A method and system for modeling soil and composite strata based on multidimensional information fusion

By using multidimensional information fusion technology, the problem of unified expression and deep fusion of multi-source heterogeneous geological data has been solved, achieving high-precision three-dimensional geological modeling and improving the automation level and engineering practicality of geological modeling.

CN123118943APending Publication Date: 2026-11-17SHANDONG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611353190.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-02
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing technologies are unable to effectively unify the representation of multi-source heterogeneous geological data, resulting in inconsistent data quality and an inability to achieve high-precision 3D geological modeling. Furthermore, traditional methods are difficult to achieve deep fusion of cross-sensor information, leading to insufficient engineering practicality.

Method used

By constructing a global three-dimensional coordinate system and introducing a spatial affine transformation model, multi-source heterogeneous data are uniformly mapped to the same spatial grid cell. Combined with a dynamic credibility assessment mechanism and an adaptive federated filtering algorithm, multi-scale spatiotemporal tensor decomposition and fusion are achieved. An improved moving cube algorithm is used to reconstruct the three-dimensional geological model, and spatial density clustering is combined to identify anomalous geological bodies.

Benefits of technology

It has achieved the construction of a high-precision, full-element three-dimensional geological information model, improved the automation level and engineering practicality of geological modeling, significantly enhanced the robustness and reliability of data fusion, and adapted to complex geological environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123118943A_ABST
    Figure CN123118943A_ABST
Patent Text Reader

Abstract

This invention proposes a method and system for modeling soil and composite strata based on multi-dimensional information fusion, belonging to the field of underground engineering and geological exploration information processing. The method includes: collecting and standardizing multi-source geological information; establishing a global three-dimensional coordinate system, performing spatial registration using spatial affine transformation, and constructing a basic geological feature tensor; extracting the probability distribution of sensor signal amplitudes, and calculating a dynamic credibility index by combining discrete information entropy and real-time signal-to-noise ratio; constructing a high-dimensional spatiotemporal geological tensor and decomposing it at multiple scales, and obtaining the fused comprehensive geological parameters by combining the credibility index and an adaptive federated filtering algorithm; using an improved moving cube algorithm combined with linear interpolation to reconstruct the boundary entities of the strata triangular mesh, and generating a full-element three-dimensional geological information model through spatial weighted interpolation and spatial density clustering. This invention effectively solves problems such as heterogeneous multi-source data, difficulty in controlling data quality, and the disconnect between fusion and modeling, and can provide strong support for underground engineering construction.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of underground engineering and geological exploration information processing technology, and particularly relates to a method and system for modeling soil and composite strata based on multi-dimensional information fusion. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] With the rapid advancement of infrastructure construction, the scale of underground projects such as underground tunnels, urban integrated pipe corridors, and deep foundation pits is constantly expanding, and the soil and complex geological environments faced during construction are becoming increasingly complex. Accurate perception and detailed characterization of geological conditions have become the core prerequisites for ensuring construction safety, optimizing tunneling parameters, and preventing geological disasters such as ground subsidence and sudden water inrush.

[0004] Currently, the geological exploration methods widely used in engineering practice mainly include ground-penetrating radar (GPR), laser vibrometers, core drilling, and seismic wave methods. These devices differ significantly in terms of detection principles, data formats, sampling frequencies, spatial resolution, and coverage. For example, GPR can acquire high-resolution electromagnetic wave reflection information from shallow layers, but its ability to identify aquifers is limited; laser vibrometers can capture micro-vibration signals and are suitable for dynamic stiffness assessment; core drilling provides the most direct stratigraphic lithology and mechanical parameters, but the information is sparse and the cost is high. The aforementioned multi-source heterogeneous data face significant challenges in unified representation, spatiotemporal alignment, and fusion utilization.

[0005] On the one hand, different sensors are susceptible to electromagnetic interference, construction vibration, local obstruction, or data loss in complex geological environments, resulting in inconsistent data quality. Existing methods lack a dynamic reliability assessment mechanism and cannot quantify the validity of data from each sensor at different spatial locations in real time. Low-quality data may directly contaminate the fusion results and even mislead construction decisions.

[0006] On the other hand, traditional geological modeling methods often treat data fusion and 3D modeling as separate processes. Most studies focus on the inversion or interpolation of single-type data, making it difficult to achieve deep fusion of cross-sensor information. Even when some methods attempt to fuse multi-source data, the output is often discrete geological attribute cloud maps or two-dimensional profiles, making it difficult to directly convert geological identification results into 3D geological models that can guide construction. Existing models often fail to fully express stratigraphic categories, geological parameters, distribution of anomalous geological bodies, and spatial structural relationships, resulting in insufficient engineering practicality and interpretability. Summary of the Invention

[0007] To overcome the shortcomings of the existing technologies, this invention provides a method and system for modeling soil and composite strata based on multi-dimensional information fusion. It aims to solve the problems of difficulty in uniformly representing multi-source heterogeneous geological data, inconsistency in spatiotemporal scales, and disconnect between identification results and modeling, and to achieve the construction of high-precision, full-element three-dimensional geological information models.

[0008] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions: The first aspect of this invention provides a method for modeling soil and composite strata based on multidimensional information fusion; A method for modeling soil and composite strata based on multidimensional information fusion, comprising: Collect multi-source geological information and perform standardized processing to uniformly map different data to the tunnel's three-dimensional coordinate system and spatial grid cells; A global three-dimensional coordinate system is constructed based on the standard tunnel measurement and control network. A spatial affine transformation model is introduced to complete the spatial registration of heterogeneous data and construct the basic geological feature tensor. Extract the probability distribution of signal amplitude and calculate the dynamic credibility index based on discrete information entropy and real-time signal-to-noise ratio; A high-dimensional spatiotemporal geological tensor is constructed based on the basic geological feature tensor and observation data with different physical dimensions, and multi-scale decomposition is performed. Combined with the dynamic credibility index, the integrated geological parameters after fusion are obtained by adaptive federated filtering. Based on the fused multidimensional geological feature tensor, an improved moving cube algorithm is adopted. Using the stratigraphic feature threshold as the extraction criterion, the intersection points of isosurfaces are calculated by linear interpolation for the grid edges that cross the stratigraphic boundary, and the voxel field is reconstructed into a continuous triangular grid boundary entity. Multidimensional continuous geological parameters are assigned to grid nodes through spatial weighted interpolation, and anomalous geological bodies are identified by spatial density clustering. Finally, a full-element three-dimensional geological information model is output.

[0009] As a further technical solution, a spatial affine transformation model is introduced to complete the spatial registration of heterogeneous data. The unified spatiotemporal mapping relationship is represented as follows:

[0010] in, This is a spatial point coordinate vector mapped to the tunnel's global three-dimensional coordinate system. This represents the original detection point coordinate vector of each sensor in its own local coordinate system. It is a three-dimensional rotation matrix. It is a three-dimensional translation vector.

[0011] As a further technical solution, the signal amplitude probability distribution is extracted, and a dynamic credibility index is calculated based on discrete information entropy and real-time signal-to-noise ratio, including: Extract the first Calculate the discrete information entropy by calculating the probability distribution of signal amplitude from each sensor within a time window:

[0012] in, The probability distribution of signal amplitude. For discrete information entropy; Based on the real-time signal-to-noise ratio, the first... The sensors are located in the space grid. Dynamic credibility assessment index:

[0013] in, It is a dynamic credibility assessment index. This represents the real-time signal-to-noise ratio of the currently acquired signal. This represents the maximum reference signal-to-noise ratio of the equipment under ideal calibration conditions. and These are the weighting adjustment coefficients for information entropy and signal-to-noise ratio, respectively. This is the set nonlinear decay control constant.

[0014] As a further technical solution, a high-dimensional spatiotemporal geological tensor is constructed and decomposed at multiple scales. Combined with a dynamic credibility index, adaptive federated filtering is used to obtain the fused comprehensive geological parameter characterization, including: Multi-scale tensor decomposition was performed on the constructed high-dimensional spatiotemporal geological tensor to extract the deep implicit common features of soil and composite strata. A multi-level federated filtering architecture is constructed, and the common state representation of each local sub-filter is constructed using the deep implicit common features. The observation data of each local sensor is mapped to a unified common feature state space to obtain the local optimal estimation state of each local sub-filter. The adaptive information sharing factor is obtained by fusing the dynamic reliability assessment index with the local state information represented by the inverse of the local state error covariance matrix. Based on the adaptive information sharing factor, the local optimal estimated states are weighted and fused, and the state of the global master filter is updated to obtain the fused comprehensive geological parameter characterization.

[0015] As a further technical solution, the adaptive information sharing factor is:

[0016] in, For the first A local sensor in The inverse of the state error covariance matrix at time t is the information matrix. The trace of the matrix is ​​used to convert the information content of the local state estimate into a scalar value. The larger the value, the smaller the uncertainty of the local state estimate and the higher the information contribution. The total number of effective sensors participating in this mesh fusion; For the first A local sensor is located in the space grid. Dynamic credibility assessment index.

[0017] As a further technical solution, the intersection points of isosurfaces are calculated for grid edges crossing stratigraphic boundaries using linear interpolation, as shown in the following formula:

[0018] in, These are the coordinates of the intersection point of the isosurface and the grid edge. and The coordinates of the two endpoints of the grid edge. and These are the stratigraphic feature values ​​corresponding to the two endpoints. This is the set stratigraphic boundary threshold.

[0019] As a further technical solution, multidimensional continuous geological parameters are assigned to grid nodes through spatial weighted interpolation, and anomalous geological bodies are identified using spatial density clustering. The final output is a full-element three-dimensional geological information model, including: The integrated geological parameter representation includes multidimensional continuous geological parameters, which are assigned to the reconstructed geometric grid nodes through a spatially weighted interpolation function. For any point on the geometric surface... physical properties Its calculation formula is defined as:

[0020] in, For point The number of voxel sampling points in the neighborhood. For the first The original fusion attribute values ​​of each sampling point Based on Euclidean distance Inverse distance weighting factor, This is the smoothness coefficient; For regions in space that exhibit discontinuous jumps or extreme parameter shifts, spatial density clustering algorithms are used to identify anomalous point clouds. The geometric closed contours of anomalous geological bodies are extracted using a defined neighborhood density threshold function, and the nesting and connectivity relationships with surrounding strata are calculated based on topological association rules. By integrating continuous attribute information and the topological relationships of anomalous geological bodies into the same spatial data structure, the final output is a full-element three-dimensional geological information model containing stratigraphic categories, geological parameters, anomalous bodies, and spatial relationships.

[0021] A second aspect of the present invention provides a soil and composite stratum modeling system based on multidimensional information fusion.

[0022] A soil and composite stratum modeling system based on multidimensional information fusion, comprising: The data acquisition module is configured to: acquire multi-source geological information and perform standardized processing, and map different data to the tunnel's three-dimensional coordinate system and spatial grid cells in a unified manner; The spatial registration and tensor construction module is configured to: construct a global three-dimensional coordinate system based on the standard tunnel measurement control network, introduce a spatial affine transformation model to complete the spatial registration of heterogeneous data, and construct a basic geological feature tensor. The dynamic credibility assessment module is configured to: extract the signal amplitude probability distribution and calculate the dynamic credibility index based on discrete information entropy and real-time signal-to-noise ratio; The information fusion module is configured to: construct a high-dimensional spatiotemporal geological tensor based on the basic geological feature tensor and observation data of different physical dimensions, perform multi-scale decomposition, combine dynamic credibility index, and use adaptive federated filtering to obtain the fused comprehensive geological parameter characterization. The 3D geometric reconstruction module is configured to: based on the fused multi-dimensional geological feature tensor, use an improved moving cube algorithm, take the stratigraphic feature threshold as the extraction criterion, calculate the intersection points of isosurfaces for the grid edges that cross the stratigraphic boundary through linear interpolation, and reconstruct the voxel field into a continuous triangular grid boundary entity. The attribute mapping and model generation module is configured to: assign multidimensional continuous geological parameters to grid nodes through spatial weighted interpolation, identify anomalous geological bodies using spatial density clustering, and finally output a full-element three-dimensional geological information model.

[0023] A third aspect of the present invention provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps of a soil and composite stratum modeling method based on multidimensional information fusion as described in the first aspect of the present invention.

[0024] A fourth aspect of the present invention provides an electronic device, including a memory, a processor, and a program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps in a method for modeling soil and composite strata based on multidimensional information fusion as described in the first aspect of the present invention.

[0025] The above one or more technical solutions have the following beneficial effects: (1) This invention constructs a global three-dimensional coordinate system and introduces a spatial affine transformation model to uniformly map multi-source heterogeneous data to the same spatial grid cell, forming a standardized basic geological feature tensor. This effectively solves the problem of differences in data format, spatial resolution, and coverage between different sensors. Subsequently, by using a dynamic reliability assessment mechanism based on the coupling of information entropy and signal-to-noise ratio, the data quality of each sensor in different spatial grid cells can be quantified in real time, and low-quality data that has been contaminated can be automatically identified and downweighted or removed. At the same time, the dynamic reliability assessment index and the local state error covariance are used together to calculate the adaptive information sharing factor. The information contribution weight of each sensor is dynamically adjusted according to the data quality and the degree of determination of local state estimation, which effectively prevents abnormal data from contaminating the fusion results and significantly improves the robustness and reliability of the fusion system.

[0026] (2) This invention integrates multi-scale spatiotemporal tensor decomposition and adaptive federated filtering algorithms. Multi-scale tensor decomposition extracts deep, implicit common features of soil and composite strata, and uses these common features to constrain the consistency of multi-source local state fusion. Simultaneously, by combining the dynamic reliability evaluation index and local state error covariance, the information contribution weights of each local sub-filter are adaptively allocated, effectively addressing complex situations such as abrupt changes in strata and information conflicts at strata interfaces. Compared to traditional single fusion algorithms, this method is more adaptable to nonlinear geological scenarios, achieving deep collaborative fusion of multi-sensor data and outputting comprehensive geological parameters that better reflect actual strata conditions.

[0027] (3) This invention employs an improved moving cube algorithm and spatial weighted interpolation technology to automatically reconstruct the fused multi-dimensional geological feature tensor into triangular mesh boundary entities carrying continuous physical properties, and combines it with a spatial density clustering algorithm to identify anomalous geological bodies and their topological relationships. The entire process realizes integrated processing from data fusion to 3D modeling, and finally outputs a full-element 3D geological information model containing stratigraphic categories, geological parameters, anomalous body distribution, and spatial structural relationships, significantly improving the automation level and engineering practical value of geological modeling.

[0028] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0029] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0030] Figure 1 This is a flowchart of the method in the first embodiment.

[0031] Figure 2This is a schematic diagram of the overall process of the method in the first embodiment.

[0032] Figure 3 This is a system structure diagram of the second embodiment. Detailed Implementation

[0033] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0034] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations of the present invention.

[0035] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0036] Example 1 This embodiment discloses a method for modeling soil and complex strata based on multi-dimensional information fusion. It constructs a basic geological feature tensor by acquiring and spatially registering multi-source geological data, dynamically evaluates data credibility by combining information entropy and signal-to-noise ratio, and then constructs a high-dimensional spatiotemporal geological tensor and performs multi-scale decomposition. It uses adaptive federated filtering to achieve deep fusion of cross-modal information, and finally outputs a full-element three-dimensional geological model by improving the moving cube algorithm and attribute-geometry coupling mapping, which effectively improves the accuracy and efficiency of complex strata modeling.

[0037] like Figure 1 and Figure 2 As shown, a method for modeling soil and composite strata based on multi-dimensional information fusion includes: Step S1: Collect multi-source geological information and perform standardized processing, mapping different data to the tunnel's three-dimensional coordinate system and spatial grid cells.

[0038] Geological information acquisition equipment, such as ground-penetrating radar and laser vibrometers, was deployed in the underground soil and various composite subsurface exploration areas to obtain multi-source geological information under unknown tunnel conditions. This information included soil type, water content, porosity, soil strength, and density. The multi-source data underwent standardization processing, including format unification and filtering to remove noise, and the different data were uniformly mapped to the tunnel's three-dimensional coordinate system and spatial grid cells.

[0039] Step S2: Construct a global three-dimensional coordinate system based on the standard tunnel measurement control network, introduce a spatial affine transformation model to complete the spatial registration of heterogeneous data, and construct the basic geological feature tensor.

[0040] A global three-dimensional coordinate system is constructed based on the standard tunnel measurement control network, and a spatial affine transformation model is introduced to complete the spatial registration of heterogeneous data. Its unified spatiotemporal mapping relationship can be expressed as:

[0041] in, This is a spatial point coordinate vector mapped to the tunnel's global three-dimensional coordinate system. This represents the original detection point coordinate vector of each sensor in its own local coordinate system. It is a three-dimensional rotation matrix. This is a three-dimensional translation vector. Through this mapping relationship, a basic geological feature tensor containing multi-dimensional physical properties is constructed, laying a unified data foundation for subsequent deep information fusion.

[0042] Step S3: Extract the signal amplitude probability distribution and calculate the dynamic credibility index based on discrete information entropy and real-time signal-to-noise ratio.

[0043] A dynamic credibility assessment mechanism based on the coupling of information entropy and signal-to-noise ratio is constructed to monitor the data fluctuation status of each sensor in real time and calculate the credibility of different sensors in different spatial grid cells, so as to isolate or downweight contaminated abnormal data and ensure the efficient availability of data.

[0044] First, extract the first... Each sensor in the time window Signal amplitude probability distribution within Calculate its discrete information entropy :

[0045] Information entropy is used to characterize the disorder and chaos of a signal response. The more severe the disturbance of a signal by vibration or electromagnetic clutter, the more divergent its probability distribution and the higher the entropy value. The larger the value, the better. Then, combining the real-time signal-to-noise ratio, the first... The sensors are located in the space grid. Dynamic credibility assessment index Its calculation formula is defined as:

[0046] in, This represents the real-time signal-to-noise ratio of the currently acquired signal. This is the maximum reference signal-to-noise ratio of the device under ideal calibration conditions; and The weighting adjustment coefficients for information entropy and signal-to-noise ratio are respectively (satisfying...) ); This is the set nonlinear decay control constant. Through this evaluation model, the system can adaptively quantify and score the data quality within each grid cell, effectively eliminating data at the failure boundary. This evaluation index is used as... As a dynamic scalar, it will be directly used as the core adjustment parameter of the next fusion algorithm.

[0047] Step S4: Construct a high-dimensional spatiotemporal geological tensor based on the basic geological feature tensor and observation data with different physical dimensions, and perform multi-scale decomposition. Combined with the dynamic credibility index, use adaptive federated filtering to obtain the fused comprehensive geological parameter characterization.

[0048] Based on the aforementioned spatiotemporally unified geological feature tensor and dynamic reliability assessment results, a cross-modal information fusion algorithm based on multi-scale spatiotemporal tensor decomposition and adaptive federated filtering is adopted. First, observation data with different physical dimensions are constructed into a high-dimensional spatiotemporal geological tensor. Multi-scale tensor decomposition technology is used to extract deep, implicit common features of soil and composite strata, which characterize the stratigraphic structure and geological property variation patterns jointly reflected by different sensor observation data within the current spatial grid. Then, based on the deep, implicit common features, a unified state feature expression for each local sub-filter is established, converting the local observation information from different sensors into local state estimation vectors with consistent state dimensions and geological semantics. On this basis, a multi-level federated filtering architecture is constructed. Based on the dynamic reliability assessment index and the inverse deterministic information contribution weights of the state error covariance matrix of each local sub-filter, the local state estimation vectors are weighted and fused to complete the state update and optimal estimation of the global main filter.

[0049] In this architecture, deep implicit shared features are used to establish a unified state feature representation for each local sub-filter, ensuring that the local state estimation vectors formed by different sensors have consistent state dimensions and geological semantics; is the adaptive information sharing factor for the i-th local sub-filter, used to characterize the degree of information contribution of this local sub-filter to the global state update. The adaptive information sharing factor is calculated based on the dynamic reliability assessment index and the inverse of the local state error covariance matrix, as follows:

[0050] in, For the first A local sensor in The inverse of the state error covariance matrix at time t is the information matrix. The trace of the matrix is ​​used to convert the information content of the local state estimate into a scalar value. The larger the value, the smaller the uncertainty of the local state estimate and the higher the information contribution. The total number of effective sensors participating in this mesh fusion; For the first A local sensor is located in the space grid. Dynamic credibility assessment index.

[0051] Update equation of global fusion state Defined as:

[0052] in, In order to be in The optimal fused state vector of global geological information at time t, i.e., the comprehensive geological parameter representation after fusion; The total number of valid sensor nodes participating in the current mesh fusion; For the first The local optimal estimation vector of a local sensor subsystem after state prediction and measurement update.

[0053] Step S5: Based on the fused multi-dimensional geological feature tensor, an improved moving cube algorithm is adopted, and the stratigraphic feature threshold is used as the extraction criterion. The isosurface intersection points of the grid edges that cross the stratigraphic boundary are calculated by linear interpolation, and the voxel field is reconstructed into a continuous triangular grid boundary entity.

[0054] Based on the cross-sensor fusion data obtained in the first aspect, an integrated 3D modeling processing system is established to transform geological information into a spatial entity architecture, ultimately outputting a complete 3D geological information model containing stratigraphic categories, geological parameters, distribution of anomalous geological bodies, and spatial structural relationships. Specifically, the system first receives a multi-dimensional geological feature tensor based on a spatial grid. In the 3D geometric boundary reconstruction stage, an improved Moving Cubes algorithm is used, employing stratigraphic feature thresholds from the fused data as extraction criteria. For grid edges crossing stratigraphic boundaries, linear interpolation formulas are used to accurately calculate the coordinates of the intersection points of isosurfaces and grid edges. Its mathematical expression is:

[0055] in, and The coordinates of the two endpoints of the grid edge. and These are the stratigraphic feature values ​​corresponding to these two endpoints. This is the set stratigraphic boundary threshold.

[0056] By traversing all voxel meshes and connecting these intersections, the system reconstructs the discrete voxel field into a topologically continuous triangular mesh boundary entity.

[0057] Step S6: Multidimensional continuous geological parameters are assigned to grid nodes through spatial weighted interpolation, and anomalous geological bodies are identified by spatial density clustering. Finally, a full-element three-dimensional geological information model is output.

[0058] Based on the geometric reconstruction, a deep coupling mapping of "attribute-geometry" is performed. Multidimensional continuous geological parameters such as water content, porosity, and soil strength contained in the fused data are assigned to the reconstructed geometric mesh nodes through a spatially weighted interpolation function. For any point on the geometric surface... physical properties Its calculation formula is defined as:

[0059] in, For point The number of voxel sampling points in the neighborhood. For the first The original fusion attribute values ​​of each sampling point Based on Euclidean distance Inverse distance weighting factor, This represents the smoothness coefficient. Furthermore, for regions exhibiting discontinuous jumps or extreme parameter shifts in space, the system utilizes a spatial density clustering algorithm to identify anomalous point clouds, using a defined neighborhood density threshold function. Extract the geometric closed contour of the anomalous geological body and calculate its nesting and connectivity with the surrounding strata based on topological association rules.

[0060] By integrating continuous attribute information and the topological relationships of anomalous geological bodies into the same spatial data structure, the final output is a full-element three-dimensional geological information model containing stratigraphic categories, geological parameters, anomalous bodies, and spatial relationships.

[0061] Example 2 This embodiment discloses a soil and composite stratum modeling system based on multi-dimensional information fusion; like Figure 3 As shown, a soil and composite stratum modeling system based on multi-dimensional information fusion includes: The data acquisition module is configured to: acquire multi-source geological information and perform standardized processing, and map different data to the tunnel's three-dimensional coordinate system and spatial grid cells in a unified manner; The spatial registration and tensor construction module is configured to: construct a global three-dimensional coordinate system based on the standard tunnel measurement control network, introduce a spatial affine transformation model to complete the spatial registration of heterogeneous data, and construct a basic geological feature tensor. The dynamic credibility assessment module is configured to: extract the signal amplitude probability distribution and calculate the dynamic credibility index based on discrete information entropy and real-time signal-to-noise ratio; The information fusion module is configured to: construct a high-dimensional spatiotemporal geological tensor based on the basic geological feature tensor and observation data of different physical dimensions, perform multi-scale decomposition, combine dynamic credibility index, and use adaptive federated filtering to obtain the fused comprehensive geological parameter characterization. The 3D geometric reconstruction module is configured to: based on the fused multi-dimensional geological feature tensor, use an improved moving cube algorithm, take the stratigraphic feature threshold as the extraction criterion, calculate the intersection points of isosurfaces for the grid edges that cross the stratigraphic boundary through linear interpolation, and reconstruct the voxel field into a continuous triangular grid boundary entity. The attribute mapping and model generation module is configured to: assign multidimensional continuous geological parameters to grid nodes through spatial weighted interpolation, identify anomalous geological bodies using spatial density clustering, and finally output a full-element three-dimensional geological information model.

[0062] Example 3 The purpose of this embodiment is to provide a computer-readable storage medium.

[0063] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in a soil and composite stratum modeling method based on multidimensional information fusion as described in Example 1.

[0064] Example 4 The purpose of this embodiment is to provide an electronic device.

[0065] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in a method for modeling soil and composite strata based on multidimensional information fusion as described in Embodiment 1.

[0066] The steps and methods involved in the apparatuses of Embodiments 2, 3, and 4 above correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.

[0067] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.

[0068] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for modeling soil and composite ground based on multi-dimensional information fusion, characterized in that, include: Collect multi-source geological information and perform standardized processing to uniformly map different data to the tunnel's three-dimensional coordinate system and spatial grid cells; A global three-dimensional coordinate system is constructed based on the standard tunnel measurement and control network. A spatial affine transformation model is introduced to complete the spatial registration of heterogeneous data and construct the basic geological feature tensor. Extract the probability distribution of signal amplitude and calculate the dynamic credibility index based on discrete information entropy and real-time signal-to-noise ratio; A high-dimensional spatiotemporal geological tensor is constructed based on the basic geological feature tensor and observation data with different physical dimensions, and multi-scale decomposition is performed. Combined with the dynamic credibility index, the integrated geological parameters after fusion are obtained by adaptive federated filtering. Based on the fused multidimensional geological feature tensor, an improved moving cube algorithm is adopted. Using the stratigraphic feature threshold as the extraction criterion, the intersection points of isosurfaces are calculated by linear interpolation for the grid edges that cross the stratigraphic boundary, and the voxel field is reconstructed into a continuous triangular grid boundary entity. Multidimensional continuous geological parameters are assigned to grid nodes through spatial weighted interpolation, and anomalous geological bodies are identified by spatial density clustering. Finally, a full-element three-dimensional geological information model is output.

2. The method for modeling soil and composite strata based on multi-dimensional information fusion as described in claim 1, characterized in that, In the process of introducing a spatial affine transformation model to complete the spatial registration of heterogeneous data, the unified spatiotemporal mapping relationship is represented as: in, This is a spatial point coordinate vector mapped to the tunnel's global three-dimensional coordinate system. This represents the original detection point coordinate vector of each sensor in its own local coordinate system. It is a three-dimensional rotation matrix. It is a three-dimensional translation vector.

3. The method for modeling soil and composite strata based on multi-dimensional information fusion as described in claim 1, characterized in that, Extract the signal amplitude probability distribution, and calculate the dynamic credibility index based on discrete information entropy and real-time signal-to-noise ratio, including: Extract the first Calculate the discrete information entropy by calculating the probability distribution of signal amplitude from each sensor within a time window: in, The probability distribution of signal amplitude. For discrete information entropy; Based on the real-time signal-to-noise ratio, the first... The sensors are located in the space grid. Dynamic credibility assessment index: in, It is a dynamic credibility assessment index. This represents the real-time signal-to-noise ratio of the currently acquired signal. This represents the maximum reference signal-to-noise ratio of the equipment under ideal calibration conditions. and These are the weighting adjustment coefficients for information entropy and signal-to-noise ratio, respectively. This is the set nonlinear decay control constant.

4. The method for modeling soil and composite strata based on multi-dimensional information fusion as described in claim 1, characterized in that, A high-dimensional spatiotemporal geological tensor is constructed and decomposed at multiple scales. Combined with a dynamic credibility index, adaptive federated filtering is used to obtain the fused comprehensive geological parameter characterization, including: Multi-scale tensor decomposition was performed on the constructed high-dimensional spatiotemporal geological tensor to extract the deep implicit common features of soil and composite strata. Based on the deep implicit common features, a unified state feature expression for each local sub-filter is established, so that the local state estimates formed by different sensors have consistent state dimension and geological semantics. The adaptive information sharing factor is calculated based on the inverse of the local state error covariance matrix and the dynamic credibility evaluation index. Based on the adaptive information sharing factor, the local state estimates of each local sub-filter are weighted and fused to complete the state update of the global main filter, and the fused comprehensive geological parameter characterization is obtained.

5. The method for modeling soil and composite strata based on multi-dimensional information fusion as described in claim 4, characterized in that, The adaptive information sharing factor is: in, For the first A local sensor in The inverse of the state error covariance matrix at time t is the information matrix. The trace of the matrix is ​​used to convert the information content of the local state estimate into a scalar value. The larger the value, the smaller the uncertainty of the local state estimate and the higher the information contribution. The total number of effective sensors participating in this mesh fusion; For the first A local sensor is located in the space grid. Dynamic credibility assessment index.

6. The method for modeling soil and composite strata based on multi-dimensional information fusion as described in claim 1, characterized in that, The intersection points of isosurfaces are calculated using linear interpolation for grid edges that cross stratigraphic boundaries, as shown in the following formula: in, These are the coordinates of the intersection point of the isosurface and the grid edge. and The coordinates of the two endpoints of the grid edge. and These are the stratigraphic feature values ​​corresponding to the two endpoints. This is the set stratigraphic boundary threshold.

7. The method for modeling soil and composite strata based on multi-dimensional information fusion as described in claim 1, characterized in that, Multidimensional continuous geological parameters are assigned to grid nodes through spatial weighted interpolation, and anomalous geological bodies are identified using spatial density clustering. The final output is a full-element 3D geological information model, including: The integrated geological parameter representation includes multidimensional continuous geological parameters, which are assigned to the reconstructed geometric grid nodes through a spatially weighted interpolation function. For any point on the geometric surface... physical properties Its calculation formula is defined as: in, For point The number of voxel sampling points in the neighborhood. For the first The original fusion attribute values ​​of each sampling point Based on Euclidean distance Inverse distance weighting factor, This is the smoothness coefficient; For regions in space that exhibit discontinuous jumps or extreme parameter shifts, spatial density clustering algorithms are used to identify anomalous point clouds. The geometric closed contours of anomalous geological bodies are extracted using a defined neighborhood density threshold function, and the nesting and connectivity relationships with surrounding strata are calculated based on topological association rules. By integrating continuous attribute information and the topological relationships of anomalous geological bodies into the same spatial data structure, the final output is a full-element three-dimensional geological information model containing stratigraphic categories, geological parameters, anomalous bodies, and spatial relationships.

8. A soil and composite stratum modeling system based on multi-dimensional information fusion, characterized in that, include: The data acquisition module is configured to: acquire multi-source geological information and perform standardized processing, and map different data to the tunnel's three-dimensional coordinate system and spatial grid cells in a unified manner; The spatial registration and tensor construction module is configured to: construct a global three-dimensional coordinate system based on the standard tunnel measurement control network, introduce a spatial affine transformation model to complete the spatial registration of heterogeneous data, and construct a basic geological feature tensor. The dynamic credibility assessment module is configured to: extract the signal amplitude probability distribution and calculate the dynamic credibility index based on discrete information entropy and real-time signal-to-noise ratio; The information fusion module is configured to: construct a high-dimensional spatiotemporal geological tensor based on the basic geological feature tensor and observation data of different physical dimensions, perform multi-scale decomposition, combine dynamic credibility index, and use adaptive federated filtering to obtain the fused comprehensive geological parameter characterization. The 3D geometric reconstruction module is configured to: based on the fused multi-dimensional geological feature tensor, use an improved moving cube algorithm, take the stratigraphic feature threshold as the extraction criterion, calculate the intersection points of isosurfaces for the grid edges that cross the stratigraphic boundary through linear interpolation, and reconstruct the voxel field into a continuous triangular grid boundary entity. The attribute mapping and model generation module is configured to: assign multidimensional continuous geological parameters to grid nodes through spatial weighted interpolation, identify anomalous geological bodies using spatial density clustering, and finally output a full-element three-dimensional geological information model.

9. A computer-readable storage medium having a program stored thereon, characterized in that, When executed by the processor, the program implements the steps in the soil and composite strata modeling method based on multidimensional information fusion as described in any one of claims 1-7.

10. An electronic device comprising a memory, a processor, and a program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the soil and composite strata modeling method based on multidimensional information fusion as described in any one of claims 1-7.