A method and system for analyzing influence of complex loose bodies based on tunneling simulation

Abstract

The application discloses a kind of complex loose body influence analysis method and system based on tunneling simulation, it is related to engineering simulation technical field, including, obtaining geological space data and construction state data in target analysis area, construct basic microcosmic voxel dataset in simulation three-dimensional space, and aggregate to form macroscopic voxel unit;Based on simulation shield propulsion distance, historical simulation boundary state and spatial attenuation relationship generates simulation dynamic boundary parameter, further forms simulated impact load gradient field and determines the calculation candidate area;To the calculation candidate area, execute pre-probing cleavage judgment, when meeting cleavage condition, call basic microcosmic voxel unit to carry out local collision search and damage calculation, not satisfied when maintaining macroscopic rapid propagation;Damage result is written back again and the distribution of simulation influence area is output.Reduces the overhead of global calculation, improves the identification accuracy of complex loose body local instability.

Description

Technical Field

[0001] This invention relates to the field of engineering simulation technology, and in particular to a method and system for analyzing the impact of complex loose bodies based on tunneling simulation. Background Technology

[0002] During tunnel boring machine (TBM) construction, a complex loose mass often exists ahead of the tunnel face, composed of gravel, sand, weak interlayers, and groundwater disturbance zones. This type of stratum exhibits significant differences in porosity, strength, density, and water content. Under the influence of construction disturbances such as cutterhead thrust, torque, tunnel face pressure, and grouting pressure, it is prone to local compaction, particle rearrangement, energy concentration, and instability propagation. Existing construction impact analysis methods largely rely on global continuous medium models, finite element meshes, or discrete element global particle calculations. Continuous medium models struggle to represent local abrupt changes and non-uniform damage evolution within the loose mass. While global fine particle calculations offer high accuracy, they are computationally intensive and cannot meet the needs of simulation analysis under multiple working conditions and time steps. Some adaptive mesh methods typically refine the mesh locally based on errors or gradients, lacking a coordinated assessment of the continuous migration of the TBM advance boundary, the sensitivity of the loose mass to disturbances, and the triggering conditions for local fracturing. This can easily lead to excessive invalid calculation areas or omission of high-risk regions. Summary of the Invention

[0003] In view of the aforementioned existing problems, the present invention is proposed.

[0004] Therefore, this invention provides a method for analyzing the impact of complex loose bodies based on tunneling simulation, which solves the problems of insufficient continuous migration of boundary conditions, large computational load of full-domain fine simulation, and inaccurate identification of high-risk areas of local instability in existing shield tunneling construction environment analysis of complex loose bodies, which make it difficult to balance simulation efficiency and prediction accuracy.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] In a first aspect, the present invention provides a method for analyzing the impact of complex loose bodies based on tunneling simulation, which includes acquiring geological spatial data and construction status data of complex loose bodies within the target analysis area, constructing a basic micro voxel dataset in the simulation three-dimensional space, and aggregating macro voxel units based on spatial adjacency and similarity of geological parameters.

[0007] Based on the simulated shield tunneling distance, historical simulation boundary states, and spatial attenuation relationships, a spatiotemporal sliding window boundary mapping is performed to generate the simulation dynamic boundary parameters at the current simulation time step.

[0008] The simulated impact load gradient field is generated based on the simulated dynamic boundary parameters, and the precise candidate region is determined by combining the sensitive characteristics of loose body disturbance.

[0009] The macroscopic voxel units within the actuarial candidate region are pre-explored for fusion judgment. When the fusion condition is met, the corresponding basic microscopic voxel units are called to perform local collision search and damage calculation. When the fusion condition is not met, the macroscopic voxel units are maintained to perform macroscopic fast propagation calculation.

[0010] The damage calculation results are written back to update the corresponding macroscopic voxel elements, and the continuous and stable basic microscopic voxel elements are frozen according to the cumulative damage state. The simulation influence region distribution of complex loose bodies under impact load is generated by combining the macroscopic rapid propagation results.

[0011] As a preferred embodiment of the complex loose body influence analysis method based on tunneling simulation described in this invention, the geological spatial data includes the porosity, bulk mass, uniaxial compressive strength, internal friction angle, damping parameters, groundwater disturbance parameters, and spatial density distribution input to the simulation model.

[0012] The construction status data includes the shield tunneling direction, tunneling speed, cutterhead thrust, cutterhead torque, cutterhead rotation speed, face pressure, and grouting pressure input into the simulation model.

[0013] The formation of macroscopic voxel units includes dividing the simulation 3D space corresponding to the complex loose body into basic microscopic voxel units, and using the basic microscopic voxel units as the smallest spatial index units and parameter storage units to record model spatial coordinates and geological parameters, and setting an initial inactive computation state for each basic microscopic voxel unit; based on the model spatial adjacency relationship, geological parameter similarity, and disturbance sensitivity feature similarity between basic microscopic voxel units, multiple basic microscopic voxel units are aggregated into macroscopic voxel units, and the macroscopic voxel units are used as the default simulation computation objects, while establishing a split index relationship between the macroscopic voxel units and the corresponding basic microscopic voxel units.

[0014] As a preferred embodiment of the complex loose body influence analysis method based on tunneling simulation described in this invention, the generation of simulation dynamic boundary parameters at the current simulation time step includes: in the initial calculation stage of the simulation model, assigning initial dynamic boundary parameters to the macroscopic voxel elements around the simulation shield face according to the construction status data and the position of the simulation shield face in the simulation three-dimensional space; and in the simulation iteration stage, calculating the simulation shield advance distance between the current simulation time step and the previous simulation time step according to the advance speed and the set simulation time step.

[0015] Based on the simulated shield tunneling distance and the shield tunneling direction, a corresponding historical mapping neighborhood is determined for the macroscopic voxel unit in front of the simulated shield tunneling in the simulated three-dimensional space; according to the model space distance between the historical macroscopic voxel unit in the historical mapping neighborhood and the macroscopic voxel unit in front of the tunneling, combined with the spatial attenuation relationship and the equivalent propagation hindrance characteristics, the mapping weight of the boundary parameters transmitted from the historical macroscopic voxel unit to the macroscopic voxel unit in front of the tunneling is calculated.

[0016] Based on the mapping weights, the historical simulation boundary states of each historical macro voxel unit in the historical mapping neighborhood at the previous simulation time step are weighted and fused. Combined with the boundary transfer correction relationship determined by the geological parameters and the newly added simulation construction boundary effect correction amount, the simulation dynamic boundary parameters of the macro voxel unit in front of the tunneling are generated at the current simulation time step.

[0017] As a preferred embodiment of the complex loose body influence analysis method based on tunneling simulation described in this invention, the generation of the simulated impact load gradient field includes: extracting boundary action components characterizing the simulated impact load based on the simulation dynamic boundary parameters at the current simulation time step; weighting and summing the boundary action components based on dimensional normalized weights to calculate the load scalar of each macroscopic voxel element; calculating the difference of the load scalar based on the adjacency relationship in the model space to generate the load gradient vector corresponding to each macroscopic voxel element, and calculating the magnitude of the load gradient vector to generate the corresponding load gradient intensity; the load gradient vector and the load gradient intensity together constitute the simulated impact load gradient field.

[0018] The determination of the actuarial candidate region includes: forming a spatial gradient characterization vector based on the load gradient intensity, the projection component of the load gradient vector in the shield advancement direction, and the model spatial distance to the simulated shield face; forming a loose body attribute characterization vector based on the porosity, normalized bulk mass, strength weakening characteristics, and structural inhomogeneity of the macroscopic voxel units; performing calibration mapping on the spatial gradient characterization vector and the loose body attribute characterization vector respectively to obtain a spatial gradient mapping vector and an attribute mapping vector; performing normalization processing on the macroscopic voxel units participating in the screening based on the correlation between the spatial gradient mapping vector and the attribute mapping vector to obtain the normalized correlation weight of each macroscopic voxel unit; and determining the macroscopic voxel units whose normalized correlation weight reaches a preset screening threshold as actuarial candidate regions that need to undergo preliminary fracture judgment.

[0019] As a preferred embodiment of the complex loose body impact analysis method based on tunneling simulation described in this invention, the step of pre-exploratory fracturing judgment of macroscopic voxel units within the actuarial candidate area includes: reading the simulation dynamic boundary parameters, normalized correlation weights, equivalent propagation hindrance characteristics, and structural inhomogeneity of the macroscopic voxel units within the actuarial candidate area; determining the simulated impact disturbance intensity entering the macroscopic voxel unit based on the simulation dynamic boundary parameters, and performing pre-exploratory attenuation calculation on the simulated impact disturbance intensity in combination with the equivalent propagation hindrance characteristics to obtain the pre-exploratory energy deposition results of the macroscopic voxel units; identifying local parameter abrupt change regions within the macroscopic voxel units based on the structural inhomogeneity and the differences in geological parameters between basic microscopic voxel units, and using the pre-exploratory energy deposition results corresponding to the local parameter abrupt change regions as the basis for prioritizing fracturing judgment.

[0020] When the preliminary energy deposition result reaches the equivalent failure condition of the macro voxel unit and the normalized correlation weight reaches the preset actuarial priority condition, the macro voxel unit is determined to meet the pyrolysis condition, and the corresponding basic micro voxel unit is called according to the pyrolysis index relationship; when the preliminary energy deposition result does not reach the equivalent failure condition of the macro voxel unit, or the normalized correlation weight does not reach the preset actuarial priority condition, the macro voxel unit is determined not to meet the pyrolysis condition, and the macro voxel unit continues to perform macro fast propagation calculation.

[0021] As a preferred embodiment of the complex loose body influence analysis method based on tunneling simulation described in this invention, the following steps for performing local collision search and damage calculation include: reading the model spatial coordinates, geological parameters, and disturbance sensitivity features of the corresponding basic micro voxel unit according to the rift index relationship; switching the corresponding basic micro voxel unit from an inactive calculation state to a locally refined calculation state; allocating local incident disturbance amounts based on the pre-exploration energy deposition results and the model spatial position of the basic micro voxel unit; determining the local collision search range based on the model spatial adjacency relationship between basic micro voxel units, the local incident disturbance amounts, and the disturbance sensitivity features; and within the local collision search range, calculating the local compression state, disturbance absorption state, and damage accumulation state of the basic micro voxel unit; and determining the damage calculation results of the basic micro voxel unit based on the local compression state, disturbance absorption state, and damage accumulation state.

[0022] As a preferred embodiment of the complex loose body impact analysis method based on tunneling simulation described in this invention, the step of writing back the damage calculation results to the corresponding macroscopic voxel unit includes: writing back the damage calculation results and updated geological parameters of the basic microscopic voxel unit to the macroscopic voxel unit; recalculating the equivalent propagation hindrance characteristics, disturbance sensitivity statistical results, and structural inhomogeneity of the macroscopic voxel unit based on the write-back results; when the basic microscopic voxel unit in the local fine calculation state does not reach the instability evolution condition within a consecutive preset number of simulation time steps, and the macroscopic voxel unit does not meet the rupture condition again, the corresponding basic microscopic voxel unit is switched back to the inactive calculation state, and the local fine calculation of the corresponding basic microscopic voxel unit is stopped; basic microscopic voxel units whose cumulative damage state reaches the failure judgment threshold are extracted, three-dimensional spatial connectivity identification is performed on them, and the simulation impact area distribution is output in combination with the macroscopic rapid propagation results.

[0023] Secondly, this invention provides a complex loose body impact analysis system based on tunneling simulation, comprising: a voxel modeling module for acquiring geological spatial data and construction status data, constructing a basic micro voxel dataset in the simulation three-dimensional space, and aggregating it to form macro voxel units; a boundary mapping module for generating simulation dynamic boundary parameters at the current simulation time step based on the simulated shield tunneling distance, historical simulation boundary states, and spatial attenuation relationships; a candidate area determination module for generating a simulated impact load gradient field based on the simulation dynamic boundary parameters, and determining the precise candidate area by combining the loose body disturbance sensitivity characteristics; a pre-exploration lysis module for performing pre-exploration lysis judgment on the macro voxel units within the precise candidate area, and calling the corresponding basic micro voxel units when the lysis conditions are met; a local precise calculation module for performing local collision search and damage calculation on the called basic micro voxel units; and a result output module for writing back the damage calculation results to update the corresponding macro voxel units, performing freezing processing on continuous and stable basic micro voxel units, and generating the simulated impact area distribution of complex loose bodies under impact loads.

[0024] Thirdly, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, it implements any step of the method for analyzing the influence of complex loose bodies based on tunneling simulation as described in the first aspect of the present invention.

[0025] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the method for analyzing the influence of complex loose bodies based on tunneling simulation as described in the first aspect of the present invention.

[0026] The beneficial effects of this invention are as follows: By constructing a basic micro-voxel dataset in the simulated three-dimensional space, and using the basic micro-voxel units as spatial indices and parameter storage objects before aggregating them to form macro-voxel units, the initial simulation stage can use macro-voxel units as the default calculation objects, avoiding the waste of computing power caused by the full-domain micro-fine calculation of complex loose bodies. Through spatiotemporal sliding window boundary mapping, the historical simulation boundary states are continuously migrated to the area ahead of the tunnel as the simulated shield advances, improving the continuity and consistency of boundary conditions between different simulation time steps. By correlating the simulated impact load gradient field with the loose body's disturbance sensitivity characteristics, macro-voxel units with drastic impact load changes, fragile geological conditions, and the risk of local instability can be preferentially identified. Through pre-exploration rift judgment, the corresponding basic micro-voxel unit is called only when the rift condition is met to perform local collision search and damage calculation, balancing computational efficiency and local analysis accuracy. Through damage result write-back and freeze processing, dynamic switching and closed-loop updating between macro-voxels and basic micro-voxels are achieved, thereby more accurately generating the simulated influence area distribution of complex loose bodies under impact loads. Attached Figure Description

[0027] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 This is a flowchart of a method for analyzing the impact of complex loose bodies based on tunneling simulation.

[0029] Figure 2 This is a computer equipment diagram for a method of analyzing the impact of complex loose bodies based on tunneling simulation. Detailed Implementation

[0030] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0031] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0032] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0033] Reference Figures 1-2 This is one embodiment of the present invention, which provides a method for analyzing the impact of complex loose bodies based on tunneling simulation, including the following steps:

[0034] S1: Obtain geological spatial data and construction status data of complex loose bodies within the target analysis area, construct a basic micro voxel dataset in the simulated three-dimensional space, and aggregate macro voxel units based on spatial adjacency and similarity of geological parameters.

[0035] When performing computer simulation on the target analysis area in front of the tunnel boring machine, all data processing and physical logic operations in this embodiment are automatically executed by a computer device equipped with a processor and memory, which serves as the simulation computing platform.

[0036] First, the computer equipment acquires geological spatial data and construction status data of the complex loose body construction area through an input interface, and stores these data in its memory. The geological spatial data includes porosity, bulk mass, uniaxial compressive strength, internal friction angle, damping parameters, groundwater disturbance parameters, and spatial density distribution input into the simulation model. The construction status data includes shield tunneling direction, tunneling speed, cutterhead thrust, cutterhead torque, cutterhead rotation speed, face pressure, and grouting pressure input into the simulation model. These data are derived from borehole exploration, geophysical surveys, on-site monitoring sensors, shield machine operation records, and the construction control system, respectively.

[0037] After data acquisition, the processor unifies the data from different sources into the three-dimensional coordinate system of the simulation model, ensuring that each set of geological spatial data corresponds to a specific location of the complex loose body in the simulated three-dimensional space. After coordinate unification, the processor divides the simulated three-dimensional space corresponding to the complex loose body into continuously arranged basic micro-voxel units, and determines the model spatial distance between the centers of adjacent basic micro-voxel units as the standard spatial scale of the basic micro-voxel units, which is used for subsequent geological parameter similarity calculations.

[0038] The basic micro-voxel unit serves as the smallest spatial indexing unit and parameter storage unit. Each basic micro-voxel unit records the model's spatial coordinates, spatial boundary, voxel unit number, porosity, bulk mass, uniaxial compressive strength, internal friction angle, damping parameters, groundwater disturbance parameters, and spatial density values ​​mapped from the spatial density distribution in the memory. The voxel unit number is a unique index of the basic micro-voxel unit in the basic micro-voxel dataset, used to locate the corresponding basic micro-voxel unit during subsequent fracturing index set, local fine calculation, and damage result write-back. At this time, the basic micro-voxel unit is only used to store geological parameters and spatial indexes, and does not perform simulated impact energy propagation calculations, local collision search calculations, or damage accumulation calculations. This forms the basic micro-voxel dataset, which serves as the underlying data structure for subsequent macro-voxel aggregation and pre-exploration fracturing calls.

[0039] After the basic micro-voxel dataset is formed, the processor sets an initial computational state for each basic micro-voxel unit. The initial computational state is uniformly set to an inactive state, indicating that the basic micro-voxel unit will not participate in the micro-level fine simulation before macro-voxel aggregation is completed and before being invoked by subsequent pre-exploration fragmentation. Through this process, although the basic micro-voxel units have been established and have their complete geological parameters saved, it does not incur global micro-level computational overhead, thus avoiding the problem of redundant calculations involving first performing global fine calculations and then aggregation.

[0040] Subsequently, the processor calculates the similarity of geological parameters for adjacent basic voxel units in the model space. To avoid direct comparison of geological parameters with different dimensions, porosity, bulk mass, uniaxial compressive strength, internal friction angle, damping parameters, groundwater disturbance parameters, and spatial density are normalized before calculation, and the normalized geological parameters are used to form the geological parameter vector of the basic voxel unit. The similarity of geological parameters between the i-th and j-th basic voxel units is determined according to the following formula:

[0041]

[0042] in, This represents the similarity of geological parameters between the i-th basic voxel unit and the j-th basic voxel unit; This represents the number of a basic micro-voxel unit involved in the similarity calculation. Indicates the number of another fundamental voxel unit that is spatially adjacent to the i-th fundamental voxel unit; Represents the natural exponential function; Indicates the influence coefficient of spatial distance; This represents the spatial distance between the center positions of the i-th and j-th basic voxel units. The standard spatial scale representing the basic microscopic voxel unit; This represents the distance stability constant used to prevent the denominator from being zero; Indicates the influence coefficient of geological parameter differences; Represents the normalized geological parameter vector of the i-th basic micro voxel unit; Represents the normalized geological parameter vector of the j-th basic micro-voxel unit; Let L2 be the norm of the difference in normalized geological parameter vectors between the i-th and j-th basic micro voxel units.

[0043] After completing the geological parameter similarity calculation, the disturbance sensitivity characteristics of the basic micro-voxel units are further generated based on the porosity, uniaxial compressive strength, groundwater disturbance parameters, and spatial density distribution in the basic micro-voxel units. These disturbance sensitivity characteristics characterize the susceptibility of the basic micro-voxel units to compaction, particle rearrangement, or local instability under the disturbances of shield tunneling construction.

[0044] Specifically, the processor first normalizes the porosity, uniaxial compressive strength, groundwater disturbance parameters, and spatial density variation, and determines the strength weakening feature based on the normalized uniaxial compressive strength. Then, it uses the normalized porosity, strength weakening feature, normalized groundwater disturbance parameters, and normalized spatial density variation as components of the perturbation sensitivity feature. Higher porosity, stronger strength weakening feature, higher groundwater disturbance parameters, and more pronounced spatial density variation correspond to higher perturbation sensitivity features for the corresponding basic micro-voxel unit; conversely, lower porosity, lower strength weakening feature, lower groundwater disturbance parameters, and more gradual spatial density variation correspond to lower perturbation sensitivity features for the corresponding basic micro-voxel unit. The processor then writes the perturbation sensitivity feature corresponding to each basic micro-voxel unit into the basic micro-voxel dataset.

[0045] The processor further determines the strength weakening characteristics of the basic micro voxel unit based on the degree of reduction of the uniaxial compressive strength relative to the strength statistical range within the target analysis area; and after the macro voxel unit is formed, it determines the structural non-uniformity of the macro voxel unit based on the differences in geological parameters and disturbance sensitivity characteristics between the basic micro voxel units within the same macro voxel unit.

[0046] Before performing regional growth aggregation, the processor further compares the disturbance sensitivity characteristics between adjacent basic micro voxel units in the model space. If two basic micro voxel units are all within a preset range of difference in terms of porosity variation, strength weakening degree, groundwater disturbance intensity, and spatial density variation, then the two basic micro voxel units are determined to have similar disturbance sensitivity characteristics; if any difference exceeds the preset range, the two basic micro voxel units are not directly aggregated into the same macro voxel unit. Through this processing, the macro voxel units not only have similar geological parameters, but also have a consistent tendency to respond to construction disturbances.

[0047] When performing aggregation judgment, the processor also uses the minimum uniaxial compressive strength, the maximum porosity, the maximum groundwater disturbance parameter, and the degree of spatial density change as hard constraint parameters. When any hard constraint parameter exceeds the preset change condition, even if the geological parameter similarity reaches the preset aggregation threshold, the corresponding basic micro voxel unit will not be merged into the same macro voxel unit.

[0048] Subsequently, macroscopic voxel units are formed according to a deterministic regional growth method. Specifically, the unassigned basic microscopic voxel units are first traversed in order of distance from the simulated shield tunnel face from farthest to nearest, and the currently unassigned basic microscopic voxel unit is used as the aggregation seed; then, basic microscopic voxel units that have a model spatial surface adjacency relationship with the aggregation seed are searched; when the geological parameter similarity between adjacent basic microscopic voxel units and the aggregation seed reaches a preset aggregation threshold, the disturbance-sensitive features are similar, and the spatial range after merging does not exceed the upper limit of the macroscopic voxel unit's scale, the adjacent basic microscopic voxel units are merged into the same macroscopic voxel unit.

[0049] The process of neighbor search and similarity judgment continues until no adjacent basic micro voxel units satisfying the aggregation conditions are found, thus completing the generation of a macro voxel unit. New unassigned basic micro voxel units are then selected as aggregation seeds, and the above process is repeated until all basic micro voxel units within the target analysis area have been assigned. If a basic micro voxel unit cannot be merged into an adjacent macro voxel unit due to abrupt changes in geological parameters or significant differences in disturbance-sensitive characteristics, the processor will register the corresponding basic micro voxel unit separately as a boundary-type macro voxel unit and mark it as a candidate for local parameter mutation in the macro voxel data table. Boundary-type macro voxel units still participate in subsequent boundary mapping and pre-exploration fragmentation judgment as the default simulation calculation objects.

[0050] The reason for adopting the aforementioned regional growth aggregation method is that macroscopic voxel units need to maintain continuity in the model space, while ensuring consistency in the geological parameters and construction disturbance response tendencies of the internal basic microscopic voxel units. Compared to simply merging fixed large grids, this aggregation method avoids forcibly merging regions with abrupt changes in porosity, intensity, and groundwater disturbance into the same macroscopic voxel unit, thereby reducing the possibility of subsequent rapid macroscopic propagation calculations masking local instability risks. Compared to ordinary clustering methods, this aggregation method also avoids spatially discontinuous basic microscopic voxels being grouped into the same macroscopic voxel unit, ensuring that subsequent preliminary fracturing can directly restore the continuous local model space region.

[0051] After the macro voxel units are generated, a split index relationship is established between the macro voxel units and their corresponding underlying micro voxel units, and stored in the form of a split index set:

[0052]

[0053] in, This represents the set of basic micro voxel indices contained in the u-th macro voxel unit; Indicates the macroscopic voxel unit number; Indicates the basic micro-voxel unit number; Indicates the macroscopic voxel unit number to which the i-th basic micro voxel unit belongs; This indicates that the i-th basic micro voxel unit is assigned to the u-th macro voxel unit.

[0054] After the fracturing index set is established, the macro voxel unit number, macro voxel spatial boundary, macro voxel center coordinates, included basic micro voxel index set, corresponding geological parameter statistics, disturbance sensitivity feature statistics, normalized volumetric mass, intensity weakening feature, equivalent propagation hindrance feature determined by geological parameter statistics, structural inhomogeneity determined by disturbance sensitivity feature statistics, and the number of internal basic micro voxels are written into the macro voxel data table in memory. The macro voxel units in the macro voxel data table serve as the default simulation calculation objects for subsequent calculations; the basic micro voxel dataset serves as the local fine-grained calculation object called upon after the preliminary fracturing conditions are met.

[0055] Conventional simulations of complex loose bodies often employ either precise calculations using a fine-mesh grid across the entire domain or coarse calculations using a fixed large-mesh grid. The former can easily lead to excessive computational burden, while the latter can easily mask abrupt changes in porosity, intensity, and groundwater disturbance. Existing methods struggle to simultaneously balance simulation efficiency and the accuracy of local instability identification.

[0056] The above processing constructs a basic micro-voxel dataset through simulation of three-dimensional space. This allows the basic micro-voxel units to primarily serve as spatial indexers and parameter storage units, without directly participating in the simulation of impact energy propagation, local collision search, and damage accumulation calculation. Macro-voxel units, on the other hand, are formed by constraints imposed by the model space's adjacency relationships, geological parameter similarity, and similarity of disturbance-sensitive features, avoiding simple merging based on fixed mesh sizes.

[0057] This approach makes macroscopic voxel elements the default simulation calculation objects, while basic microscopic voxel elements are only invoked for local fine-grained calculations after the pre-exploration fragmentation conditions are met. Compared to conventional global fine-grained calculations, the computational load is significantly reduced; compared to ordinary coarse-mesh simulations, locally weakened regions are still preserved through fragmentation index relationships, providing a clear basis for subsequent fine-grained calculations.

[0058] S2: Based on the simulated shield tunneling distance, historical simulation boundary states, and spatial attenuation relationships, perform spatiotemporal sliding window boundary mapping to generate simulation dynamic boundary parameters at the current simulation time step; generate simulated impact load gradient field based on simulation dynamic boundary parameters, and determine the precise candidate region by combining the loose body disturbance sensitivity characteristics.

[0059] After constructing the macroscopic voxel elements, the macroscopic voxel element numbers, model spatial locations, geological parameter statistics, disturbance sensitivity feature statistics, structural inhomogeneity, and fracture index relationships are read from the macroscopic voxel data table. The macroscopic voxel elements are then used as the boundary mapping and gradient calculation objects for the current simulation time step. The basic microscopic voxel elements remain as inactive local fine-grained calculation objects in this step and do not participate in boundary mapping and load gradient calculation.

[0060] In the initial calculation phase of the simulation model, the processor converts the cutterhead thrust, cutterhead torque, cutterhead rotation speed, face pressure, grouting pressure, and propulsion speed into simulation dynamic boundary parameters applied to the macroscopic voxel elements surrounding the simulated shield face, based on the construction status data and the position of the simulated shield face in the simulation 3D space. These simulation dynamic boundary parameters represent the construction boundary action states experienced by the macroscopic voxel elements within the corresponding simulation time step, indicating the thrust, rotational shear, face pressure, grouting pressure, and propulsion actions transmitted by the shield construction equipment to the complex loose body. The initial dynamic boundary parameters include thrust boundary components, torque rotation boundary components, face pressure boundary components, grouting pressure boundary components, and propulsion speed boundary components. The torque rotation boundary component is jointly determined by the cutterhead torque and cutterhead rotation speed and serves as the historical simulation boundary state for subsequent simulation iterations.

[0061] During the simulation iteration phase, the processor obtains the simulated shield tunneling distance between the current simulation time step and the previous simulation time step based on the advancement speed and the set simulation time step. Subsequently, based on the simulated shield tunneling distance and direction, the processor compensates for the advancement position of historical macroscopic voxel cells with historical simulation boundary states from the previous simulation time step, and establishes a historical mapping neighborhood around the macroscopic voxel cells in front of the simulated shield tunneling. When the simulated shield centerline trajectory is a curve, the processor performs arc length compensation along the simulated shield centerline trajectory and determines the local shield tunneling direction based on the trajectory position of the current simulation time step, avoiding mapping historical simulation boundary states to incorrect model space positions along a single straight line.

[0062] After establishing the historical mapping neighborhood, the processor determines the mapping weights for transferring boundary parameters from the historical macro voxel units to the macro voxel units ahead of the tunnel, based on the model space distance, spatial attenuation relationship, and equivalent propagation hindrance characteristics between the historical macro voxel units and the macro voxel units ahead of the tunnel.

[0063]

[0064] in, For the first The historical macroscopic voxel unit towards the first The mapping weights of the boundary parameters are passed to the macroscopic voxel units at the front of the tunnel. Numbering of macroscopic voxel units at the tunneling front. Numbering of historical macroscopic voxel units. For the current simulation time step, This is the spatial distance attenuation coefficient of the model. For the first After propulsion position compensation, the historical macroscopic voxel unit is related to the first Model space distance between macroscopic voxel units at the front of the tunnel. The radius of the spatiotemporal sliding window mapping. Let be the spatial distance stability constant. The attenuation coefficient is the equivalent propagation retardation difference. For the first Equivalent propagation hindrance characteristics of a macroscopic voxel unit at the tunneling front. For the first The equivalent propagation hindrance characteristics of a historical macroscopic voxel unit at the previous simulation time step. It is the hysteresis characteristic stability constant.

[0065] The equivalent propagation hindrance characteristics are determined by the normalized volumetric mass, normalized damping parameter, normalized internal friction angle, normalized porosity, normalized groundwater disturbance parameter, and strength weakening characteristics of the macroscopic voxel unit. Specifically, larger volumetric mass, damping parameter, and internal friction angle correspond to a higher degree of stability propagation hindrance of the construction boundary action within the macroscopic voxel unit. Larger porosity, groundwater disturbance parameter, and strength weakening characteristics correspond to a higher degree of instability sensitivity of the macroscopic voxel unit, thus weakening and correcting the stability propagation of the construction boundary action. These parameters are obtained from the statistical results of geological parameters in the macroscopic voxel data table. Correction directions and calibration parameters are determined by indoor experiments, historical simulation samples, or field monitoring inversion data and then written into the simulation parameter table.

[0066] The model spatial distance attenuation coefficient and the equivalent propagation hindrance difference attenuation coefficient are calibrated based on the attenuation results of the construction boundary effect intensity with respect to model spatial distance and medium differences in historical simulation samples. The spatiotemporal sliding window mapping radius is determined based on the simulated shield tunneling distance, the upper limit of the macroscopic voxel unit scale, and the forward influence range allowed for mapping in the current simulation time step, and is not less than the simulated shield tunneling distance corresponding to the current simulation time step. The spatial distance stability constant and the hindrance characteristic stability constant are positive numbers to prevent the denominator from being zero, and are written into the simulation parameter table.

[0067] After obtaining the mapping weights, the historical simulation boundary states of each historical macroscopic voxel unit in the historical mapping neighborhood at the previous simulation time step are weighted and fused according to the mapping weights. Combined with the boundary transfer correction relationship determined by geological parameters and the newly added simulation construction boundary action correction amount, the simulation dynamic boundary parameters of the macroscopic voxel unit in front of the tunneling are generated at the current simulation time step.

[0068]

[0069] in, For the first The simulation dynamic boundary parameters of a macroscopic voxel element at the front of the tunneling tunnel at the current simulation time step. For the first The historical mapping neighborhood corresponding to each macroscopic voxel unit at the front of the tunnel. For the first The mapping weights corresponding to each historical macroscopic voxel unit. For the first Boundary transfer correction matrix determined by the geological parameters of macroscopic voxel units. For the first The historical simulation boundary state of a historical macroscopic voxel element at the previous simulation time step. For boundary fusion stability constant, For the first The correction amount for the simulated construction boundary effect added to each macroscopic voxel element at the current simulation time step.

[0070] The boundary transfer correction matrix is ​​determined by the equivalent propagation hindrance characteristics, porosity, damping parameters, uniaxial compressive strength, groundwater disturbance parameters, and structural inhomogeneity of the macroscopic voxel element at the current simulation time step. It is used to correct the transfer amplitude of the thrust boundary component, torque rotation boundary component, face pressure boundary component, grouting pressure boundary component, and propulsion velocity boundary component in the corresponding macroscopic voxel element. The boundary transfer correction matrix is ​​formed from historical simulation samples, field monitoring inversion data, or indoor test calibration data, and is written into the memory as a fixed parameter table.

[0071] The newly added simulated construction boundary action correction is determined by the changes in construction status data within the current simulation time step relative to the previous simulation time step. It is used to supplement the boundary actions newly applied to the macroscopic voxel elements in front of the tunnel during the current advancement process. Specifically, it includes corrections for axial compression caused by changes in cutterhead thrust, rotational shear caused by changes in cutterhead torque and speed, support pressure caused by changes in face pressure, diffusion compression caused by changes in grouting pressure, and propulsion caused by changes in advance speed.

[0072] After generating the simulation dynamic boundary parameters, the processor extracts the boundary action components from the simulation dynamic boundary parameters corresponding to each macroscopic voxel unit, and uses these boundary action components as inputs for calculating the simulated impact load. These boundary action components are not the original construction state data that has been reread, but rather the construction boundary action states that have been allocated to the corresponding macroscopic voxel units after being mapped via a spatiotemporal sliding window. The boundary action components include thrust boundary components, torque rotation boundary components, face pressure boundary components, grouting pressure boundary components, and propulsion speed boundary components. Specifically, the thrust boundary component represents the axial compression effect of the cutterhead thrust on the complex loose material in front of the face; the torque rotation boundary component is jointly determined by the mapped cutterhead torque boundary state and cutterhead rotation speed boundary state, representing the shearing effect of the cutterhead rotation on the complex loose material; the face pressure boundary component represents the support pressure effect of the face on the complex loose material in front; the grouting pressure boundary component represents the diffusion compression effect of the grouting pressure on the surrounding complex loose material; and the propulsion speed boundary component represents the propulsion effect formed by the forward movement of the simulated shield face within a unit simulation time. The processor performs dimensional normalization on the boundary action components and then weights and fuses them to obtain the load scalar of each macroscopic voxel element. Next, based on the adjacency relationship in the model space, it calculates the load scalar difference between adjacent macroscopic voxel elements, generating the load gradient vector corresponding to each macroscopic voxel element. The magnitude of the load gradient vector is then calculated to obtain the load gradient intensity. Thus, the load gradient vector and the load gradient intensity together constitute the simulated impact load gradient field.

[0073] For example, within a certain simulation time step, a macroscopic voxel element in front of the simulated tunnel boring machine (TBM) face receives the thrust boundary component and the face pressure boundary component, mapped from historical simulation boundary states. Simultaneously, due to the increase in cutterhead rotation speed within the current simulation time step, the torque rotation boundary component is enhanced along with the mapped cutterhead torque and rotation speed boundary states. The processor uses these boundary components as the construction boundary state of the macroscopic voxel element in the current simulation time step and performs dimensional normalization on these boundary components to obtain the load scalar of the macroscopic voxel element. When the load scalar difference between this macroscopic voxel element and its adjacent macroscopic voxel elements is significant, the processor generates a corresponding load gradient vector and load gradient intensity to represent the direction and degree of change of the simulated impact load in the simulation three-dimensional space.

[0074] After obtaining the simulated impact load gradient field, the processor generates a spatial gradient characterization vector based on the load gradient intensity, the projection component of the load gradient vector in the shield tunneling direction, and the model spatial distance to the simulated shield face. It then generates a loose body property characterization vector based on porosity, normalized bulk mass, strength weakening characteristics, and structural inhomogeneity. Subsequently, the processor performs calibration mapping on the spatial gradient characterization vector and the loose body property characterization vector, obtaining a spatial gradient mapping vector and a property mapping vector. Based on the correlation between these two vectors, normalization is performed among the macroscopic voxel units participating in the selection process to obtain normalized correlation weights.

[0075]

[0076] in, For the first Normalized correlation weights of macroscopic voxel units at the current simulation time step This is the calibration mapping matrix corresponding to the spatial gradient representation vector. For the first Spatial gradient representation vector of macroscopic voxel units This is the calibration mapping matrix corresponding to the loose body attribute characterization vector. For the first Loose volume attribute representation vector of macroscopic voxel units For vector dot product operation, To associate the dimensions of the computational space, This refers to the set of macroscopic voxel units that participate in the screening at the current simulation time step. Number the other macroscopic voxel unit involved in the normalization process. For the first Spatial gradient representation vector of macroscopic voxel units For the first Loose volume attribute representation vector of macroscopic voxel units It is a natural exponential function.

[0077] in, and The calibration mapping matrix, determined from historical simulation samples, field monitoring inversion data, or indoor test calibration data, is used to transform the spatial gradient representation vector and the loose body attribute representation vector to the same associated computational space. The dimension of the associated computational space is determined based on the unified vector length of the spatial gradient representation vector and the loose body attribute representation vector after calibration mapping, and is written into the simulation parameter table before simulation calculation. The set of macroscopic voxel units participating in the screening at the current simulation time step consists of macroscopic voxel units located within a preset influence range in front of the simulated shield tunnel face and having completed the simulation dynamic boundary parameter mapping. The preset screening threshold is determined by the normalized association weight distribution of the macroscopic voxel units that actually enter the local fine calculation region in historical simulation samples or field monitoring inversion data, or by the upper limit of the number of macroscopic voxel units allowed to enter the pre-exploration rift judgment at the current simulation time step.

[0078] The calibration mapping matrix is ​​a fixed mapping parameter used in the simulation calculation process to realize the correlation calculation between representation vectors of different dimensions, without involving neural network training, backpropagation, or online parameter updates. The normalized correlation weight is used to characterize the relative priority of macroscopic voxel units entering the actuarial candidate region in the current simulation time step.

[0079] Finally, the processor identifies macroscopic voxel units whose normalized correlation weights reach a preset screening threshold as actuarial candidate regions for which preliminary fragmentation judgment needs to be performed.

[0080] Existing tunnel boring machine (TBM) construction simulations tend to repeatedly reconstruct boundary conditions during the tunneling process, or simply translate the previous time state, making it difficult to accurately reflect the attenuation of the construction boundary effect in different loose media due to spatial distance and propagation resistance. Current technologies lack a stable boundary state transition mechanism between simulation time steps, affecting both computational continuity and efficiency.

[0081] Based on the simulated tunnel boring machine (TBM) advance distance, historical simulation boundary states, and spatial attenuation relationships, a spatiotemporal sliding window boundary mapping is performed. The processor first compensates for the position of historical macroscopic voxel elements based on the advance distance and direction. Then, it calculates the mapping weights by combining the model's spatial distance and equivalent propagation hindrance characteristics, and performs weighted fusion of historical simulation boundary states. Simultaneously, it overlays the correction amount for newly added simulated construction boundary effects to form the simulation dynamic boundary parameters at the current simulation time step. Based on this, the processor generates a simulated impact load gradient field according to the simulation dynamic boundary parameters, and determines the actuarial candidate region through calibration mapping and normalization correlation processing between the spatial gradient characterization vector and the loose body attribute characterization vector. This allows subsequent pre-exploration fracturing judgments to be focused on macroscopic voxel elements with significant load changes, close to the simulated TBM face, high sensitivity to loose body disturbance, and significant structural differences, avoiding unified actuarial judgments for all macroscopic voxels.

[0082] S3: Perform pre-exploration and fusion judgment on macroscopic voxel units within the actuarial candidate region. When the fusion condition is met, call the corresponding basic microscopic voxel unit to perform local collision search and damage calculation. When the fusion condition is not met, maintain the macroscopic voxel unit to perform macroscopic fast propagation calculation.

[0083] The processor receives the actuarial candidate region output by S2 and reads the simulation dynamic boundary parameters, normalized correlation weights, equivalent propagation hindrance characteristics, structural inhomogeneity, geological parameter statistics, disturbance sensitivity characteristic statistics, and fracture index set for each macroscopic voxel element within the actuarial candidate region. All of the above data comes from the aforementioned macroscopic voxel data table and the calculation results of the current simulation time step.

[0084] For the u-th macroscopic voxel element within the actuarial candidate region, the processor determines the intensity of the simulated impact disturbance entering the macroscopic voxel element based on the simulation dynamic boundary parameters, and performs pre-probe attenuation calculation by combining the equivalent propagation hindrance characteristics and propagation path length to obtain the pre-probe energy deposition result of the macroscopic voxel element:

[0085]

[0086] in, For the first The preliminary energy deposition results of a macroscopic voxel unit at the current simulation time step. Numbering of macroscopic voxel units, For the current simulation time step, To determine the entry point based on the simulation dynamic boundary parameters Simulated impact disturbance intensity of macroscopic voxel units It is a natural exponential function. For the first The equivalent propagation hindrance characteristics of a macroscopic voxel unit at the current simulation time step. To simulate the impact disturbance in the first Preview propagation length in a macroscopic voxel unit For the first Structural nonuniformity of macroscopic voxel units.

[0087] The intensity of the simulated impact disturbance entering the macroscopic voxel element is jointly determined by the load scalar, load gradient intensity, and boundary action components of the corresponding macroscopic voxel element. The load scalar characterizes the intensity of the construction boundary action currently borne by the macroscopic voxel element, while the load gradient intensity characterizes the degree of variation of the construction boundary action between adjacent macroscopic voxel elements. The preliminary propagation length is determined based on the intersection of the incident direction of the simulated impact disturbance entering the macroscopic voxel element and the spatial boundary of the macroscopic voxel, taking the length of the simulated impact disturbance path traversing within the macroscopic voxel element.

[0088] After obtaining the preliminary energy deposition results, the processor identifies local parameter abrupt change regions within macro voxel units based on structural inhomogeneity and differences in geological parameters between basic micro voxel units. The equivalent failure condition of the macro voxel unit is used to determine whether the macro voxel unit needs to be fragmented into basic micro voxel units for local fine-grained calculations. The equivalent failure threshold is based on the minimum failure energy threshold of the corresponding basic micro voxel unit within the macro voxel unit, and is reduced and corrected according to the maximum porosity, maximum groundwater disturbance parameter, maximum disturbance sensitivity characteristic, and structural inhomogeneity. When the preliminary energy deposition results reach or exceed the reduced and corrected equivalent failure threshold, the preliminary energy deposition results are deemed to have met the equivalent failure condition of the macro voxel unit. The equivalent failure threshold is determined from historical simulation samples, field monitoring inversion data, or indoor experimental calibration data and is written into the simulation parameter table.

[0089] The preset actuarial priority condition is used to determine whether macroscopic voxel units that meet the energy deposition risk need to immediately call the corresponding basic microscopic voxel units. The preset actuarial priority threshold is not lower than the preset screening threshold corresponding to the actuarial candidate region, and is determined based on the upper limit of computing resources allowed to call basic microscopic voxel units in the current simulation time step, the distribution of local unstable regions in historical simulation samples, and the in-situ monitoring inversion data; when the normalized correlation weight reaches or exceeds the preset actuarial priority threshold, it is determined that the preset actuarial priority condition has been met.

[0090] Specifically, the processor reads the corresponding basic micro voxel units according to the rift index set, and compares the porosity, uniaxial compressive strength, groundwater disturbance parameters, spatial density, and disturbance sensitivity characteristics of the basic micro voxel units with the corresponding statistical results of the macro voxel units to which they belong. When the continuously distributed basic micro voxel units exhibit weakening strength, abrupt changes in porosity, enhanced groundwater disturbance, or a sudden increase in disturbance sensitivity characteristics, the corresponding continuous region is identified as a local parameter abrupt change region, and the pre-exploration energy deposition results corresponding to this region are used as the basis for rift priority judgment.

[0091] Taking a macroscopic voxel unit near a weak interlayer in front of the tunnel face of a simulated shield tunnel as an example, the processor reads the simulated dynamic boundary parameters of the corresponding macroscopic voxel unit and determines that the simulated impact disturbance has entered the macroscopic voxel unit. Since the macroscopic voxel unit contains continuous basic microscopic voxel units with low uniaxial compressive strength, high porosity, large groundwater disturbance parameters, and a sudden increase in disturbance sensitivity characteristics, the processor identifies this continuous region as a local parameter mutation region. Subsequently, the processor performs a preliminary attenuation calculation on the simulated impact disturbance intensity based on the equivalent propagation hindrance characteristics, obtaining the preliminary energy deposition result of the macroscopic voxel unit. If the preliminary energy deposition result reaches the equivalent failure threshold jointly determined by the minimum uniaxial compressive strength, maximum porosity, maximum groundwater disturbance parameter, maximum disturbance sensitivity characteristic, and structural non-uniformity, and the normalized correlation weight of the macroscopic voxel unit reaches the preset actuation priority threshold, then the processor determines that the macroscopic voxel unit meets the rupture condition and calls the basic microscopic voxel units contained in the macroscopic voxel unit according to the rupture index set to perform local fine calculation.

[0092] When the preliminary energy deposition results reach the equivalent failure condition of the macroscopic voxel unit, and the normalized correlation weight reaches the preset actuarial priority condition, the processor determines that the macroscopic voxel unit meets the cleavage condition, and calls the corresponding basic microscopic voxel unit according to the cleavage index set. The called basic microscopic voxel unit switches from an inactive calculation state to a local actuarial calculation state, and at the same time reads the model space coordinates, geological parameters, and disturbance-sensitive features of the corresponding basic microscopic voxel unit.

[0093] When the preliminary energy deposition results do not meet the equivalent failure condition of the macro voxel element, or the normalized correlation weight does not meet the preset actuarial priority condition, the processor determines that the macro voxel element does not meet the pyrolysis condition, does not call the corresponding basic micro voxel element to participate in the local actuarial calculation, and maintains the macro voxel element to perform macro fast propagation calculation. For macro voxel elements that do not meet the pyrolysis condition, the processor reads the macro fast propagation result written to the macro voxel data table in the previous simulation time step, and uses the value representing the propagation degree of the construction boundary effect as the macro propagation intensity of the previous simulation time step; then, combined with the current simulation dynamic boundary parameters and equivalent propagation hindrance characteristics, it recursively obtains the macro fast propagation result of the current simulation time step, and writes the macro fast propagation result of the current simulation time step into the macro voxel data table. The macro fast propagation result includes the macro propagation intensity and the propagation influence range; the macro propagation intensity is used to characterize the degree to which the unpyroly macro voxel element is affected by the construction boundary effect in the current simulation time step, and the propagation influence range is used to characterize the model space range where the macro propagation intensity reaches the preset background propagation condition.

[0094] For macroscopic voxel units that meet the pyrolysis conditions, the processor allocates local incident perturbation amounts to the invoked basic microscopic voxel units based on the pre-exploration energy deposition results, the model spatial location of the basic microscopic voxel units, and perturbation-sensitive characteristics:

[0095]

[0096] in, Indicates the first The local incident perturbation amount allocated to each basic micro voxel element at the current simulation time step; Indicates the basic micro-voxel unit number; Indicates the first Perturbation-sensitive characteristics of each basic micro voxel unit; This represents the stability constant of the disturbance-sensitive characteristic; Indicates the simulated impact disturbance path and the first The overlap length of the model space of each basic micro voxel unit; Indicates the overlap length stability constant; Indicates the number of another fundamental micro voxel unit in the same set of cleavage indexes; Indicates the first Each macroscopic voxel unit corresponds to a set of basic microscopic voxel indices. Indicates the first Perturbation-sensitive characteristics of each basic micro voxel unit; Indicates the simulated impact disturbance path and the first The overlap length of the model space of each basic micro voxel unit; This represents the stability constant for the local incident perturbation allocation. After the initial allocation of local incident perturbations is completed, the processor normalizes and corrects the local incident perturbations of all basic micro voxel units within the same rift index set, so that the sum of the local incident perturbations of each basic micro voxel unit is consistent with the prospective energy deposition results of the macro voxel units to which they belong.

[0097] After normalization correction, the processor determines the local collision search range based on the model space adjacency relationship between basic voxel units, local incident perturbation, and perturbation-sensitive features. Specifically, the processor expands the search towards adjacent basic voxel units in the model space, centered on basic voxel units with high local incident perturbation, high perturbation-sensitive features, or located within a region of abrupt change in local parameters. When the local incident perturbation of an adjacent basic voxel unit is lower than a preset expansion condition and the perturbation-sensitive feature does not meet the expansion condition, the expansion in that direction stops. This forms the local collision search range.

[0098] Within the local collision search range, the processor calculates the local compression state, disturbance absorption state, and damage accumulation state of the basic voxel elements. The local compression state is jointly determined by the difference in local incident disturbance between adjacent basic voxel elements, the model space adjacency relationship, and the degree of spatial constraint, characterizing the degree of compression transmission from adjacent elements to the basic voxel element; the degree of spatial constraint is determined by the number of basic voxel elements adjacent to the basic voxel element in the model space and their corresponding adjacency directions. The disturbance absorption state is jointly determined by the local incident disturbance, damping parameters, porosity, groundwater disturbance parameters, and disturbance sensitivity characteristics, characterizing the degree to which the basic voxel element absorbs simulated impact disturbances in the current simulation time step. The damage accumulation state is updated based on the disturbance absorption state, local compression state, and the damage accumulation state recorded in the previous simulation time step. The corresponding damage accumulation state is determined according to the following formula:

[0099]

[0100] in, For the first The damage accumulation state of each basic micro voxel unit in the next simulation time step For the first The damage accumulation state of each basic micro voxel unit recorded at the current simulation time step. To obtain the smaller value of the function, For the first The perturbation absorption state of each basic micro voxel unit at the current simulation time step. For the first The failure energy threshold of a basic micro voxel unit The failure energy stability constant, For the first The local compression state of a basic micro voxel unit at the current simulation time step.

[0101] The processor uses the local compression state, disturbance absorption state, and damage accumulation state as the damage calculation results for the basic micro voxel unit and temporarily stores the damage calculation results in the basic micro voxel dataset. For macro voxel units that do not meet the pyrolysis conditions, the processor does not change the inactive calculation state of the corresponding basic micro voxel unit, but only updates the macro fast propagation calculation results through the equivalent propagation hindrance characteristics of the macro voxel unit.

[0102] The local compression state is determined by the difference in local incident disturbance between adjacent basic voxel units, the model space adjacency relationship, and the degree of spatial constraint. A larger difference in local incident disturbance, a more closely spaced adjacency direction than the simulated impact disturbance propagation direction, and a higher degree of spatial constraint correspond to a higher level of local compression. The disturbance absorption state is determined by the local incident disturbance, damping parameters, porosity, groundwater disturbance parameters, and disturbance sensitivity characteristics. A larger local incident disturbance, a larger damping parameter, higher porosity, a larger groundwater disturbance parameter, and a higher disturbance sensitivity characteristic correspond to a higher level of disturbance absorption. The failure energy threshold of the basic voxel unit is determined by the uniaxial compressive strength, volume, porosity, and groundwater disturbance parameters of the basic voxel unit, and calibrated using inversion data from indoor tests or field monitoring.

[0103] Conventional adaptive mesh refinement typically relies on error thresholds or fixed local refinement rules, making it difficult to combine shield tunneling impact disturbances, loose medium resistance, structural non-uniformity, and calculation priority for comprehensive assessment. For complex loose bodies, simply determining refinement based on mesh error or load magnitude can easily lead to missed detections of high-risk areas or over-refinement of low-risk areas.

[0104] The existing scheme only performs preliminary fracturing judgment on macroscopic voxel units within the actuarial candidate area. The processor reads the simulation dynamic boundary parameters, normalized correlation weights, equivalent propagation hindrance characteristics, and structural inhomogeneity. It first determines the preliminary energy deposition results after the simulated impact disturbance enters the macroscopic voxel unit, and then identifies local parameter abrupt change areas by combining the differences in internal geological parameters. This makes the fracturing judgment not only focus on external impact disturbances, but also on whether there are unstable and sensitive locations such as intensity weakening, porosity abrupt change, and enhanced groundwater disturbance within the macroscopic voxel.

[0105] When the preliminary energy deposition results reach the equivalent failure condition and the normalized correlation weight reaches the preset fine calculation priority condition, the processor calls the corresponding basic micro voxel unit according to the split index relationship, switching the basic micro voxel unit from an inactive calculation state to a local fine calculation state. Macro voxel units that do not meet the split condition continue to perform macro fast propagation calculations, thus forming a simulation path of parallel and coordinated macro fast propagation and local fine calculations.

[0106] S4: Write back the damage calculation results to update the corresponding macroscopic voxel elements, and perform freezing processing on the continuous and stable basic microscopic voxel elements according to the cumulative damage state. Combine the macroscopic rapid propagation results to generate the simulation influence area distribution of complex loose bodies under impact load.

[0107] The processor receives damage calculation results from basic micro voxel units, macroscopic rapid propagation results from unfractured macro voxel units, and the calculation status of each basic micro voxel unit at the current simulation time step. The damage calculation results include local compression status, disturbance absorption status, and damage accumulation status. Based on the fracture index set, the processor writes the damage calculation results of the basic micro voxel units back to their respective macro voxel units and simultaneously records the updated geological parameters corresponding to the basic micro voxel units.

[0108] For basic micro voxel units that have already undergone local fine-grained calculations, the processor updates the corresponding geological parameters based on the damage accumulation state. Specifically, when the damage accumulation state of a basic micro voxel unit increases, the processor reduces the uniaxial compressive strength and internal friction angle according to a preset damage parameter degradation relationship, adjusts the porosity, damping parameters, and groundwater disturbance parameters, and writes the updated geological parameters into the basic micro voxel dataset. The preset damage parameter degradation relationship updates the uniaxial compressive strength, internal friction angle, porosity, damping parameters, and groundwater disturbance parameters according to the damage accumulation state; wherein, when the damage accumulation state increases, the uniaxial compressive strength and internal friction angle decrease, while the porosity, damping parameters, and groundwater disturbance parameters increase; when the damage accumulation state reaches the failure determination threshold, the corresponding basic micro voxel unit is marked as a failed basic micro voxel unit.

[0109] After the geological parameters of the basic micro voxel units are updated, the processor recalculates the geological parameter statistics of the corresponding macro voxel units based on the fragmentation index set. For the k-th type of geological parameter of the u-th macro voxel unit, the statistical results after writing back are determined according to the following formula:

[0110]

[0111] in, Indicates the first The macroscopic voxel unit at the next simulation time step Statistical results of geological parameters; Indicates the macroscopic voxel unit number; Indicates the category number of the geological parameter; Indicates the next simulation time step; Indicates the basic micro-voxel unit number; Indicates the first The first basic micro-voxel unit belongs to the first The set of cleavage indexes corresponding to each macroscopic voxel unit; Indicates the first Each macroscopic voxel unit corresponds to a set of basic microscopic voxel indices. Indicates the first The volume of a basic microscopic voxel unit; Indicates the first The basic micro voxel unit at the next simulation time step Updated geological parameters.

[0112] When re-statistically analyzing the geological parameters of the macroscopic voxel units, the processor, in addition to calculating the volume-weighted average, also simultaneously records the minimum uniaxial compressive strength, the maximum porosity, the maximum groundwater disturbance parameter, the maximum disturbance-sensitive characteristic, and the structural inhomogeneity.

[0113] Based on the statistical results of the rewritten geological parameters, the equivalent propagation hindrance characteristics of the macroscopic voxel unit are redefined; based on the rewritten disturbance sensitivity characteristics, the statistical results of the disturbance sensitivity characteristics of the macroscopic voxel unit are redefined; based on the differences in geological parameters and disturbance sensitivity characteristics among the basic microscopic voxel units within the macroscopic voxel unit, the structural inhomogeneity of the macroscopic voxel unit is redefined. The updated equivalent propagation hindrance characteristics, disturbance sensitivity characteristics statistical results, and structural inhomogeneity are written into the macroscopic voxel data table, serving as input for the next simulation time step to perform spatiotemporal sliding window boundary mapping, precise candidate region determination, and pre-exploration fracture judgment.

[0114] Subsequently, a freeze judgment is performed on the basic micro voxel units in the local fine calculation state. When a basic micro voxel unit does not reach the instability evolution condition within a consecutive preset number of simulation time steps, and its corresponding macro voxel unit does not meet the fragmentation condition again within the corresponding simulation time step, the processor switches the corresponding basic micro voxel unit from the local fine calculation state back to the inactive calculation state and stops the local fine calculation of the corresponding basic micro voxel unit. The instability evolution condition includes at least one of the following: the damage accumulation state continues to increase, the disturbance absorption state exceeds the preset absorption condition, or the local compression state exceeds the preset compression condition. After switching back to the inactive calculation state, the corresponding basic micro voxel unit still retains the damage calculation results and updated geological parameters, but subsequently, by default, the corresponding macro voxel unit participates in the macro rapid propagation calculation.

[0115] When the increment of the simulation dynamic boundary parameter of the corresponding macroscopic voxel unit in a subsequent simulation time step exceeds the preset reactivation condition, or when the macroscopic fast propagation result reaches the background propagation condition, the processor will switch the corresponding basic microscopic voxel unit back to the local fine calculation state. The increment of the simulation dynamic boundary parameter is determined by the difference between the simulation dynamic boundary parameter of the current simulation time step and the simulation dynamic boundary parameter recorded before freezing. The background propagation condition is determined by comparing the macroscopic propagation intensity in the macroscopic fast propagation result with a preset background propagation threshold; when the macroscopic propagation intensity reaches or exceeds the preset background propagation threshold, the macroscopic fast propagation result is determined to have reached the background propagation condition.

[0116] For macroscopic voxel elements that do not meet the pyrolysis conditions, the processor retains their macroscopic fast propagation calculation results and writes them into the macroscopic voxel data table. For macroscopic voxel elements that have undergone local fine-grained calculations, the processor uses the damage accumulation state of the basic microscopic voxel elements as the basis for judging local effects; for macroscopic voxel elements that have not undergone local fine-grained calculations, the processor uses the macroscopic fast propagation results as the basis for judging background propagation effects. This ensures that the simulation influence region distribution simultaneously includes both the local fine-grained damage region and the macroscopic propagation influence region.

[0117] Extract the basic micro voxel units whose damage accumulation state has reached the failure determination threshold to form a set of basic failure micro voxels:

[0118]

[0119] in, This represents the set of fundamental microvoxels representing the failure at the next simulation time step. Indicates the next simulation time step; Indicates the basic micro-voxel unit number; Indicates the first The damage accumulation state of a basic micro voxel unit at the next simulation time step; Indicates the failure determination threshold; Indicates greater than or equal to; Indicates the first Each basic micro-voxel unit reaches the failure determination condition. The failure determination threshold is determined based on the value range of the damage accumulation state. When the damage accumulation state adopts a normalized value of zero to one, the failure determination threshold is a positive number not greater than one, and is determined by historical simulation samples, indoor test or field monitoring inversion data; when the damage accumulation state reaches or exceeds the failure determination threshold, the corresponding basic micro-voxel unit is included in the failed basic micro-voxel set.

[0120] After obtaining the set of basic micro-voxels of failure, the processor groups the set of basic micro-voxels into continuous regions according to their model space coordinates and model space adjacency relationships. Specifically, basic micro-voxels of failure that are adjacent to each other by face, edge, or corner are grouped into the same failure connected region; basic micro-voxels of failure that are not adjacent to each other by model space are divided into different failure connected regions. Each failure connected region is used to represent a locally damaged area in a complex loose body that has reached the failure determination criteria.

[0121] After completing the partitioning of the failed connected regions, the processor extracts the basic micro-voxel units of the outer boundary of each failed connected region and generates the corresponding three-dimensional influence profile based on the model space coordinates of the basic micro-voxel units of the outer boundary. The three-dimensional influence profile is used to characterize the core damage range formed by the complex loose body under the results of local fine calculation.

[0122] For macro voxel elements that have undergone macro-fast propagation calculations but have not triggered pre-exploration lysis, the processor reads the corresponding macro-fast propagation results and determines whether the macro-fast propagation results meet the preset background disturbance conditions. If the preset background disturbance conditions are met, the corresponding macro voxel element is designated as the peripheral disturbance region outside the core destruction range; if the preset background disturbance conditions are not met, the corresponding macro voxel element is not included in the simulation influence region distribution.

[0123] Finally, the processor spatially merges the core damage area and the peripheral disturbance region to generate a simulated influence region distribution of a complex loose body under impact load. This simulated influence region distribution includes the core damage region formed by the microscopic voxel units of the failed foundation, and the peripheral disturbance region formed by the macroscopic rapid propagation results, thus simultaneously reflecting both local failure results and macroscopic propagation effects.

[0124] Conventional local refinement simulations often lack a closed-loop mechanism for writing back and updating the upper-level coarse model after completing local calculations. This results in macroscopic layer parameters still using the old state, and subsequent time steps are prone to propagation state distortion. Existing solutions write back the local compression state, disturbance absorption state, and damage accumulation state of the basic microscopic voxel units to their respective macroscopic voxel units and update the corresponding geological parameters simultaneously.

[0125] Based on the write-back results, the processor recalculates the equivalent propagation hindrance characteristics, disturbance sensitivity statistics, and structural nonuniformity of the macroscopic voxel units, enabling the macroscopic voxel data table to reflect the damage evolution results after local fine-grained calculations. For basic microscopic voxel units that have not reached the instability evolution condition within a consecutive preset number of simulation time steps, and whose corresponding macroscopic voxel units have not again met the fragmentation condition, the processor switches the local fine-grained calculation state back to the inactive calculation state, avoiding long-term occupation of computing power by local fine-grained regions.

[0126] Finally, the processor extracts the basic micro-voxel units whose cumulative damage state reaches the failure determination threshold, and forms continuous failure regions according to the adjacency relationship in the model space, further generating the core damage range. For macro-voxel units that are not fractured but whose macroscopic rapid propagation results reach the background disturbance conditions, they are included in the simulation influence region distribution as the peripheral disturbance range. In this way, the output results simultaneously include local fine damage regions and macroscopic propagation disturbance regions, which can more completely reflect the spatial influence range of complex loose bodies under impact loads.

[0127] This embodiment also provides a system for analyzing the impact of complex loose bodies based on tunneling simulation, including:

[0128] The module consists of four parts: a voxel modeling module, a boundary mapping module, and a result output module. The voxel modeling module acquires geological spatial data and construction status data, constructs a basic micro-voxel dataset in the simulated 3D space, and aggregates it to form macro-voxel units. The boundary mapping module generates simulation dynamic boundary parameters for the current simulation time step based on the simulated shield tunneling distance, historical simulation boundary states, and spatial attenuation relationships. The candidate region determination module generates a simulated impact load gradient field based on the simulation dynamic boundary parameters and determines the precise calculation candidate region by combining the loose body disturbance sensitivity characteristics. The pre-exploration lysis module performs pre-exploration lysis judgment on the macro-voxel units within the precise calculation candidate region and calls the corresponding basic micro-voxel units when the lysis conditions are met. The local precise calculation module performs local collision search and damage calculation on the called basic micro-voxel units. The result output module writes back the damage calculation results to update the corresponding macro-voxel units, performs freezing processing on continuous and stable basic micro-voxel units, and generates the simulated influence area distribution of complex loose bodies under impact loads.

[0129] This embodiment also provides a computer device applicable to the method for analyzing the impact of complex loose bodies based on tunneling simulation, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the method for analyzing the impact of complex loose bodies based on tunneling simulation as proposed in the above embodiment.

[0130] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0131] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the method for analyzing the influence of complex loose bodies based on tunneling simulation as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0132] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for analyzing the impact of complex loose bodies based on tunneling simulation, characterized in that, include: Geological spatial data and construction status data of complex loose bodies within the target analysis area are obtained. A basic micro voxel dataset is constructed in the simulated three-dimensional space, and macro voxel units are formed by aggregating them based on spatial adjacency and similarity of geological parameters. Based on the simulated shield tunneling distance, historical simulation boundary states, and spatial attenuation relationships, a spatiotemporal sliding window boundary mapping is performed to generate the simulation dynamic boundary parameters at the current simulation time step. The simulated impact load gradient field is generated based on the simulated dynamic boundary parameters, and the precise candidate region is determined by combining the sensitive characteristics of loose body disturbance. The macroscopic voxel units within the actuarial candidate region are pre-explored for fusion judgment. When the fusion condition is met, the corresponding basic microscopic voxel units are called to perform local collision search and damage calculation. When the fusion condition is not met, the macroscopic voxel units are maintained to perform macroscopic fast propagation calculation. The damage calculation results are written back to update the corresponding macroscopic voxel elements, and the continuous and stable basic microscopic voxel elements are frozen according to the cumulative damage state. The simulation influence region distribution of complex loose bodies under impact load is generated by combining the macroscopic rapid propagation results.

2. The method for analyzing the impact of complex loose bodies based on tunneling simulation as described in claim 1, characterized in that: The geological spatial data includes the porosity, bulk mass, uniaxial compressive strength, internal friction angle, damping parameters, groundwater disturbance parameters, and spatial density distribution input into the simulation model. The construction status data includes the shield tunneling direction, tunneling speed, cutterhead thrust, cutterhead torque, cutterhead rotation speed, face pressure, and grouting pressure input into the simulation model. The formation of macroscopic voxel units includes dividing the simulation 3D space corresponding to the complex loose body into basic microscopic voxel units, and using the basic microscopic voxel units as the smallest spatial index units and parameter storage units to record model spatial coordinates and geological parameters, and setting an initial inactive computation state for each basic microscopic voxel unit; based on the model spatial adjacency relationship, geological parameter similarity, and disturbance sensitivity feature similarity between basic microscopic voxel units, multiple basic microscopic voxel units are aggregated into macroscopic voxel units, and the macroscopic voxel units are used as the default simulation computation objects, while establishing a split index relationship between the macroscopic voxel units and the corresponding basic microscopic voxel units.

3. The method for analyzing the impact of complex loose bodies based on tunneling simulation as described in claim 2, characterized in that: The generation of simulation dynamic boundary parameters at the current simulation time step includes: in the initial calculation stage of the simulation model, assigning initial dynamic boundary parameters to the macroscopic voxel units around the simulation shield face based on the construction status data and the position of the simulation shield face in the simulation three-dimensional space; and in the simulation iteration stage, calculating the simulation shield advance distance between the current simulation time step and the previous simulation time step based on the advance speed and the set simulation time step. Based on the simulated shield tunneling distance and the shield tunneling direction, a corresponding historical mapping neighborhood is determined for the macroscopic voxel unit in front of the simulated shield tunneling in the simulated three-dimensional space; according to the model space distance between the historical macroscopic voxel unit in the historical mapping neighborhood and the macroscopic voxel unit in front of the tunneling, combined with the spatial attenuation relationship and the equivalent propagation hindrance characteristics, the mapping weight of the boundary parameters transmitted from the historical macroscopic voxel unit to the macroscopic voxel unit in front of the tunneling is calculated. Based on the mapping weights, the historical simulation boundary states of each historical macro voxel unit in the historical mapping neighborhood at the previous simulation time step are weighted and fused. Combined with the boundary transfer correction relationship determined by the geological parameters and the newly added simulation construction boundary effect correction amount, the simulation dynamic boundary parameters of the macro voxel unit in front of the tunneling are generated at the current simulation time step.

4. The method for analyzing the impact of complex loose bodies based on tunneling simulation as described in claim 3, characterized in that: The generation of the simulated impact load gradient field includes extracting boundary action components that characterize the simulated impact load based on the simulated dynamic boundary parameters at the current simulation time step. The load scalar of each macroscopic voxel element is calculated by weighting and summing the boundary action components based on dimensional normalized weights. The difference of the load scalar is calculated based on the adjacency relationship in the model space, the load gradient vector corresponding to each macroscopic voxel unit is generated, and the magnitude of the load gradient vector is calculated to generate the corresponding load gradient intensity. The simulated impact load gradient field is formed by the load gradient vector and the load gradient intensity together. The determination of the actuarial candidate region includes: forming a spatial gradient characterization vector based on the load gradient intensity, the projection component of the load gradient vector in the shield advancement direction, and the model spatial distance to the simulated shield face; forming a loose body attribute characterization vector based on the porosity, normalized bulk mass, strength weakening characteristics, and structural inhomogeneity of the macroscopic voxel units; performing calibration mapping on the spatial gradient characterization vector and the loose body attribute characterization vector respectively to obtain the spatial gradient mapping vector and the attribute mapping vector; performing normalization processing on the macroscopic voxel units participating in the screening based on the correlation between the spatial gradient mapping vector and the attribute mapping vector to obtain the normalized correlation weight of each macroscopic voxel unit; and determining the macroscopic voxel units whose normalized correlation weight reaches a preset screening threshold as actuarial candidate regions that need to undergo preliminary fracture judgment.

5. The method for analyzing the impact of complex loose bodies based on tunneling simulation as described in claim 4, characterized in that: The preliminary fracturing judgment of macroscopic voxel units within the actuarial candidate area includes: reading the simulated dynamic boundary parameters, normalized correlation weights, equivalent propagation hindrance characteristics, and structural inhomogeneity of macroscopic voxel units within the actuarial candidate area; determining the simulated impact disturbance intensity entering the macroscopic voxel unit based on the simulated dynamic boundary parameters, and performing preliminary attenuation calculation on the simulated impact disturbance intensity in combination with the equivalent propagation hindrance characteristics to obtain the preliminary energy deposition results of the macroscopic voxel unit; identifying local parameter mutation regions within the macroscopic voxel unit based on the structural inhomogeneity and the differences in geological parameters between basic microscopic voxel units, and using the preliminary energy deposition results corresponding to the local parameter mutation regions as the basis for prioritizing fracturing judgment. When the preliminary energy deposition result reaches the equivalent failure condition of the macro voxel unit and the normalized correlation weight reaches the preset actuarial priority condition, the macro voxel unit is determined to meet the pyrolysis condition, and the corresponding basic micro voxel unit is called according to the pyrolysis index relationship; when the preliminary energy deposition result does not reach the equivalent failure condition of the macro voxel unit, or the normalized correlation weight does not reach the preset actuarial priority condition, the macro voxel unit is determined not to meet the pyrolysis condition, and the macro voxel unit continues to perform macro fast propagation calculation.

6. The method for analyzing the impact of complex loose bodies based on tunneling simulation as described in claim 5, characterized in that: The execution of local collision search and damage calculation includes: reading the model spatial coordinates, geological parameters, and disturbance sensitivity features of the corresponding basic micro voxel unit according to the rift index relationship, switching the corresponding basic micro voxel unit from an inactive calculation state to a locally refined calculation state; allocating local incident disturbance amounts according to the pre-exploration energy deposition results and the model spatial position of the basic micro voxel unit; determining the local collision search range based on the model spatial adjacency relationship between basic micro voxel units, the local incident disturbance amount, and the disturbance sensitivity features, and within the local collision search range, calculating the local compression state, disturbance absorption state, and damage accumulation state of the basic micro voxel unit, and determining the damage calculation results of the basic micro voxel unit based on the local compression state, disturbance absorption state, and damage accumulation state.

7. The method for analyzing the impact of complex loose bodies based on tunneling simulation as described in claim 6, characterized in that: The step of writing back the damage calculation results to the corresponding macroscopic voxel unit includes: writing back the damage calculation results and updated geological parameters of the basic microscopic voxel unit to the macroscopic voxel unit; recalculating the equivalent propagation hindrance characteristics, disturbance sensitivity statistical results, and structural inhomogeneity of the macroscopic voxel unit based on the write-back results; when the basic microscopic voxel unit in the local fine calculation state does not reach the instability evolution condition within a consecutive preset number of simulation time steps, and the macroscopic voxel unit does not meet the rupture condition again, the corresponding basic microscopic voxel unit is switched back to the inactive calculation state, and the local fine calculation of the corresponding basic microscopic voxel unit is stopped; extracting the basic microscopic voxel units whose cumulative damage state has reached the failure judgment threshold, performing three-dimensional spatial connectivity identification on them, and combining the macroscopic rapid propagation results to output the simulation influence area distribution.

8. A system for analyzing the impact of complex loose bodies based on tunneling simulation, based on the method for analyzing the impact of complex loose bodies based on tunneling simulation as described in any one of claims 1 to 7, characterized in that: The voxel modeling module is used to acquire geological spatial data and construction status data, construct a basic micro voxel dataset in the simulation three-dimensional space, and aggregate it to form macro voxel units; The boundary mapping module is used to generate simulation dynamic boundary parameters for the current simulation time step based on the simulated shield tunneling distance, historical simulation boundary states, and spatial attenuation relationships. The candidate region determination module generates a simulated impact load gradient field based on the simulated dynamic boundary parameters and determines the precise candidate region by combining the perturbation sensitivity characteristics of the loose body. The pre-fracture detection module performs pre-fracture detection judgment on the macroscopic voxel elements within the precise candidate region and calls the corresponding basic microscopic voxel elements when the fracture conditions are met. The local precise calculation module performs local collision search and damage calculation on the called basic microscopic voxel elements. The result output module writes back the damage calculation results to update the corresponding macroscopic voxel elements, performs freezing processing on continuous and stable basic microscopic voxel elements, and generates the simulated influence region distribution of complex loose bodies under impact load.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the method for analyzing the influence of complex loose bodies based on tunneling simulation as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the method for analyzing the influence of complex loose bodies based on tunneling simulation as described in any one of claims 1 to 7.

Images