamic monitoring and early warning method based on digital twinning, comprising the following steps:

[0057] Obtaining terrain point cloud, rock-soil layering and construction progress, completing time alignment and space registration, forming a standardized multi-source input set and establishing a stage working condition library;

[0058] Establishing three-dimensional geometric twinning and physical twinning according to the standardized multi-source input set and the stage working condition library, adopting finite element simulation and completing boundary back propagation and pore pressure rebalancing, forming a twinning baseline;

[0059] Using a double-consistency-topological constraint excavation twinning simulation method, calculating observation residuals and physical conservation residuals in a double-consistency assimilation layer and jointly updating material properties, boundary conditions and state variables to obtain an assimilated high-consistency twinning body and an assimilation sample set;

[0060] Applying a topological constraint slip surface evolutioner to the high-consistency twinning body in a topological constraint slip surface evolution layer and implementing adaptive grid refinement and interface reconstruction in the critical region to obtain a topological constraint-compliant slip surface feasible region and a critical region reinforced model;

[0061] Based on the assimilation sample set and the critical region reinforced model, constructing displacement and seepage pressure uncertainty cone domains in an uncertainty early warning layer, calculating threshold crossing probability and outputting graded early warning results.

[0062] In this embodiment, the generation of the stage working condition library specifically comprises:

[0063] Obtain the terrain point cloud, the rock-soil layering and the construction progress, and perform integrity checking, generate the progress time axis with clear start and end time and corresponding working layer position, form the initial data set containing time mark, spatial coordinate and layering attribute;

[0064] Perform time alignment on the initial data set, select a unified time axis as the global time reference, perform time offset estimation and interpolation resampling on the original time series of each monitoring channel, and obtain the unified time axis sequence;

[0065] Perform spatial registration on the terrain point cloud and the rock-soil layering in the unified spatial coordinate system, map the original point cloud coordinates through rotation and translation based on three-dimensional rigid body transformation to the unified coordinate set, so that the registered point cloud and the layering interface are consistent with the reference in position, direction and scale, and output the spatial registration result;

[0066] Perform normalization processing on the unified time axis sequence and the spatial registration result, perform de-dimensioning, centering and variance normalization on displacement, inclination, seepage pressure and layering attribute and other multi-source data respectively, and generate the normalized multi-source input set;

[0067] Based on the construction progress time axis and the normalized multi-source input set, each time point on the unified time axis is assigned to a unique stage index according to the interval in which it falls, the working condition items of each stage, such as excavation position, working layer position, support configuration and external hydrological condition, are recorded, time series data clusters, spatial element sets and working condition description sets are formed according to stages, and a stage working condition library containing stage index, stage time range, stage space range and stage working condition item is output.

[0068] In the embodiment, the generation of the twin baseline specifically includes:

[0069] Based on the normalized multi-source input set and the stage space range of the stage working condition library, generate the terrain triangulation network and the voxel division of the rock-soil layering, independently partition and identify the soft interlayer and the potential structure surface, and form a three-dimensional geometric twin containing grid topology, layering boundary and partition attribute;

[0070] Assign material properties, contact types of layering interfaces and drainage boundary types to each partition of the three-dimensional geometric twin, establish the initial value and constraint set of the physical twin, map the unloading phase and the recharge phase in the stage working condition library into the mechanical load sequence and the seepage working condition sequence respectively, and obtain the physical constraint and working condition input;

[0071] Assemble the unit contribution and boundary conditions of the discrete equation according to the grid and partition list, determine the mechanical solver of the stress field and the displacement field and the seepage solver of the seepage field and the pore pressure field, determine the solving sequence, convergence criterion and stage advancing strategy, and form a finite element simulation scheme;

[0072] An initial field result is obtained by executing a finite element simulation scheme, wherein mechanical equilibrium satisfies a balance relationship that the total stiffness formed by mesh assembly is equal to the external load vector after being multiplied by the displacement field on each degree of freedom, and a mechanical solver outputs an initial displacement field, an initial stress field and an initial reaction force set according to the balance relationship;

[0073] Boundary back-propagation is implemented according to the initial displacement field and the observation corresponding relationship in the stage working condition library, displacement values consistent with the monitoring points are extracted from the initial displacement field through the extraction mapping of the observation position, and the displacement values are compared with the observation displacement at the same time point to form displacement residuals, the residuals at each position are weighted by using the observation weight, the weighted residuals are back-projected to the model boundary and the load degree of freedom through the transposition mode of the extraction mapping, and boundary correction quantities are obtained, and the boundary conditions and the equivalent external load are updated;

[0074] The initial pore pressure value is re-estimated according to the seepage working condition and the updated boundary condition, and the stage stable state is pushed forward on the unified time axis, the seepage solver adds the source and sink items according to the contribution of the storage coefficient matrix to the change of the pore pressure with time, the contribution of the conductivity coefficient matrix to the seepage flux and the influence of the consolidation coupling term on the interaction between the displacement and the pore pressure, and judges whether the stage convergence condition is satisfied to obtain the pore pressure field and the seepage field after re-equilibrium, and the updated displacement field, the stress field are consistent;

[0075] In the twin baseline generation stage, the finite element simulation is re-executed based on the updated boundary condition and the re-equilibrium pore pressure field and seepage field, the displacement field, the stress field, the pore pressure field and the reaction force set of each stage are summarized according to the time sequence and the space range of the stage working condition library, and the twin baseline is formed.

[0076] In the twin baseline generation stage, the finite element simulation is re-executed based on the updated boundary condition and the re-equilibrium pore pressure field and seepage field, the displacement field, the stress field, the pore pressure field and the reaction force set of each stage are summarized according to the time sequence and the space range of the stage working condition library, and the twin baseline is formed.

[0077] Based on the double-consistency-topology-constrained excavation twin simulation method, the displacement field, the stress field, the pore pressure field and the reaction force set of the twin baseline are read from the double-consistency assimilation layer, the observation displacement, the observation pore pressure and the corresponding observation position consistent with the unified time axis are extracted from the stage working condition library, the observation mapping relationship and the observation weight are established, and the assimilation input set containing the model prediction data and the measured observation data is formed;

[0078] Based on the assimilation input set, an observation residual vector is calculated, and an observation consistency term is constructed, the observation consistency term being a quadratic form measurement with the observation residual as the independent variable and the observation weight as the coefficient;

[0079] A physical conservation residual combination vector is calculated according to the mechanical equilibrium condition and the seepage continuity condition of the twin baseline, the physical conservation residual combination vector being composed of the mechanical equilibrium residual and the mass conservation residual in parallel by components, and a physical consistency term is constructed, the physical consistency term being a quadratic form measurement with the physical conservation residual as the independent variable and the physical constraint weight as the coefficient;

[0080] The observation consistency term and the physical consistency term are combined to form a double-consistency assimilation joint objective, and a joint update is performed according to a fixed update order, including three steps of state quantity update, boundary condition update and material property update, and an intermediate twin is output;

[0081] The state quantity update is obtained by incrementally solving the gradient direction of the joint objective with respect to the model state, and the state quantity update increment is obtained and superimposed on the current displacement field and pore pressure field, the boundary condition update is obtained by incrementally solving the gradient direction of the joint objective with respect to the boundary parameter, and the boundary correction quantity is obtained and the boundary displacement, boundary load and drainage boundary are updated, and the material property update is obtained by incrementally solving the gradient direction of the joint objective with respect to the material parameter, and the material property update increment is obtained and the elastic modulus, cohesion, internal friction angle and permeability coefficient are updated;

[0082] The intermediate twin is executed at the current stage of the unified time axis, and the observation residual and the physical conservation residual are recalculated, and the normalized ratio of the observation residual to the observation data and the normalized ratio of the physical conservation residual to the physical reference scale are determined, and when any ratio is higher than the corresponding threshold value, the joint update is performed according to the fixed update order, and the update results of the state quantity, the boundary condition and the material property in this round are recorded, and when both ratios are not higher than the corresponding threshold value, the iteration in this round is ended, and a plurality of intermediate twin samples are obtained;

[0083] The model state, the boundary condition and the material property of the plurality of intermediate twin samples are collected, and the assimilation sample set is formed by summarizing;

[0084] The statistical mean of the displacement, the stress and the pore pressure on each degree of freedom is calculated for the assimilation sample set, and the average state field is generated, and the mechanical conservation error, the seepage continuity error and the slip surface topological connectivity are three constraint conditions applied to the average state field, and the local results that do not satisfy the constraint conditions are filtered out, and the material parameters and the boundary conditions corresponding to the high-consistency state field selected are backfilled into the twin main model, and the high-consistency twin is reconstructed.

[0085] In the embodiment, the generation of the slip surface feasible region and the critical zone reinforced model specifically includes:

[0086] The displacement field, the stress field and the pore pressure field of the assimilated high-consistency twin are read in the topological constraint slip surface evolution layer, and a joint index field for slip identification is constructed, and a candidate high-gradient zone is formed by taking the shear strain, the plastic strain energy density and the pore pressure gradient as components, and an initial mask of the candidate slip zone is output;

[0087] The connectivity annotation is performed on the initial mask of the candidate slip zone, and according to the connectivity maintenance rule of the topological constraint slip surface evolution, adjacent fragments are merged according to the voxel or unit adjacency relationship, and isolated noise pieces are eliminated, and a connected component set and a dominant path are generated.

[0088] The curvature distribution and the normal consistency distribution of the connected component set are calculated, the high-frequency fluctuation section is smoothed and trimmed according to the curvature threshold and the normal consistency threshold, the dominant path satisfying the threshold condition is retained, and a curvature-restricted slip path sketch is generated;

[0089] The slip band strip region is generated in the normal band of the curvature-restricted slip path sketch, the topological correction is performed on the bifurcation and confluence according to the connectivity preservation rule, and the topologically constrained slip band template is output;

[0090] The critical region adaptive grid refinement is performed in the topologically constrained slip band template coverage, the refinement level is determined according to the error estimation and the gradient index, the local encryption is implemented in the band and the band boundary, and the consistent transition with the surrounding grid is maintained, and the encrypted grid and the mapping relationship in the critical region are output;

[0091] Based on the encrypted grid in the critical region, the interface reconstruction is performed on the slip band boundary, the interface type, the contact parameter and the permeability reduction parameter are established according to the processing strategy of the contact boundary, and the interface element set is formed;

[0092] On the encrypted grid in the critical region and the interface element set, the displacement field, the stress field and the pore pressure field of the high-consistency twin body are used to correct the slip band template, the slip occurrence possibility weight of each grid element is calculated, and the threshold processing is performed accordingly, and the slip surface feasible region conforming to the topological constraint is output;

[0093] Taking the slip surface feasible region as the core region, the material properties, the interface parameters and the boundary conditions in the core region are locally strengthened, the critical region strengthening model containing the encrypted grid, the interface element and the local parameter set is established, and the strengthening model is associated with the high-consistency twin body at the structure level.

[0094] In the embodiment, the generation of the hierarchical early warning result specifically includes:

[0095] The assimilation sample set, the slip surface feasible region and the critical region strengthening model are read at the uncertainty early warning layer, the observed displacement and the observed seepage pressure of the corresponding stage are obtained according to the time sequence of the stage working condition library, and the stage input for uncertainty evaluation is formed;

[0096] Taking the slip surface feasible region as the spatial weight base, the displacement and seepage pressure samples of the assimilation sample set are spatially weighted and counted in the coverage range, and the displacement set statistic and the seepage pressure set statistic are obtained;

[0097] The grid refinement level, the interface element distribution and the local parameter set are extracted from the critical region strengthening model, the critical region scale factor is calculated, the spatial weight and the threshold sensitivity in the slip surface feasible region are adjusted, and the critical region weight field containing the spatial weight and the scale adjustment is output;

[0098] Based on the displacement set statistics, the osmotic pressure set statistics and the critical zone weight field, a displacement and osmotic pressure uncertainty cone domain is constructed with the set statistical mean value as the center and the set covariance and the critical zone weight field as the measurement boundary, to respectively constrain the joint value range of displacement and osmotic pressure under the unified time axis;

[0099] According to the displacement and osmotic pressure uncertainty cone domain, the assimilation sample set is used to perform Monte Carlo traversal in the sliding surface feasible region to calculate the threshold crossing probability of displacement or osmotic pressure exceeding the preset safety threshold of the stage working condition library in the cone domain boundary.

[0100] The threshold crossing probability is compared with the hierarchical threshold interval of the stage working condition library to generate a hierarchical early warning result, and the corresponding sliding surface feasible region spatial range, trigger sub-region and time window are marked in the early warning result.

[0101] Embodiment 1:

[0102] In order to verify the feasibility of the application in implementation, the application is applied to the whole process dynamic monitoring and early warning work of a mountainous highway slope excavation. The geological conditions of this area are complex, the rock-soil layering is significant, the soft interlayer is developed, and the groundwater activity is frequent. There is a large deviation between the conventional finite element prediction result and the measured deformation, the slip zone position recognition is unstable, the early warning result is delayed, and the false positive rate is high. Therefore, the digital twin-based slope excavation dynamic monitoring and early warning method of the application is introduced to unify modeling and dynamically update the data flow, model evolution and risk assessment during excavation.

[0103] Multi-source sensing devices are laid out on site to continuously collect topographic point cloud, rock-soil layering, support parameters and pore pressure monitoring data. After time alignment, spatial registration and standardization processing, all data are input into the stage working condition library to drive the real-time evolution of the digital twin model. In the twin body construction stage, through the coordinated establishment of geometric twins and physical twins, the fine restoration of rock mass layering, interface contact and pore pressure distribution is realized. The initial displacement field and pore pressure field output by the finite element simulation are formed after boundary back propagation and rebalancing to form the twin baseline, which provides accurate initial conditions for subsequent double-consistency assimilation.

[0104] During excavation, the double-consistency assimilation mechanism proposed by the application is used to calculate the observation residual and the physical conservation residual respectively, form the joint structure of the observation consistency term and the physical consistency term, and through the synchronous update of state variables, boundary conditions and material properties, the virtual and real models realize collaborative convergence in the observation space and the physical space. The process automatically generates an assimilation sample set and forms a high-consistency twin body through statistics, providing reliable input for the sliding surface evolution.

[0105] Based on high consistency twin, the sliding zone is identified in the topological constraint slip surface evolution layer. Through the connectivity preservation rule and curvature threshold constraint, the slip surface is kept continuous and smooth in the topological structure, eliminating the false fracture and multi-branching problem commonly seen in traditional methods. After adaptive mesh refinement in the critical zone, the interface reconstruction generates an interface element set consistent with the actual sliding contact characteristics, forming a critical zone reinforcement model to provide a local high-resolution basis for subsequent uncertainty analysis.

[0106] In the early warning stage, the system constructs the displacement and seepage pressure uncertainty cone based on the assimilation sample set and the slip surface feasible region, and adjusts the spatial weight by using the critical zone scale factor to realize the probability quantization of the potential instability area. The threshold crossing probability is calculated to automatically generate a hierarchical warning result, showing the possible evolution range of the slip surface, the risk level and the time window. The actual application shows that this method can identify the potential sliding trend in advance, the prediction result is in good agreement with the measured deformation, the warning accuracy is significantly improved, and the virtual and real models remain stable and consistent for a long time.

[0107] In summary, the present application realizes the deep integration of digital twin and field monitoring under complex geological conditions, breaks through the limitations of fitting distortion, slip surface topological fracture and qualitative warning in traditional simulation prediction, and achieves the unified goal of high consistency simulation, topologically constrained evolution and quantitative warning, providing a reliable and traceable intelligent technical path for slope safety monitoring.

[0108] Table 1 Performance comparison of slope excavation dynamic monitoring and early warning method based on digital twin and traditional method

[0109]

[0110] As can be seen from Table 1, the double consistency-topologically constrained excavation twin simulation method (DCA-TSE) of the present application is significantly better than the traditional method in many key indicators. First, in terms of displacement prediction and pore pressure prediction error, the average error of the present application is reduced to about one third of the traditional finite element method, which shows that the double consistency assimilation mechanism effectively suppresses the error accumulation caused by simple observation fitting, so that the virtual and real models converge in the observation space and the physical space at the same time, and the prediction result is closer to the field measurement.

[0111] Secondly, in terms of slip surface identification accuracy and topological connectivity integrity, the DCA-TSE method relies on the topologically constrained slip surface evolution, realizes the geometric continuity and physical rationality of the slip surface through connectivity preservation and curvature threshold control, reduces the spatial deviation of the slip surface from 1.35 meters to 0.32 meters, and improves the connectivity integrity of the sliding zone to more than 97%. This improvement directly improves the geometric expression quality of the sliding area and avoids the false fracture problem commonly seen in traditional threshold identification.

[0112] In terms of early warning timeliness and accuracy, the uncertainty cone field based on assimilation sample set and critical zone scale factor is constructed, so that the risk assessment evolves from a single indicator to a joint probability model. The quantitative calculation of threshold crossing probability significantly improves the advance and reliability of early warning, which can output stable early warning signal about 8 hours earlier than the traditional method, and the early warning accuracy reaches more than 96%, providing sufficient response time for construction scheduling.

[0113] In addition, the application also performs excellently in virtual-real consistency index and calculation convergence efficiency. Through hierarchical updating and constraint screening mechanism, the model maintains the balance state of physical conservation and observation consistency after multiple rounds of assimilation iteration, and the overall virtual-real consistency is improved by about 13%, and the calculation convergence efficiency is improved to 2.2 times, which shows that it not only has higher accuracy, but also has more stable calculation process.

[0114] In summary, the DCA-TSE method of the application forms a fusion mode of data-driven and mechanism-driven through three innovation links of double consistency assimilation, topological constraint sliding surface evolution and uncertainty quantification, significantly improves the accuracy, real-time performance and interpretability of slope excavation monitoring and early warning, and provides high reliability technical support for safety management under complex geological environment.

[0115] The above is only the preferred specific embodiment of the application, but the protection scope of the application is not limited thereto, any person skilled in the art can make equivalent replacement or change according to the technical scheme and inventive concept of the application within the technical range disclosed by the application, which should be covered within the protection scope of the application.

Claims

1. A method for dynamic monitoring and early warning of slope excavation based on digital twinning, characterized in that, The method comprises the following steps: Obtaining terrain point cloud, stratification of rock-soil and construction progress, completing time alignment and space registration, forming a standardized multi-source input set and establishing a stage working condition library; According to the standardized multi-source input set and the stage working condition library, three-dimensional geometric twin and physical twin are established, finite element simulation is adopted to complete boundary back propagation and pore pressure rebalancing, and a twin baseline is formed; Using a double-consistency-topology-constrained excavation twin simulation method, the observation residual and the physical conservation residual are calculated in the double-consistency assimilation layer, and the material properties, boundary conditions and state variables are jointly updated, to obtain an assimilated high-consistency twin and an assimilation sample set; In the topology-constrained slip surface evolution layer, a topology-constrained slip surface evolution device is applied to the high-consistency twin, and adaptive grid refinement and interface reconstruction in the critical region are implemented, to obtain a slip surface feasible region and a critical region reinforcement model conforming to the topology constraint; In the uncertainty early warning layer, a displacement and seepage pressure uncertainty cone domain is constructed based on the assimilation sample set and the critical region reinforcement model, a threshold crossing probability is calculated, and a graded early warning result is output.

2. The method according to claim 1, wherein, The generation of the stage working condition library specifically comprises: Obtaining terrain point cloud, stratification of rock-soil and construction progress and performing integrity check, generating a progress time axis with clear start and end time and corresponding working horizon for the construction progress, forming an initial data set containing time markers, spatial coordinates and stratification attributes; Performing time alignment on the initial data set, selecting a unified time axis as a global time reference, and performing time offset estimation and interpolation resampling on the original time series of each monitoring channel to obtain a unified time axis sequence; Performing space registration on the terrain point cloud and the stratification of rock-soil in a unified spatial coordinate system, mapping the original point cloud coordinates to a unified coordinate set through rotation and translation based on a three-dimensional rigid body transformation, making the registered point cloud and stratification interface consistent with the reference in position, direction and scale, and outputting a space registration result; Performing normalization processing on the unified time axis sequence and the space registration result, performing de-dimensioning, centering and variance normalization on the multi-source data of displacement, inclination, seepage pressure and stratification attributes respectively, and generating a standardized multi-source input set; Based on the construction progress time axis and the standardized multi-source input set, each time point on the unified time axis is assigned to a unique stage index according to the stage start and end interval it falls into, the working condition items of each stage, such as excavation site, working horizon, support configuration and external hydrological conditions, are recorded, time series data clusters, spatial element sets and working condition description sets are aggregated by stage, and a stage working condition library containing stage index, stage time range, stage space range and stage working condition items is output.

3. The method according to claim 1, characterized in that, The generation of the twin baseline specifically comprises: Based on the standardized multi-source input set and the stage space range of the stage working condition library, a terrain triangulation network and a voxel division of rock-soil stratification are generated, soft intercalated layers and potential structural planes are independently partitioned and identified, and a three-dimensional geometric twin containing grid topology, stratification boundary and partition attribute is formed; Material properties, contact types of layered interfaces and drainage boundary types are assigned to each sub-zone of the three-dimensional geometric twin, a set of initial values and constraints of the physical twin are established, and the unloading phase and the recharge phase in the stage working condition library are respectively mapped to the mechanical load sequence and the seepage working condition sequence to obtain the physical constraint and working condition input; The unit contribution of the discrete equation and the boundary condition are assembled according to the grid and the sub-zone list to determine the mechanical solver of the stress field and the displacement field and the seepage solver of the seepage field and the pore pressure field, determine the solving sequence, the convergence criterion and the stage advancing strategy, and form the finite element simulation scheme; The initial field results are obtained by executing the finite element simulation scheme, wherein the mechanical equilibrium satisfies the balance relationship that the total stiffness formed by the grid assembly is equal to the external load vector after being multiplied by the displacement field on each degree of freedom, and the initial displacement field, the initial stress field and the initial reaction force set are output by the mechanical solver according to the balance relationship; The boundary back-propagation is implemented according to the initial displacement field and the observation corresponding relationship in the stage working condition library, the displacement values consistent with the monitoring points are extracted from the initial displacement field through the extraction mapping of the observation position, the displacement residual is formed by comparing the displacement values point by point with the observation displacement at the same time, the residual at each position is weighted by using the observation weight, the weighted residual is back-projected to the model boundary and the load degree of freedom through the transposition mode of the extraction mapping to obtain the boundary correction, and the boundary condition and the equivalent external load are updated; The pore pressure initial value is re-estimated according to the seepage working condition and the updated boundary condition, and the stage stable state is advanced on the unified time axis, the seepage solver adds the source and sink items according to the contribution of the storage coefficient matrix to the change of the pore pressure with time, the contribution of the hydraulic conductivity matrix to the seepage flux and the influence of the consolidation coupling term on the interaction between the displacement and the pore pressure, judges whether the stage convergence condition is satisfied, and obtains the pore pressure field and the seepage field after re-equilibrium, which are consistent with the updated displacement field, the stress field; In the twin baseline generation stage, the finite element simulation is re-executed based on the updated boundary condition and the re-equilibrium pore pressure field and seepage field, the displacement field, the stress field, the pore pressure field and the reaction force set of each stage are summarized according to the time sequence and the space range of the stage working condition library to form the twin baseline.

4. The method according to claim 1, wherein, The generation of the high-consistency twin and the assimilation sample set thereof specifically includes: Based on the double-consistency-topology-constrained excavation twin simulation method, the displacement field, the stress field, the pore pressure field and the reaction force set of the twin baseline are read from the double-consistency assimilation layer, the observation displacement, the observation pore pressure and the corresponding observation position consistent with the unified time axis are extracted from the stage working condition library, the observation mapping relationship and the observation weight are established, and the to-be-assimilated input set containing the model predicted data and the measured observation data is formed; Based on the to-be-assimilated input set, the observation residual vector is calculated, and the observation consistency term is constructed, which is a quadratic form measurement with the observation residual as the independent variable and the observation weight as the coefficient; According to the mechanical equilibrium condition and the seepage continuity condition of the twin baseline, a physical conservation residual combination vector is calculated, which is composed of the mechanical equilibrium residual and the mass conservation residual by component juxtaposition, and a physical consistency term is constructed, which is a quadratic form measurement with the physical conservation residual as the independent variable and the physical constraint weight as the coefficient; The observation consistency term and the physical consistency term are combined to form a dual-consistency assimilation joint target, and a joint update is performed according to a fixed update order, including three steps of state quantity update, boundary condition update and material property update, and an intermediate twin is output; The intermediate twin is executed in the current stage of the unified time axis to perform prediction and recalculate the observation residual and the physical conservation residual, and the normalized ratio of the observation residual to the observation data and the normalized ratio of the physical conservation residual to the physical reference scale are judged, and when any one of the two ratios is higher than the corresponding threshold value, the joint update is performed according to the fixed update order, and the update results of the state quantity, the boundary condition and the material property in this round are recorded, and when both of the two ratios are not higher than the corresponding threshold value, the iteration in this round is ended, and a plurality of intermediate twin samples are obtained; The model state, boundary condition and material property of the plurality of intermediate twin samples are collected and summarized to form an assimilation sample set; The statistical mean values of displacement, stress and pore pressure are calculated on each degree of freedom of the assimilation sample set to generate an average state field, and three constraint conditions of mechanical conservation error, seepage continuity error and slip surface topological connectivity are applied to the average state field, and the local results not satisfying the constraint conditions are screened out, and the material parameters and boundary conditions corresponding to the high-consistency state field screened out are backfilled into the twin main model to reconstruct a high-consistency twin.

5. The method according to claim 1, wherein, The generation of the slip surface feasible region and the critical zone reinforced model specifically includes: The displacement field, stress field and pore pressure field of the assimilated high-consistency twin are read in the topological constraint slip surface evolution layer, a joint index field for slip identification is constructed, a candidate high gradient zone is formed by taking shear strain, plastic strain energy density and pore pressure gradient as components, and a candidate slip zone initial mask is output; The connectivity of the candidate slip zone initial mask is labeled, and according to the connectivity preservation rule of the topological constraint slip surface evolution, adjacent fragments are merged according to the voxel or element adjacency relationship, and isolated noise fragments are eliminated to generate a connected component set and its dominant path; The curvature distribution and normal consistency distribution of the connected component set are calculated, the high-frequency undulation segment is smoothed and cropped according to the curvature threshold and the normal consistency threshold, the dominant path satisfying the threshold condition is retained, and a curvature-restricted slip path sketch is generated; According to the curvature-restricted slip path sketch, a slip zone strip region is generated in its normal band, and the topological correction is performed on the bifurcation and confluence according to the connectivity preservation rule, and a topologically constrained slip zone template is output; The critical zone adaptive mesh refinement is performed within the coverage range of the topologically constrained slip zone template, the refinement level is determined according to the error estimation and the gradient index, the local encryption is implemented within the slip zone and the slip zone boundary, and the consistent transition with the surrounding mesh is maintained, and the encrypted mesh and the mapping relationship of the critical zone are output; Based on the mesh of the critical zone after encryption, the interface reconstruction is performed on the slip zone boundary, the interface type, contact parameter and permeability reduction parameter are established according to the processing strategy of the contact boundary, and an interface element set is formed. On the encrypted grid and interface element set in the critical zone, the displacement field, stress field and pore pressure field of the high consistency twin are used to correct the slip zone template, calculate the slip possibility weight of each grid element and perform threshold processing accordingly, and output the feasible slip surface region that meets the topological constraints; Taking the feasible slip surface region as the core area, the material properties, interface parameters and boundary conditions in it are locally strengthened, the critical zone strengthening model containing encrypted grid, interface element and local parameter set is established, and the strengthening model is structurally associated with the high consistency twin.

6. The method according to claim 1, wherein, The generation of the hierarchical warning result specifically includes: In the uncertainty warning layer, read the assimilation sample set, the feasible slip surface region and the critical zone strengthening model, and obtain the observed displacement and observed seepage pressure of the corresponding stage according to the time sequence of the stage working condition library to form the stage input for uncertainty evaluation; Taking the feasible slip surface region as the spatial weight base, the displacement and seepage pressure samples of the assimilation sample set are spatially weighted and statistically analyzed within its coverage range to obtain the displacement set statistics and seepage pressure set statistics; Extract the grid refinement level, interface element distribution and local parameter set from the critical zone strengthening model, calculate the critical zone scale factor, adjust the spatial weight and threshold sensitivity in the feasible slip surface region, and output the critical zone weight field containing spatial weight and scale adjustment; Based on the displacement set statistics, seepage pressure set statistics and critical zone weight field, taking the set statistical mean as the center and the set covariance and critical zone weight field as the measurement boundary, the displacement and seepage pressure uncertainty cone domain is constructed to constrain the joint value range of displacement and seepage pressure on the unified time axis; According to the displacement and seepage pressure uncertainty cone domain, the assimilation sample set is used to perform Monte Carlo traversal in the feasible slip surface region to calculate the threshold crossing probability of displacement or seepage pressure exceeding the preset safety threshold of the stage working condition library within the cone domain boundary; Compare the threshold crossing probability with the hierarchical threshold interval of the stage working condition library to generate the hierarchical warning result, and mark the corresponding feasible slip surface region spatial range, trigger sub-region and time window in the warning result.

Images