An environmental civil risk assessment method and system based on multi-source data fusion

Abstract

The application provides a kind of environment civil engineering risk assessment method and system based on multi-source data fusion, belongs to civil engineering risk assessment technical field, the method comprises: obtaining the multi-source data of target construction area, identifying the discrete semantic point of target construction area, and constructing spatial semantic field through all discrete semantic points combined with three-dimensional geological space grid;Space semantic field is used as priori constraint, the spatial interpolation process of time series InSAR deformation data in multi-source data is weighted guided, and continuous deformation rate field is obtained;With continuous deformation rate field as initial boundary condition, drive finite element numerical model to carry out stress-strain forward modeling, obtain the maximum equivalent plastic strain distribution under current stress state;The maximum equivalent plastic strain distribution is superimposed on the spatial semantic field to determine the risk assessment level of target construction area.The technical scheme provided by the application can realize the evaluation of deep instability risk caused by geological structure change in civil engineering construction area.

Description

Technical Field

[0001] This application relates to the field of civil engineering risk assessment technology, and more specifically, to an environmental civil engineering risk assessment method and system based on multi-source data fusion. Background Technology

[0002] During the construction and operation of civil engineering projects, uncertainties arise from multiple environmental factors such as geology, hydrology, and climate. Environmental civil risk assessment has emerged as a response to this need. This assessment focuses on identifying and quantifying the interaction risks between the natural environment and engineering structures, including potential threats to the stability of foundations, slopes, tunnels, and dams such as landslides, floods, earthquakes, soil liquefaction, erosion, and extreme weather events. Through on-site monitoring, remote sensing data, numerical simulation, and statistical analysis, it assesses the probability of disasters and their potential consequences for engineering safety and function. With the intensification of climate change and the increasing complexity of engineering environments, environmental civil risk assessment has become a key technical support for ensuring the long-term safety and resilience of infrastructure.

[0003] Chinese patent CN120524242A discloses an intelligent civil engineering construction management method and system, relating to the field of civil engineering construction management technology. The method includes the following steps: acquiring vibration data from N construction devices and mixed vibration data from M monitoring points; determining whether the mixed vibration data exceeds a preset threshold; if so, performing multi-feature matching with a preset vibration fingerprint database; determining the responsible device based on the matching results and outputting the contribution and measured vibration values ​​in descending order. This invention eliminates environmental interference through preprocessing, and then achieves multi-dimensional decomposition of the mixed vibration data through independent component analysis. Combining the dual criteria of time-domain waveform differences and frequency-domain energy matching, it significantly improves the accuracy of responsible device identification. Furthermore, it constructs a three-dimensional geological model using ground-penetrating radar scanning and soil layer analysis data to accurately reflect the spatial distribution and attenuation characteristics of different soil layers, avoiding misjudgment of responsible devices due to soil layer differences. This solution overcomes the problem of equipment misjudgment caused by the interference of soil layer differences on vibration propagation characteristics by acquiring vibration data of construction equipment and mixed vibration data of monitoring points, and by using preset threshold judgment combined with multi-feature matching of vibration fingerprint database, as well as the technical means of constructing a three-dimensional geological model by ground penetrating radar scanning and soil layer test data. However, this solution cannot cope with the deep stability risks caused by changes in geological structure and rock mass integrity. These risks do not depend on equipment vibration signals, deform slowly and have no significant vibration precursors, and cannot be effectively detected by vibration monitoring. This makes it impossible to identify the deep instability risks caused by changes in geological structure in civil construction areas. Therefore, how to assess the deep instability risks caused by changes in geological structure in civil construction areas has become a difficult problem for the industry. Summary of the Invention

[0004] This application provides an environmental civil engineering risk assessment method and system based on multi-source data fusion, which can assess the risk of deep instability caused by geological structural changes in civil engineering construction areas.

[0005] Firstly, this application provides an environmental civil risk assessment method based on multi-source data fusion, comprising the following steps: Acquire multi-source data of the target construction area, identify discrete semantic points with spatial location attributes in the target construction area based on the geological survey report text in the multi-source data, and construct a spatial semantic field characterizing the trend of rock mass integrity by combining all discrete semantic points with a three-dimensional geological spatial grid. Using the spatial semantic field as a priori constraint, the spatial interpolation process of the temporal InSAR deformation data in the multi-source data is weighted and guided to obtain a continuous deformation rate field corrected by geological semantics. Using the continuous deformation rate field as the initial boundary condition, the finite element numerical model is driven to perform forward modeling of stress and strain to obtain the maximum equivalent plastic strain distribution under the current stress state. The maximum equivalent plastic strain distribution is superimposed onto the spatial semantic field, and the risk assessment level of the target construction area is determined by comparing the spatial overlap between the equivalent plastic strain concentration area and the rock mass integrity trend abrupt change area.

[0006] In some embodiments, the multi-source data includes geological survey report text and time-series InSAR deformation data.

[0007] In some embodiments, obtaining multi-source data of the target construction area specifically includes: Based on the boundary of the target construction area, extract the geological survey report text corresponding to the boundary from the digital archive of the regional geological data center; The single-view complex image sequence covering the boundary of the range is queried from the synthetic aperture radar satellite archive data. Differential interferometry is performed on the single-view complex image sequence to generate temporal InSAR deformation data containing temporal cumulative deformation and deformation rate information. Multi-source data of the target construction area are obtained by combining the geological survey report text and the time-series InSAR deformation data.

[0008] In some embodiments, identifying discrete semantic points with spatial location attributes in the target construction area based on the geological survey report text in the multi-source data specifically includes: Geological exploration report text is obtained from the multi-source data, and the borehole location coordinates associated with the qualitative description statements in the geological exploration report text are identified. Based on the borehole location coordinates, the qualitative description statement is semantically transformed to obtain discrete semantic points with spatial location attributes in the target construction area.

[0009] In some embodiments, constructing a spatial semantic field characterizing the rock mass integrity trend by combining all discrete semantic points with a three-dimensional geological spatial grid specifically includes: All discrete semantic points are quantitatively labeled with rock mass integrity level according to the qualitative description statements they are associated with, resulting in a set of discrete control points carrying rock mass integrity classification values. A three-dimensional geological space grid is established using the circumscribed cuboid of the region covered by the discrete control point set as the boundary. The radial basis function interpolation algorithm is used to interpolate each grid point of the three-dimensional geological space grid by using the discrete control point set as the interpolation source point, forming a spatial semantic field that is continuously distributed in three-dimensional space and represents the trend of rock mass integrity.

[0010] In some embodiments, the spatial semantic field is used as a priori constraint to weight and guide the spatial interpolation process of temporal InSAR deformation data in the multi-source data, resulting in a geologically semantically corrected continuous deformation rate field, specifically including: The spatial semantic field is projected onto discrete locations that match coherent target points in the temporal InSAR deformation data in the same spatial coordinate system, and the rock mass integrity trend value at each coherent target point is obtained. A kernel function weight matrix for spatial anisotropy is constructed using the rock mass integrity trend value at each coherent target point as a variable. Using the kernel function weight matrix as the prior constraint matrix in the Kriging interpolation equation system, spatial interpolation is performed on the deformation rate of coherent target points in the time-series InSAR deformation data to obtain a continuous deformation rate field corrected by geological semantics.

[0011] In some embodiments, projecting the spatial semantic field onto discrete locations matching coherent target points in the temporal InSAR deformation data under the same spatial coordinate system, and obtaining the rock mass integrity trend value at each coherent target point specifically includes: Obtain the geodetic coordinates of each coherent target point in the temporal InSAR deformation data, and register the three-dimensional geological spatial grid where the spatial semantic field is located to the geodetic coordinate system through coordinate transformation to form a unified spatial reference. Under the unified spatial reference, the column of each coherent target point in the three-dimensional geological spatial grid is determined by the horizontal position of each coherent target point. The grid is indexed layer by layer along the vertical downward direction of the column with the grid resolution of the spatial semantic field as the step size. The interpolation result stored in the first grid point with a non-empty spatial semantic value is extracted as the rock mass integrity trend value of the corresponding coherent target point.

[0012] Secondly, this application provides an environmental civil engineering risk assessment system based on multi-source data fusion, used to execute an environmental civil engineering risk assessment method based on multi-source data fusion. The system includes: The acquisition module is used to acquire multi-source data of the target construction area, identify discrete semantic points with spatial location attributes in the target construction area based on the geological survey report text in the multi-source data, and construct a spatial semantic field representing the trend of rock mass integrity by combining all discrete semantic points with a three-dimensional geological spatial grid. The processing module is used to use the spatial semantic field as a priori constraint to weight and guide the spatial interpolation process of the temporal InSAR deformation data in the multi-source data, so as to obtain a continuous deformation rate field corrected by geological semantics. The processing module is also used to drive the finite element numerical model to perform forward modeling of stress and strain using the continuous deformation rate field as the initial boundary condition, so as to obtain the maximum equivalent plastic strain distribution under the current stress state. The execution module is used to superimpose the maximum equivalent plastic strain distribution onto the spatial semantic field, and determine the risk assessment level of the target construction area by comparing the spatial overlap between the equivalent plastic strain concentration area and the rock mass integrity trend change area.

[0013] Thirdly, this application provides a computer device, the computer device including a memory and a processor, the memory storing code, and the processor being configured to acquire the code and execute the above-described environmental civil risk assessment method based on multi-source data fusion.

[0014] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described environmental and civil risk assessment method based on multi-source data fusion.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: The environmental civil engineering risk assessment method and system based on multi-source data fusion provided in this application first acquires multi-source data of the target construction area. Based on the geological survey report text in the multi-source data, discrete semantic points with spatial location attributes in the target construction area are identified. A spatial semantic field representing the rock mass integrity trend is constructed by combining all discrete semantic points with a three-dimensional geological spatial grid. Second, the spatial semantic field is used as a priori constraint to weight and guide the spatial interpolation process of the temporal InSAR deformation data in the multi-source data to obtain a continuous deformation rate field corrected by geological semantics. Then, the continuous deformation rate field is used as the initial boundary condition to drive the finite element numerical model to perform forward simulation of stress and strain, and obtain the maximum equivalent plastic strain distribution under the current stress state. Finally, the maximum equivalent plastic strain distribution is superimposed on the spatial semantic field, and the risk assessment level of the target construction area is determined by comparing the spatial overlap between the equivalent plastic strain concentration area and the rock mass integrity trend abrupt change area.

[0016] Therefore, this application can assess the risk of deep instability caused by geological structural changes in civil construction areas. Firstly, by transforming qualitative descriptions in geological survey reports into discrete semantic points with spatial location attributes and constructing a spatial semantic field, the assessment of rock mass integrity in unstructured geological texts is non-destructively converted into a computable three-dimensional scalar field. This allows prior geological knowledge to participate in subsequent quantitative analysis in numerical form, avoiding the limitation of traditional assessment methods where geological information is only used for qualitative reference and cannot be deeply integrated with deformation data. Secondly, the spatial semantic field is used as a priori constraint to weight and guide the interpolation process of temporal InSAR deformation data, ensuring that the deformation rate field maintains the sharpness of geological boundaries at abrupt changes in rock mass integrity. This avoids the problem of smooth transition distortion at geological boundaries caused by conventional spatial interpolation methods, thus obtaining a continuous deformation rate field that better reflects the characteristics of real geological structures. Then, using geological language... The finite element forward modeling driven by the continuously deformable rate field with semantic correction establishes a quantitative correlation between surface deformation observations and underground stress-strain responses through a mechanical model. This overcomes the limitation of a single data source in revealing the essential mechanism of deformation driving. By capturing the maximum equivalent plastic strain distribution throughout the loading process, it ensures that no plastic damage hazard area is missed, thereby identifying the deep instability risk caused by geological structural changes in civil construction areas. Finally, the maximum equivalent plastic strain distribution is spatially overlaid and compared with the spatial semantic field. The degree of agreement between the plastic strain concentration area and the area of ​​abrupt change in rock mass integrity trend is used as the risk criterion. The risk level is comprehensively determined from the dual perspectives of surface deformation response and underground rock mass structural vulnerability, realizing the quantitative risk assessment of multi-source heterogeneous data in a geomechanical sense. In summary, the technical solution provided in this application can realize the assessment of deep instability risks caused by geological structural changes in civil construction areas. Attached Figure Description

[0017] Figure 1 This is an exemplary flowchart of an environmental civil risk assessment method based on multi-source data fusion, as shown in some embodiments of this application; Figure 2 This is an exemplary flowchart illustrating the determination of a continuous deformation rate field according to some embodiments of this application; Figure 3 This is a schematic diagram of the structure of an environmental civil risk assessment system based on multi-source data fusion, as shown in some embodiments of this application; Figure 4 This is a schematic diagram of the structure of a computer device for implementing an environmental civil risk assessment method based on multi-source data fusion, according to some embodiments of this application. Detailed Implementation

[0018] To better understand the technical solution of this application, the technical solution of this application will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] refer to Figure 1 The figure is an exemplary flowchart of an environmental civil risk assessment method based on multi-source data fusion, according to some embodiments of this application. The figure mainly includes the following steps: In step S101, multi-source data of the target construction area is acquired, and discrete semantic points with spatial location attributes of the target construction area are identified based on the geological survey report text in the multi-source data. A spatial semantic field representing the trend of rock mass integrity is constructed by combining all discrete semantic points with a three-dimensional geological spatial grid.

[0020] It should be noted that the target construction area in this application refers to the specific geographical space area that needs to be assessed for environmental civil engineering risks, as defined by the civil engineering planning red line. This area includes the engineering structures under construction or planned and the key soil and rock masses around them that are affected by construction disturbances.

[0021] In some embodiments, acquiring multi-source data of the target construction area is achieved through the following steps: Based on the boundary of the target construction area, extract the geological survey report text corresponding to the boundary from the digital archive of the regional geological data center; The single-view complex image sequence covering the boundary of the range is queried from the synthetic aperture radar satellite archive data. Differential interferometry is performed on the single-view complex image sequence to generate temporal InSAR deformation data containing temporal cumulative deformation and deformation rate information. Multi-source data of the target construction area are obtained by combining the geological survey report text and the time-series InSAR deformation data.

[0022] Specifically, the boundary of the target construction area is determined based on the engineering planning red line. The latest version of the engineering geological survey report corresponding to this boundary is retrieved from the digital archive of the regional geological data center. Unstructured text such as borehole coordinates, stratigraphic descriptions, and rock mass integrity assessments in the report are transformed into structured geological survey report text through entity recognition and attribute extraction. The geological survey report text refers to the geological description of the target construction area recorded in natural language. Then, a single-view complex image sequence covering the boundary is retrieved from the synthetic aperture radar satellite archive data. A single-view complex image sequence refers to a set of unstructured images acquired sequentially at fixed time intervals by the synthetic aperture radar satellite in repeating orbit observation mode for the same target area. The original image data, processed by multiple-look averaging, is stored in complex form for each image. The real and imaginary parts correspond to the amplitude and phase components of the echo signal, respectively. Temporally adjacent image pairs in the sequence of single-look complex images are registered and multiplied by conjugate to generate differential interferograms. The differential interferograms are then processed by terrain phase removal, filtering, and phase unwrapping. The unwrapped differential interferometric phase is used for temporal inversion to calculate the temporal cumulative deformation and deformation rate of each coherent target point relative to the reference time at each imaging time. The deformation rate refers to the change in surface displacement per unit time, thus obtaining temporal InSAR deformation data. Finally, the geological survey report text and the temporal InSAR deformation data are combined to obtain multi-source data of the target construction area.

[0023] It should be noted that the multi-source data in this application includes geological survey report text and time-series InSAR deformation data. Time-series InSAR deformation data refers to the set of deformation information organized in time series after multiple repeated observation images of the target construction area. This data set includes the geodetic coordinates of each coherent target point and its cumulative deformation at each imaging time relative to the initial reference time.

[0024] In some embodiments, identifying discrete semantic points with spatial location attributes in the target construction area based on the geological survey report text in the multi-source data is achieved through the following steps: Geological exploration report text is obtained from the multi-source data, and the borehole location coordinates associated with the qualitative description statements in the geological exploration report text are identified. Based on the borehole location coordinates, the qualitative description statement is semantically transformed to obtain discrete semantic points with spatial location attributes in the target construction area.

[0025] In practice, firstly, after obtaining the structured geological exploration report text from multi-source data, the stratigraphic segment description fields of each borehole record in the geological exploration report text are traversed. A joint information extraction method based on rule matching and sequence labeling is used to identify qualitative description statements expressing the integrity of the rock mass. The qualitative description statements refer to short text segments of standardized judgment words recorded in natural language. The qualitative description statements contain standardized judgment word text segments of "complete", "relatively complete", "relatively fragmented", "fragmented", and "extremely fragmented". At the same time, the borehole opening geodetic coordinates are extracted from the same borehole record as the borehole location coordinates associated with the qualitative description statement. Then, using the extracted borehole location coordinates as spatial anchor points, the corresponding qualitative description statements are converted into integrity rating values ​​according to a preset integrity level mapping table. In the mapping table, "complete" corresponds to an integrity rating value of 5, "relatively complete" corresponds to an integrity rating value of 4, "relatively fragmented" corresponds to an integrity rating value of 3, "fragmented" corresponds to an integrity rating value of 2, and "extremely fragmented" corresponds to an integrity rating value of 1. This yields discrete semantic points with spatial location attributes for the target construction area.

[0026] It should be noted that, in this application, discrete semantic points refer to spatial vector points with three-dimensional coordinates and assigned rock mass integrity classification values. After traversing all stratigraphic segments of all boreholes, a set of discrete semantic points with spatial location attributes is formed in the target construction area. The purpose of determining discrete semantic points is to transform the rock mass integrity information qualitatively described in natural language in the geological survey report into discrete data with three-dimensional spatial coordinates and standardized numerical labels. This allows the geological engineers' experience judgments, which originally only existed in the text report, to participate in subsequent three-dimensional spatial modeling and numerical calculations in the form of spatial vector points. This provides a basic interpolation source point for the construction of the spatial semantic field, so that the spatial transition of the deformation field can maintain the sharpness of the geological boundary at the interface where the rock mass integrity changes abruptly.

[0027] In some embodiments, constructing a spatial semantic field characterizing the rock mass integrity trend by combining all discrete semantic points with a three-dimensional geological spatial grid is achieved through the following steps: All discrete semantic points are quantitatively labeled with rock mass integrity level according to the qualitative description statements they are associated with, resulting in a set of discrete control points carrying rock mass integrity classification values. A three-dimensional geological space grid is established using the circumscribed cuboid of the region covered by the discrete control point set as the boundary. The radial basis function interpolation algorithm is used to interpolate each grid point of the three-dimensional geological space grid by using the discrete control point set as the interpolation source point, forming a spatial semantic field that is continuously distributed in three-dimensional space and represents the trend of rock mass integrity.

[0028] In specific implementation, firstly, for all generated discrete semantic points, the rock mass integrity classification value already assigned in the integrity level mapping is used as the rock mass integrity level calibration result, thereby obtaining a set of discrete control points carrying the rock mass integrity classification value. This set of discrete control points refers to a group of spatial points distributed at known borehole strata in three-dimensional space that carry numerical attributes of rock mass integrity. Then, the maximum and minimum values ​​of all discrete control points in the discrete control point set are calculated in the X, Y, and Z coordinate directions, thereby determining the minimum bounding cube to accommodate all control points. The values ​​are then calculated on the six boundary faces of this minimum bounding cube. The distance of the extended construction disturbance influence range is used to form the boundary of the modeling cuboid. Within the boundary of the modeling cuboid, the grid is divided at equal intervals along the vertical direction with half the average layer thickness of the borehole strata as the grid height and along the horizontal direction with the average spacing of InSAR coherent target points as the planar grid resolution, thus establishing a three-dimensional geological space grid. The three-dimensional geological space grid refers to the discretization of the cuboid space defined by the intersection of grid lines in the X, Y, and Z axes into a set of regularly arranged cubic units. Each intersection of the grid lines is a grid point to be assigned a value. Then, a quadratic function is selected as the radial basis function kernel. A quadratic function is a function of the form... The basis functions, where The function is a quadratic function, where r is the three-dimensional Euclidean distance from the grid point to be interpolated to the control point. The three-dimensional Euclidean distance refers to the spatial straight-line distance obtained by taking the square root of the sum of the squares of the differences between the X, Y, and Z coordinates of two points. c is a smoothing factor, and its value is the average of the three-dimensional Euclidean distances between all control points. The three-dimensional spatial coordinates and integrity classification values ​​of each control point in the discrete control point set are substituted into the N-dimensional radial basis function interpolation equation system established by the N control points. The form of the radial basis function interpolation equation system is: for the j-th control point... ,in, The weight coefficients of the i-th basis function are... Let be the three-dimensional Euclidean distance between the i-th grid point to be interpolated and the j-th control point. To determine the integrity classification value for the j-th control point, the LU decomposition method is used to solve the system of equations to obtain all weight coefficients. Then, for each grid point of the three-dimensional geological space grid, the three-dimensional Euclidean distance to each control point is calculated based on its three-dimensional coordinates and substituted into the interpolation expression. The rock mass integrity trend of the grid point is calculated, and after all grid points have been calculated, a spatial semantic field is formed that is continuously distributed in three-dimensional space.

[0029] It should be noted that, in this application, the spatial semantic field refers to the spatial continuous distribution trend of rock mass integrity in the underground three-dimensional space of the target construction area. Determining the spatial semantic field can provide geological prior constraints for subsequent spatial interpolation of InSAR deformation data. By adjusting the smoothing weight of the interpolation kernel function through the rock mass integrity trend value indicated by the spatial semantic field, the continuous deformation rate field maintains the geological boundary sharpness at the interface where rock mass integrity changes abruptly. At the same time, it provides a reference surface for the rock mass integrity trend change area for spatial superposition and comparison with the equivalent plastic strain distribution for the final risk assessment.

[0030] In step S102, the spatial semantic field is used as a priori constraint to weight and guide the spatial interpolation process of the temporal InSAR deformation data in the multi-source data, so as to obtain a continuous deformation rate field corrected by geological semantics.

[0031] In some embodiments, reference Figure 2 As shown in the figure, this is an exemplary flowchart for determining a continuous deformation rate field according to some embodiments of this application. In this embodiment, the spatial semantic field is used as a priori constraint to weight and guide the spatial interpolation process of the temporal InSAR deformation data in the multi-source data. The geologically semantically corrected continuous deformation rate field can be obtained by the following steps: In step S1021, the spatial semantic field is projected onto discrete positions that match the coherent target points in the temporal InSAR deformation data under the same spatial coordinate system, and the rock mass integrity trend value at each coherent target point is obtained. In step S1022, a kernel function weight matrix for spatial anisotropy is constructed using the rock mass integrity trend value at each coherent target point as a variable; In step S1023, the kernel function weight matrix is ​​used as the prior constraint matrix in the Kriging interpolation equation system to spatially interpolate the deformation rate of coherent target points in the time-series InSAR deformation data, thereby obtaining a continuous deformation rate field corrected by geological semantics.

[0032] In some embodiments, the spatial semantic field is projected onto discrete locations matching coherent target points in the temporal InSAR deformation data under the same spatial coordinate system, and the rock mass integrity trend value at each coherent target point is obtained by the following steps: Obtain the geodetic coordinates of each coherent target point in the temporal InSAR deformation data, and register the three-dimensional geological spatial grid where the spatial semantic field is located to the geodetic coordinate system through coordinate transformation to form a unified spatial reference. Under the unified spatial reference, the column of each coherent target point in the three-dimensional geological spatial grid is determined by the horizontal position of each coherent target point. The grid is indexed layer by layer along the vertical downward direction of the column with the grid resolution of the spatial semantic field as the step size. The interpolation result stored in the first grid point with a non-empty spatial semantic value is extracted as the rock mass integrity trend value of the corresponding coherent target point.

[0033] In practice, firstly, the geodetic coordinates of all coherent target points are extracted from the temporal InSAR deformation data. Geodetic coordinates refer to the three-dimensional spatial location expressed in latitude, longitude, and geodetic height. Simultaneously, the projection coordinate system parameters used by the three-dimensional geological spatial grid where the spatial semantic field resides are obtained. The projection coordinate system is a spatial reference system that combines plane rectangular coordinates obtained by mathematical transformations such as Gauss-Kruger projection or universal transverse Mercator projection with normal height. A seven-parameter Bursa model is used to transform the geodetic coordinates of all grid points in the three-dimensional geological spatial grid into projected coordinates, or conversely, the coordinates of coherent target points are transformed into projected coordinates consistent with the spatial semantic field. The seven-parameter Bursa model is a spatial rectangular coordinate system transformation model containing three translation parameters, three rotation parameters, and one scale factor. Ultimately, both are unified under the same projection coordinate system to form a unified spatial reference. A unified spatial reference refers to all spatial data... Based on the shared coordinate system, and then, under a unified spatial reference, the column of each coherent target point in the three-dimensional geological space grid is located using its planar coordinates. The column refers to the vertical grid line whose planar coordinates are consistent with the planar coordinates of the grid point. Starting from the grid point with the highest elevation, the grid is indexed layer by layer in the direction of decreasing elevation with the grid resolution of the spatial semantic field as the step size. The grid resolution refers to the vertical spacing between adjacent grid points. The spatial semantic value stored at each grid point is read. The spatial semantic value is the interpolation result of the rock mass integrity trend stored in the grid point. If the spatial semantic value of the current grid point is empty, the indexing continues to the next grid point. If it is not empty, the indexing stops and the interpolation result stored at the grid point is extracted as the rock mass integrity trend value at the coherent target point. When the spatial semantic values ​​of all grid points in a column are empty, the value of the nearest non-empty grid point in that column is taken as the rock mass integrity trend value of the coherent target point.

[0034] In specific implementation, firstly, the spatial semantic field is projected onto discrete locations matching coherent target points in the temporal InSAR deformation data under the same spatial coordinate system, and the rock mass integrity trend value at each coherent target point is obtained; secondly, using the obtained rock mass integrity trend value at each coherent target point as input variables, a kernel function weight matrix for spatial anisotropy is constructed. Spatial anisotropy refers to the characteristic that the smoothing intensity of the interpolation kernel function exhibits directional differences in space with changes in geological properties. The kernel function weight matrix refers to... The matrix used to scale the semivariance values ​​in the Kriging interpolation equation is generated by transforming the difference in rock mass integrity trend values ​​between coherent target point pairs with a negative exponential function as the independent variable. Its construction process is as follows: for any two coherent target points p and q, calculate the absolute value of the difference in their rock mass integrity trend values. According to the formula Calculate the weight coefficients of the point pair, where Here, k is the empirical coefficient controlling the attenuation rate, with a value of 0.5, and e is an exponential function with the natural constant e as its base. All weight coefficients are arranged in order of point pair indices to form a kernel function weight matrix. Then, this kernel function weight matrix is ​​applied to the semivariance matrix in the ordinary Kriging interpolation equation system in the form of a Hadamard product. The semivariance matrix is ​​a symmetric square matrix composed of the semivariance values ​​calculated based on the spatial distance between all coherent target point pairs. The Hadamard product is the direct element-wise multiplication of two matrices of the same order. Thus, the kernel function weight matrix is ​​embedded as a priori constraint matrix into the Kriging interpolation equation system, ensuring that values ​​within the same rock mass integrity level zone are considered. The semivariance between point pairs within the domain remains unchanged, while the semivariance between point pairs across the rock mass integrity abrupt interface is amplified. This suppresses the smoothing effect of the interpolation kernel at the integrity abrupt interface indicated by the spatial semantic location. The deformation rate of each coherent target point is spatially interpolated using the Kriging interpolation equation set after embedding the kernel function weight matrix. The deformation rate refers to the displacement of each coherent target point from the line of sight to the ground surface per unit time obtained by InSAR inversion. By solving the equation set, the optimal unbiased estimate of the deformation rate at the grid point to be interpolated is obtained. Finally, a continuous deformation rate field with geological semantic correction is generated that is continuously distributed on the two-dimensional plane and maintains the sharpness of the geological boundary.

[0035] It should be noted that in this application, the continuous deformation rate field characterizes the spatial distribution of continuous displacement of the surface of the target construction area within a unit time. The determination of the continuous deformation rate field serves as the displacement driving condition for the top boundary of the subsequent finite element numerical model. It is applied to the top surface of the three-dimensional finite element model in the form of nodal displacement constraints, so that the stress-strain response obtained by forward modeling can truly reflect the spatial distribution characteristics of surface deformation observed by InSAR. At the same time, the introduction of the spatial semantic field as a priori constraint during the interpolation process maintains the sharpness of the geological boundary of the deformation field at the interface of abrupt change in rock mass integrity, avoiding the smooth transition distortion produced by traditional interpolation methods at the geological boundary.

[0036] In step S103, the continuous deformation rate field is used as the initial boundary condition to drive the finite element numerical model to perform forward simulation of stress and strain, thereby obtaining the maximum equivalent plastic strain distribution under the current stress state.

[0037] In some embodiments, the finite element numerical model is driven to perform forward modeling of stress and strain using the continuous deformation rate field as the initial boundary condition to obtain the maximum equivalent plastic strain distribution under the current stress state. This is achieved through the following steps: A three-dimensional finite element numerical model is established based on the stratigraphic division of the target construction area, and the corresponding rock mass physical and mechanical parameters are assigned to each rock layer. The continuous deformation rate field is applied to the top boundary of the three-dimensional finite element numerical model in the form of nodal displacement constraints; The three-dimensional finite element numerical model is solved elastically and plastically under the combined action of gravity load and nodal displacement constraint. The equivalent plastic strain value of each element is extracted from the converged stress-strain field. The maximum value of the equivalent plastic strain of each element in the whole simulation time step is taken to form the largest equivalent plastic strain distribution.

[0038] In practice, firstly, the strata depth and lithological names revealed by each borehole are extracted from the structured geological exploration report. Lateral stratigraphic comparison and connection are then performed based on the principle of identical lithology and continuous strata, resulting in a stratigraphic division of the target construction area defined by several three-dimensional stratigraphic interfaces. This stratigraphic division refers to a stratification scheme that divides the underground space into several layered three-dimensional geometric bodies with uniform lithological properties. Based on this stratigraphic division, a three-dimensional finite element numerical model is established using eight-node hexahedral elements for mesh generation. This three-dimensional finite element numerical model discretizes continuous rock and soil mass into a finite number of elements interconnected by nodes. The numerical calculation model assigns elastic modulus, Poisson's ratio, cohesion, and internal friction angle extracted from the physical and mechanical parameters suggested in the geological exploration report to each rock stratum unit as the rock mass physical and mechanical parameters. Then, the deformation rate value of each grid point in the geologically semantically corrected continuous deformation rate field is multiplied by the construction simulation duration to convert it into displacement. The construction simulation duration refers to the time span from the start time of the InSAR data to the risk assessment time. The shape function interpolation method is used to distribute the grid point displacement to the four nodes closest to the grid point on the top surface of the three-dimensional finite element numerical model. The shape function interpolation method refers to using the shape function of the finite element to integrate the surface load or displacement. The standard finite element preprocessing operation, which assigns points or grid points to nodes, is applied to all top-face nodes to form nodal displacement constraints defined in the form of displacement values. Nodal displacement constraints specify boundary conditions that allow nodes to undergo known displacements in a designated direction. Finally, the three-dimensional finite element numerical model is solved elastoplastically under the combined action of gravity load and nodal displacement constraints. The elastoplastic solution refers to a nonlinear finite element calculation process that uses the Drucker-Prager yield criterion as the plastic yield criterion for rock mass and gradually loads the model using the Newton-Raphson iterative method until the convergence tolerance is met. The Drucker-Prager yield criterion is defined by the first invariant of the stress tensor and the second invariant of the deviatoric stress tensor. The yield function, expressed by invariants and applicable to geotechnical materials, is used to extract the equivalent plastic strain value of each element from the output stress-strain field after convergence. That is, for each hexahedral element, the scalar value representing the degree of cumulative plastic deformation of the element is obtained by integrating along the loading history at its integration point by taking the square root of the sum of the squares of the components of the plastic strain increment tensor in the elastoplastic solution. The equivalent plastic strain value refers to the scalar value representing the degree of cumulative plastic deformation of the element. The maximum value of the equivalent plastic strain of each element in the entire increment step history is taken, and the maximum equivalent plastic strain values ​​of all elements are arranged according to the element number to form the largest equivalent plastic strain distribution.

[0039] It should be noted that the maximum equivalent plastic strain distribution in this application refers to a three-dimensional scalar field using a finite element mesh as the carrier, where each element stores the historical maximum equivalent plastic strain value. The determination of the maximum equivalent plastic strain distribution is based on the fact that the maximum equivalent plastic strain value of each element within the entire simulation time step can capture the most dangerous plastic damage state during the entire loading process. After forming a spatial distribution of the maximum plastic damage states of all captured elements, it is spatially overlaid and compared with the rock mass integrity trend change area extracted by the spatial semantic field. When the plastic strain concentration area and the integrity change area highly coincide, it indicates that the stress accumulation driven by surface deformation is clearly related to the weak zone of the rock mass. Based on this, the environmental civil risk level of the target construction area is determined, providing a quantitative basis for construction safety decision-making.

[0040] In step S104, the maximum equivalent plastic strain distribution is superimposed on the spatial semantic field, and the risk assessment level of the target construction area is determined by comparing the spatial overlap between the maximum equivalent plastic strain concentration area and the rock mass integrity trend change area.

[0041] In some embodiments, the risk assessment level of the target construction area is determined by superimposing the maximum equivalent plastic strain distribution onto the spatial semantic field and comparing the spatial overlap between the region of maximum equivalent plastic strain concentration and the region of abrupt change in rock mass integrity trend, using the following steps: The maximum equivalent plastic strain distribution is mapped onto the grid where the spatial semantic field is located under a unified spatial reference, so as to identify grid cells whose equivalent plastic strain value exceeds the preset plastic yield threshold as plastic strain concentration areas. In the spatial semantic field, grid cells whose spatial gradient magnitude exceeds a preset integrity mutation threshold are extracted as rock mass integrity trend mutation regions. Calculate the spatial overlap between the plastic strain concentration region and the region of abrupt change in rock mass integrity trend; When the spatial overlap exceeds the preset high-risk overlap threshold, the risk assessment level of the target construction area is determined to be high-risk; otherwise, it is determined to be medium-low risk.

[0042] In practical implementation, firstly, the 3D finite element mesh containing the largest equivalent plastic strain distribution and the 3D geological spatial mesh containing the spatial semantic field are unified to the same projected coordinate system through coordinate transformation to form a unified spatial reference. A unified spatial reference means that both share the same plane coordinate system and elevation reference. Then, the maximum equivalent plastic strain value at the center point of each finite element is mapped to the spatial semantic field mesh grid point with the nearest spatial distance using the nearest neighbor interpolation method. The nearest neighbor interpolation method is an unweighted assignment method that directly assigns the value of the point to be mapped to the grid point closest to that point. After mapping, the spatial semantic field mesh is then mapped. On the spatial semantic field grid, the maximum assigned equivalent plastic strain value is checked point by point. Grid points whose maximum equivalent plastic strain value exceeds a preset plastic yield threshold are marked as plastic strain concentration regions. The preset plastic yield threshold is the critical value when the equivalent plastic strain of the rock mass reaches 0.2%. All marked grid points form plastic strain concentration regions, which refer to a set of grid points that are spatially contiguous and have undergone irreversible plastic damage to the rock mass. Secondly, the spatial gradient of the rock mass integrity trend value of each grid point is calculated using the three-dimensional Sobel gradient operator on the spatial semantic field. The three-dimensional Sobel gradient operator refers to a discrete differential operator that convolves with 3×3×3 convolution kernels in the X, Y, and Z directions, centered on the target grid point, and calculates the gradient magnitude by taking the square root of the sum of the squares. Grid points whose calculated spatial gradient magnitudes exceed a preset integrity mutation threshold are marked as grid points in the rock mass integrity trend mutation region. The preset integrity mutation threshold is the gradient value corresponding to the 85th percentile of the histogram of gradient magnitudes of all grid points in the spatial semantic field. All marked grid points form a rock mass integrity trend mutation region, which refers to the region where the rock mass integrity changes rapidly. The system identifies a set of strip-shaped or planar grid points that undergo abrupt spatial changes. Then, it calculates the number of grid points intersecting the set of grid points in the plastic strain concentration area and the set of grid points in the rock mass integrity abrupt trend abrupt change area. This number of intersecting grid points is divided by the total number of grid points in the rock mass integrity abrupt trend abrupt change area to obtain the spatial overlap. Spatial overlap refers to the proportion of the plastic strain concentration area that spatially overlaps with the rock mass integrity abrupt trend abrupt change area. Finally, when the spatial overlap is greater than a preset high-risk overlap threshold, the risk assessment level of the target construction area is determined to be high-risk; conversely, when the spatial overlap is less than or equal to 0.6, it is determined to be medium-low risk.

[0043] It should be noted that the risk assessment level in this application refers to the safety classification result of the target construction area. The purpose of determining the risk assessment level is to transform the correlation between the spatial distribution of rock mass integrity and the stress-strain response driven by deformation obtained from the fusion analysis of multi-source data into a classification conclusion that can be directly referenced for engineering decision-making. When it is determined to be high risk, it indicates that there is an unstable situation in the target construction area where surface deformation is concentrated and the weak zone of the rock mass is highly coupled. It is necessary to take control measures such as strengthening monitoring, adjusting the construction plan, or carrying out support and reinforcement. When it is determined to be medium or low risk, it indicates that the current construction disturbance has not yet caused significant stress concentration along the weak surface of rock mass integrity. The construction can continue according to the original plan and maintain the regular monitoring frequency, thereby providing a quantitative classification basis for construction safety management and risk avoidance.

[0044] Furthermore, in another aspect of this application, in some embodiments, this application provides an environmental civil risk assessment system based on multi-source data fusion, with reference to... Figure 3 The figure is a schematic diagram of the structure of an environmental civil risk assessment system based on multi-source data fusion according to some embodiments of this application. The environmental civil risk assessment system based on multi-source data fusion includes: an acquisition module 201, a processing module 202, and an execution module 203, which are described below: The acquisition module 201 in this application is mainly used to acquire multi-source data of the target construction area, identify discrete semantic points with spatial location attributes in the target construction area based on the geological survey report text in the multi-source data, and construct a spatial semantic field representing the trend of rock mass integrity by combining all discrete semantic points with a three-dimensional geological spatial grid. Processing module 202, in this application, is mainly used to use the spatial semantic field as a priori constraint to weight and guide the spatial interpolation process of the temporal InSAR deformation data in the multi-source data, so as to obtain a continuous deformation rate field corrected by geological semantics. The processing module 202 is also used to drive the finite element numerical model to perform forward modeling of stress and strain using the continuous deformation rate field as the initial boundary condition, so as to obtain the maximum equivalent plastic strain distribution under the current stress state. The execution module 203 in this application is mainly used to superimpose the maximum equivalent plastic strain distribution onto the spatial semantic field, and determine the risk assessment level of the target construction area by comparing the spatial overlap between the equivalent plastic strain concentration area and the rock mass integrity trend change area.

[0045] In addition, this application also provides a computer device, the computer device including a memory and a processor, the memory storing code, the processor being configured to acquire the code and execute the above-described environmental civil risk assessment method based on multi-source data fusion.

[0046] In some embodiments, reference Figure 4 The figure is a schematic diagram of the structure of a computer device implementing an environmental civil risk assessment method based on multi-source data fusion, according to some embodiments of this application. The environmental civil risk assessment method based on multi-source data fusion in the above embodiments can... Figure 4 The computer device shown is used to implement this, and the computer device includes at least one processor 301, a communication bus 302, a memory 303, and at least one communication interface 304.

[0047] The processor 301 may be a general-purpose central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more devices used to control the execution of the environmental civil risk assessment method based on multi-source data fusion in this application.

[0048] The communication bus 302 can be used to transmit information between the aforementioned components.

[0049] The memory 303 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disks or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory 303 may exist independently and be connected to the processor 301 via the communication bus 302. The memory 303 may also be integrated with the processor 301.

[0050] The memory 303 stores program code for executing the scheme of this application, and its execution is controlled by the processor 301. The processor 301 executes the program code stored in the memory 303. The program code may include one or more software modules. In the above embodiments, the determination of the environmental civil risk assessment method based on multi-source data fusion can be implemented by the processor 301 and one or more software modules in the program code in the memory 303.

[0051] Communication interface 304 uses any transceiver-like device for communicating with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0052] In a specific implementation, as one example, a computer device may include multiple processors, each of which may be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. Here, a processor may refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0053] The aforementioned computer device can be a general-purpose computer device or a special-purpose computer device. In specific implementations, the computer device can be a desktop computer, a portable computer, a network server, a handheld digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. This application does not limit the type of computer device.

[0054] In addition, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described environmental and civil risk assessment method based on multi-source data fusion.

[0055] Although preferred embodiments of this application have been described, those skilled in the art, once they have learned the basic inventive concept, can make other changes and modifications to these embodiments.

[0056] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application.

Claims

1. An environmental civil risk assessment method based on multi-source data fusion, characterized in that, Includes the following steps: Multi-source data of the target construction area is acquired, including geological survey report text and time-series InSAR deformation data. Based on the geological survey report text in the multi-source data, discrete semantic points with spatial location attributes in the target construction area are identified. All discrete semantic points are quantitatively calibrated according to their associated qualitative description statements to obtain a set of discrete control points carrying rock mass integrity classification values. A three-dimensional geological spatial grid is established with the circumscribed cuboid of the area covered by the discrete control point set as the boundary. A radial basis function interpolation algorithm is used to interpolate each grid point of the three-dimensional geological spatial grid, using the discrete control point set as the interpolation source point, to form a spatial semantic field that is continuously distributed in three-dimensional space and represents the trend of rock mass integrity. The spatial semantic field is projected onto discrete locations that match coherent target points in the temporal InSAR deformation data in the same spatial coordinate system, and the rock mass integrity trend value at each coherent target point is obtained. A kernel function weight matrix for spatial anisotropy is constructed using the rock mass integrity trend value at each coherent target point as a variable; the kernel function weight matrix is ​​used as the prior constraint matrix in the Kriging interpolation equation set to spatially interpolate the deformation rate of the coherent target points in the time-series InSAR deformation data to obtain a continuous deformation rate field corrected by geological semantics. Using the continuous deformation rate field as the initial boundary condition, the finite element numerical model is driven to perform forward modeling of stress and strain to obtain the maximum equivalent plastic strain distribution under the current stress state. The maximum equivalent plastic strain distribution is superimposed onto the spatial semantic field, and the risk assessment level of the target construction area is determined by comparing the spatial overlap between the equivalent plastic strain concentration area and the rock mass integrity trend abrupt change area.

2. The method as described in claim 1, characterized in that, Obtaining multi-source data for the target construction area specifically includes: Based on the boundary of the target construction area, extract the geological survey report text corresponding to the boundary from the digital archive of the regional geological data center; The single-view complex image sequence covering the boundary of the range is queried from the synthetic aperture radar satellite archive data. Differential interferometry is performed on the single-view complex image sequence to generate temporal InSAR deformation data containing temporal cumulative deformation and deformation rate information. Multi-source data of the target construction area are obtained by combining the geological survey report text and the time-series InSAR deformation data.

3. The method as described in claim 1, characterized in that, Based on the geological survey report text in the multi-source data, the discrete semantic points with spatial location attributes of the target construction area are identified, specifically including: Geological exploration report text is obtained from the multi-source data, and the borehole location coordinates associated with the qualitative description statements in the geological exploration report text are identified. Based on the borehole location coordinates, the qualitative description statement is semantically transformed to obtain discrete semantic points with spatial location attributes in the target construction area.

4. The method as described in claim 1, characterized in that, Projecting the spatial semantic field onto discrete locations matching coherent target points in the temporal InSAR deformation data within the same spatial coordinate system, and obtaining the rock mass integrity trend value at each coherent target point specifically includes: Obtain the geodetic coordinates of each coherent target point in the temporal InSAR deformation data, and register the three-dimensional geological spatial grid where the spatial semantic field is located to the geodetic coordinate system through coordinate transformation to form a unified spatial reference. Under the unified spatial reference, the column of each coherent target point in the three-dimensional geological spatial grid is determined by the horizontal position of each coherent target point. The grid is indexed layer by layer along the vertical downward direction of the column with the grid resolution of the spatial semantic field as the step size. The interpolation result stored in the first grid point with a non-empty spatial semantic value is extracted as the rock mass integrity trend value of the corresponding coherent target point.

5. An environmental civil engineering risk assessment system based on multi-source data fusion, used to execute the environmental civil engineering risk assessment method based on multi-source data fusion as described in any one of claims 1 to 4, characterized in that, The system includes: The acquisition module is used to acquire multi-source data of the target construction area, identify discrete semantic points with spatial location attributes in the target construction area based on the geological survey report text in the multi-source data, and construct a spatial semantic field representing the trend of rock mass integrity by combining all discrete semantic points with a three-dimensional geological spatial grid. The processing module is used to use the spatial semantic field as a priori constraint to weight and guide the spatial interpolation process of the temporal InSAR deformation data in the multi-source data, so as to obtain a continuous deformation rate field corrected by geological semantics. The processing module is also used to drive the finite element numerical model to perform forward modeling of stress and strain using the continuous deformation rate field as the initial boundary condition, so as to obtain the maximum equivalent plastic strain distribution under the current stress state. The execution module is used to superimpose the maximum equivalent plastic strain distribution onto the spatial semantic field, and determine the risk assessment level of the target construction area by comparing the spatial overlap between the equivalent plastic strain concentration area and the rock mass integrity trend change area.

6. A computer device, characterized in that, The computer device includes a memory and a processor, the memory storing code, and the processor being configured to retrieve the code and execute the environmental civil risk assessment method based on multi-source data fusion as described in any one of claims 1 to 4.

7. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the environmental civil risk assessment method based on multi-source data fusion as described in any one of claims 1 to 4.

Images