Seismic activity fault model construction method, system, device and medium based on spatial data mining and geological constraints

Abstract

The application discloses a kind of based on spatial data mining and geological constraint seismic activity fault model construction method, system, equipment and medium.The method is first arranged and analyzed to seismic data, extracts minimum complete subdirectory;Using improved mean shift algorithm for spatial clustering, identify each active fault corresponding small earthquake cluster;Subsequently, each fault cluster is three-dimensional slice, and least square method fitting is generated fault interpretation line;With interpretation line as foundation, initial seismic activity fault three-dimensional model is constructed, and is optimized by multi-source geological constraint, and active fault three-dimensional fine model is obtained;Finally, the model is integrated with digital earth spatial registration.This application can realize the whole process automation and quantization from seismic data analysis to active fault three-dimensional fine model construction and application, improve the objectivity, efficiency, precision and reliability of model construction, provide important technical support for seismic disaster risk assessment, active fault detection.

Description

Technical Field

[0001] This invention relates to the field of 3D modeling technology, and in particular to a method, system, equipment, and medium for constructing seismic active fault models based on spatial data mining and geological constraints. Background Technology

[0002] Active faults are the core geological structures that trigger earthquake disasters. Accurately identifying active faults and constructing their three-dimensional geometric models is of great practical significance for geological disaster prevention and control. However, current fault model construction methods still have many technical shortcomings and cannot meet the needs of geological disaster prevention and control for high-precision, automated, and highly reliable fault models.

[0003] Currently, existing seismic fault interpretation techniques are mainly divided into two categories: traditional manual interpretation methods and semi-automated methods. Traditional manual interpretation methods suffer from the drawbacks of high subjectivity, extremely low efficiency, difficulty in processing massive amounts of seismic data, and susceptibility to human error leading to insufficient fault identification accuracy and affecting the reliability of subsequent models. While semi-automated methods reduce human intervention to some extent, they still have several insurmountable technical limitations. First, the clustering algorithms used, such as K-clustering and density clustering, require manual pre-specification of the number of clusters or key parameters. However, the spatial distribution of seismic events is greatly influenced by regional tectonic backgrounds, exhibiting non-uniform and complex clustering characteristics. Fixed preset parameters cannot adapt to the seismic activity distribution patterns in different regions, easily leading to over-segmentation or under-segmentation of seismic clusters, resulting in inaccurate fault identification. Second, fault interpretation lines generated using linear fitting and simple interpolation methods have poor continuity and low consistency with actual seismic activity distribution, failing to truly reflect the spatial distribution characteristics of faults.

[0004] In summary, existing methods for constructing active fault models for earthquakes suffer from drawbacks such as strong subjectivity in manual interpretation, low interpretation efficiency, and poor adaptability of clustering algorithms, resulting in low accuracy in active fault identification and poor precision in fault models. Summary of the Invention

[0005] The purpose of this invention is to provide a method, system, device, and medium for constructing seismic active fault models based on spatial data mining and geological constraints, which can improve the objectivity, efficiency, accuracy, and reliability of seismic active fault model construction.

[0006] To achieve the above objectives, this invention provides a method for constructing a seismic active fault model based on spatial data mining and geological constraints, comprising: Seismic data from the seismic zone are processed and analyzed to extract the minimum complete subdirectory; the spatial nearest neighbor index algorithm is used to determine whether the minimum complete subdirectory has fault three-dimensional structure cluster characteristics, and the minimum complete subdirectory with fault three-dimensional structure cluster characteristics is obtained. An improved mean-shift algorithm is used to cluster the minimal complete subdirectory with fault three-dimensional structure cluster characteristics to obtain multiple earthquake clusters; the multiple earthquake clusters are identified to obtain the small earthquake clusters corresponding to each active fault; Three-dimensional automatic slicing is performed on the small earthquake clusters corresponding to each active fault, and the fault interpretation line is fitted using the least squares method. Based on the fault interpretation line, an initial three-dimensional model of the active earthquake fault is constructed. The initial active seismic fault model was supplemented and corrected based on geological constraints to obtain a fine three-dimensional model of the active seismic faults in the seismic zone.

[0007] Optionally, the improved mean-shift algorithm is used to cluster the minimal complete subdirectory with fault 3D structural cluster characteristics to obtain multiple seismic clusters, including: A smooth estimation is performed on the minimum complete subdirectory that possesses the characteristics of a fault three-dimensional structural cluster. An improved mean-drift algorithm is used to cluster the minimal complete subdirectories with fault 3D structural cluster characteristics after smooth estimation. The cluster center points are iteratively updated according to the mean-drift vector until the preset convergence condition is met, resulting in multiple seismic clusters.

[0008] Optionally, the identification of the multiple earthquake families to obtain the cluster of small earthquakes corresponding to each active fault includes: Calculate the spatial distance between earthquake clusters and select earthquake clusters whose spatial distance is lower than a preset geological scale threshold as candidate earthquake clusters of the same active fault. Based on surface geological evidence, the candidate earthquake clusters are assessed to identify active faults and the corresponding small earthquake clusters.

[0009] Optionally, the step of automatically slicing the small earthquake clusters corresponding to each active fault in three dimensions and fitting the fault interpretation lines using the least squares method, and constructing an initial three-dimensional model of the active earthquake fault based on the fault interpretation lines, includes: A robust local weighted regression algorithm is used to fit the three-dimensional spatial point set of the small earthquake clusters corresponding to the active fault, and a trend line of the small earthquake clusters reflecting the extension direction of the small earthquake clusters is obtained. The positions of multiple three-dimensional slices are determined along the small earthquake cluster trend line, and the fault interpretation lines on each three-dimensional slice are obtained by fitting with the least squares method. Based on surface rupture data and fault interpretation lines fitted on all three-dimensional slices, a discrete smoothing interpolation algorithm is used to spatially continuum the discrete fault interpretation lines, thereby constructing an initial seismic active fault model.

[0010] Optionally, the step of supplementing and correcting the initial seismic active fault model according to geological constraints to obtain a refined three-dimensional model of the seismic active faults in the seismic zone includes: Determine the activity nature of each active fault based on geological constraints; For each branch active fault, fault interpretation lines and three-dimensional seismic active fault models are constructed to determine the hierarchical relationship between the branch active fault and the main active fault structure. Based on the activity characteristics and the hierarchical correlation, the fault structure of the initial seismic active fault model is supplemented and corrected to obtain a refined seismic active fault model of the seismic zone.

[0011] Optionally, after supplementing and correcting the initial active seismic fault model according to geological constraints to obtain a detailed three-dimensional model of the active seismic fault in the seismic zone, the method further includes: The detailed model of active faults in the earthquake zone was spatially registered and fused with the Digital Earth Engine.

[0012] To achieve the above objectives, the present invention also provides a system for constructing seismic active fault models based on spatial data mining and geological constraints, comprising: The data processing module is used to organize and analyze seismic data in the seismic zone to extract the minimum complete subdirectory; the spatial nearest neighbor index algorithm is used to determine whether the minimum complete subdirectory has fault three-dimensional structure cluster characteristics, and the minimum complete subdirectory with fault three-dimensional structure cluster characteristics is obtained. The clustering module is used to cluster the minimal complete subdirectory with fault three-dimensional structure cluster features using an improved mean-shift algorithm to obtain multiple earthquake clusters; and to identify the multiple earthquake clusters to obtain the small earthquake clusters corresponding to each active fault. The initial fault model construction module is used to automatically slice the small earthquake clusters corresponding to each active fault in three dimensions and fit the fault interpretation line using the least squares method, and construct the initial three-dimensional model of the active earthquake fault based on the fault interpretation line. The fault model correction module is used to supplement and correct the initial seismic active fault model according to geological constraints, so as to obtain a three-dimensional fine model of the seismic active fault in the seismic zone.

[0013] Optionally, the initial fault model construction module is used for: A robust local weighted regression algorithm is used to fit the three-dimensional spatial point set of the small earthquake clusters corresponding to the active fault, and a trend line of the small earthquake clusters reflecting the extension direction of the small earthquake clusters is obtained. The positions of multiple three-dimensional slices are determined along the small earthquake cluster trend line, and the fault interpretation lines on each three-dimensional slice are obtained by fitting with the least squares method. Based on surface rupture data and fault interpretation lines fitted on all three-dimensional slices, a discrete smoothing interpolation algorithm is used to spatially continuum the discrete fault interpretation lines, thereby constructing an initial seismic active fault model.

[0014] To achieve the above objectives, the present invention also provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the seismic active fault model construction method based on spatial data mining and geological constraints as described above.

[0015] To achieve the above objectives, the present invention also provides a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the seismic active fault model construction method based on spatial data mining and geological constraints as described above.

[0016] Compared with existing technologies, this invention provides a method, system, device, and medium for constructing seismic active fault models based on spatial data mining and geological constraints. First, it extracts the minimum complete sub-directory of seismic data and uses the spatial nearest neighbor index algorithm to accurately select the minimum complete sub-directory with the characteristics of three-dimensional fault structure clusters. This can replace traditional manual interpretation, reducing subjective errors caused by human judgment from the source, while significantly improving the processing efficiency of massive seismic data. Second, it uses a mean-shift algorithm for seismic event clustering without pre-specifying the number of clusters and key parameters, effectively avoiding over-segmentation or omission of seismic clusters and accurately identifying the small earthquake clusters corresponding to each active fault. Then, by automatically slicing the small earthquake clusters in three dimensions and fitting the fault interpretation lines using the least squares method, the continuity of the fault interpretation lines and their fit with the actual distribution of seismic activity can be improved. The initial three-dimensional model constructed in this way better reflects the true spatial distribution characteristics of the fault. Finally, the initial model is supplemented and corrected based on geological constraints to further optimize the model structure and accuracy. This invention enables full automation and quantification of the entire process from seismic data analysis to the construction and application of a detailed three-dimensional model of active faults, improving the objectivity, efficiency, accuracy and reliability of model construction, and providing important technical support for earthquake disaster risk assessment and active fault detection. Attached Figure Description

[0017] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a method for constructing an active seismic fault model based on spatial data mining and geological constraints, provided by an embodiment of the present invention. Figure 2 This is a schematic diagram of earthquake event clustering results provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the seismic cluster slice division and trend line fitting results provided in an embodiment of the present invention; Figure 4 This is a structural block diagram of a seismic active fault model construction system based on spatial data mining and geological constraints provided in an embodiment of the present invention; Figure 5 This is a structural block diagram of a terminal device provided in an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] See Figure 1 , Figure 1 This is a flowchart illustrating a method for constructing a seismic active fault model based on spatial data mining and geological constraints, as provided in an embodiment of the present invention. Figure 1 As shown, the method for constructing a seismic active fault model based on spatial data mining and geological constraints includes steps S1 to S4: Step S1: Organize and analyze the seismic data of the seismic zone to extract the minimum complete subdirectory; use the spatial nearest neighbor index algorithm to determine whether the minimum complete subdirectory has the characteristics of fault three-dimensional structure cluster, and obtain the minimum complete subdirectory with the characteristics of fault three-dimensional structure cluster. For example, data such as earthquake catalogs (including longitude, latitude, depth, magnitude, and time of occurrence) of earthquake regions, surface rupture data, focal mechanism interpretation libraries, digital elevation models (DEMs), and satellite imagery are collected.

[0021] Due to the diverse sources, formats, and spatiotemporal references of the data, preprocessing is performed on the multi-source data. For example, the collected data is sequentially cleaned, deduplicated, and outlier removed, and a unified spatiotemporal reference is established to ensure that different data can be used interchangeably in terms of spatial location and temporal scale. Based on this, the minimum complete magnitude in the earthquake catalog is analyzed, and the minimum complete sub-catalog is extracted. This preprocessing effectively removes invalid data and interfering information, ensuring the standardization, completeness, and consistency of the data used in subsequent analyses, thereby improving the accuracy of earthquake activity analysis.

[0022] To more accurately identify seismic events closely related to fault activity and avoid interference from randomly or uniformly distributed seismic events in subsequent clustering and modeling results, a spatial nearest neighbor index algorithm is further employed. This algorithm quantitatively analyzes the minimum complete sub-catalogue in three-dimensional space to determine whether the seismic events within it possess the characteristics of a three-dimensional fault structure cluster. For example, the spatial nearest neighbor index is calculated using the following formula: ; in, This represents the actual average nearest neighbor distance for all seismic events extracted from the minimum complete subdirectory, which is the average distance between each seismic event and its spatial nearest neighbor. It represents the theoretical average nearest neighbor distance when the same number of points are completely randomly distributed (Poisson distribution) in three-dimensional space, serving as a benchmark reference for spatial distribution patterns.

[0023] Based on calculations Values ​​can be used to determine the spatial distribution patterns of earthquake events: when When the actual average nearest neighbor distance of an earthquake event is less than the theoretical random distribution distance, the earthquake activity is clustered, which is consistent with the geological characteristics of fault activity clustering along the fault zone, and corresponds to effective earthquake data with the characteristics of fault three-dimensional structure clustering. when When the actual average nearest neighbor distance is consistent with the theoretical random distribution distance, the seismic activity is randomly distributed, which conforms to the Poisson distribution assumption and has no obvious fault-related clustering characteristics. when When the actual average nearest neighbor distance is greater than the theoretical random distribution distance, the seismic activity is uniformly distributed and there is a spatial competition effect. It also does not have the characteristics of a three-dimensional fault structure cluster.

[0024] It is worth noting that fault-related seismic activity typically exhibits a clustered distribution along fault zones, and the spatial nearest neighbor index algorithm can objectively quantify the spatial distribution patterns of seismic events, effectively distinguishing between clustered, random, and uniform distributions. Based on these findings, a minimal complete subdirectory with fault structure clustering characteristics is compiled. This step pre-filters noisy data, improving the efficiency and accuracy of subsequent seismic event clustering, fault interpretation, and model construction.

[0025] Step S2: The improved mean-shift algorithm is used to cluster the minimal complete subdirectory with fault three-dimensional structure cluster characteristics to obtain multiple earthquake clusters; the multiple earthquake clusters are identified to obtain the small earthquake clusters corresponding to each active fault. In one optional embodiment, the improved mean-shift algorithm is used to cluster the minimal complete subdirectory with fault 3D structural cluster characteristics to obtain multiple seismic clusters, including steps a1 to a2: a1. Perform a smooth estimation on the minimum complete subdirectory that has the characteristics of fault three-dimensional structure cluster; a2. An improved mean drift algorithm is used to cluster the minimal complete subdirectories with fault 3D structure cluster characteristics after smooth estimation. The cluster center points are iteratively updated according to the mean drift vector until the preset convergence condition is met, resulting in multiple seismic clusters.

[0026] For example, based on the minimal complete subdirectory with fault 3D structural cluster characteristics obtained in step S1, a mean shift algorithm (such as Mean Shift clustering) is used to perform spatial density clustering of seismic events. This algorithm is a non-parametric density estimation method that does not require pre-specifying the number of clusters. It allows the center point to "drift" along the direction of increasing data density, adaptively identifying high-density spatial regions and thus objectively dividing multiple seismic clusters. During the clustering process, Euclidean spatial distance in the projected coordinate system is used as the metric, and the clustering scale is controlled by a bandwidth parameter. In the post-processing stage, noisy clusters with fewer than a threshold number of events (e.g., 1-80) are removed, ultimately yielding a set of seismic clusters related to potential fault activity.

[0027] Specifically, the clustering process is calculated as follows: First, a Gaussian kernel function is used to smooth the seismic sub-catalog data. The kernel density estimate is defined as follows: ; Where h is the bandwidth, which controls the width of the kernel function. The larger the bandwidth, the narrower the spatial distribution of seismic events; d is the spatial dimension of the data, which is three-dimensional space in this embodiment, i.e., d=3; n is the total number of seismic event samples. The Gaussian kernel function is expressed as: .

[0028] Secondly, for a given point Calculate the corresponding mean shift vector using the following formula: ; in, , where is the derivative of the Gaussian kernel function; and These are the feature vectors of the current iteration center point and the input seismic event sample, respectively; the direction of the mean shift vector points towards the local maximum of the data density, providing a basis for the iterative update of the center point.

[0029] Finally, by continuously centering Move in the direction of the Mean Shift vector until the convergence condition is met (vector length is less than a preset threshold). The iterative formula is: ; Convergence condition: , This is a preset convergence threshold used to control the iteration accuracy; Ultimately, after convergence This corresponds to the center of the earthquake cluster, completing the spatial density clustering of earthquake events.

[0030] It should be noted that the kernel density estimation mean drift vector calculation provides the core density distribution basis: then, the discrete seismic events are smoothly fitted by the Gaussian kernel function to obtain a continuous spatial density function, thereby quantifying the degree of seismic activity clustering in three-dimensional space; on this basis, based on the gradient direction of the kernel density function, the mean drift vector pointing to the local density maxima is calculated, providing directional guidance for the subsequent iterative convergence of the center point. The two work together to achieve adaptive density clustering of seismic events.

[0031] In one optional embodiment, the identification of the plurality of earthquake families to obtain the small earthquake clusters corresponding to each active fault includes steps b1 to b2: b1. Calculate the spatial distance between earthquake clusters and select earthquake clusters whose spatial distance is lower than the preset geological scale threshold as candidate earthquake clusters of the same active fault. b2. Based on surface geological evidence, the candidate earthquake clusters are judged to identify active faults and the corresponding small earthquake clusters.

[0032] For example, after obtaining multiple seismic clusters using the mean-shift algorithm, due to the complexity of seismic activity, some adjacent seismic clusters may correspond to different segments of the same active fault, or exhibit spatial separation due to differences in local tectonic stress. If fault modeling is performed directly based on the initial clustering results, problems such as overly detailed fault division and confusion in the attribution of main faults and branch faults may easily occur. Therefore, multi-criteria similarity analysis is further used to merge seismic clusters and determine fault attribution, in order to achieve the identification and division of active faults.

[0033] Specifically, spatial proximity analysis is first performed. By calculating the spatial distance between the centroids of earthquake clusters, or the minimum distance between earthquake events within a cluster, the spatial correlation between different earthquake clusters is quantified. If the calculated distance is lower than a preset geological scale threshold (exemplarily, the threshold can be set to 5–10 km, which can be adaptively adjusted according to the tectonic background and fault development scale of the earthquake region), the two earthquake clusters are identified as candidate clusters potentially belonging to the same fault. Then, based on the spatial proximity analysis, multi-source surface geological evidence is further introduced for constraint verification and adjustment to avoid misjudgment caused by simple spatial distance determination. Combining known surface rupture traces, focal mechanism solutions of related earthquakes, coseismic deformation fields, active fault geological maps, and linear geomorphic features such as fault scarps and linear valleys, candidate faults are verified one by one: if the spatial distribution of the candidate clusters highly matches the tectonic features of the surface fault traces, fault attitude reflected by the focal mechanism solutions, and coseismic deformation fields, they are confirmed to belong to the same active fault; if there are obvious geological feature conflicts, the cluster classification is adjusted, and candidate associations that do not conform to tectonic laws are eliminated. Through the constraints of multiple sources and multiple criteria of spatial dimensions and geological evidence, the total number of active faults in the study area was finally determined, and a corresponding earthquake cluster (i.e., a cluster of small earthquakes on the fault) was assigned to each active fault.

[0034] See Figure 2 , Figure 2 This is a schematic diagram of earthquake event clustering results provided in an embodiment of the present invention. Figure 2 As shown, the mean-shift algorithm was used to perform spatial density clustering on the minimum complete subdirectory with fault 3D structural cluster characteristics, identifying three groups of effective seismic clusters of different sizes (red, green, and orange points), corresponding to the main active faults and secondary faults in the seismic zone. At the same time, spatially scattered and few noise points (gray points) were automatically removed, thus effectively filtering out non-tectonic seismic interference data and providing high-quality seismic cluster data for subsequent fault interpretation and model construction.

[0035] Step S3: Perform three-dimensional automatic slicing of the small earthquake clusters corresponding to each active fault and fit the fault interpretation line using the least squares method. Construct an initial three-dimensional model of the active earthquake fault based on the fault interpretation line. It should be noted that in existing 3D fault modeling processes, linear fitting and simple interpolation methods are often used to generate fault interpretation lines. The interpretation lines generated by these methods have poor continuity and low consistency with the spatial distribution of actual seismic activity. Furthermore, the modeling process lacks constraints from multi-source geological data and fails to effectively integrate key geological information such as surface fault traces, focal mechanism solutions, and surface rupture zones. As a result, the constructed fault models do not match the actual geological structures, have extremely low reliability and verifiability, and cannot truly reflect the activity characteristics and spatial distribution of faults. Consequently, they are unable to support the precise design of geological disaster prevention and control projects and the scientific assessment of potential risks.

[0036] Therefore, in an optional embodiment, step S3 includes steps S301 to S303: Step S301: Use a robust local weighted regression algorithm to fit the three-dimensional spatial point set of the small earthquake cluster corresponding to the active fault, and obtain the small earthquake cluster trend line reflecting the extension direction of the small earthquake cluster; Step S302: Determine the positions of multiple three-dimensional slices along the small earthquake cluster trend line, and use the least squares method to fit and obtain the fault interpretation lines on each three-dimensional slice; Step S303: Based on the surface rupture data and the fault interpretation lines fitted on all three-dimensional slices, the discrete fault interpretation lines are spatially continuous using a discrete smoothing interpolation algorithm to construct an initial seismic active fault model.

[0037] For example, the specific process is as follows: (1) Based on the constructed small earthquake clusters of each fault, the trend line of the small earthquake cluster is fitted by the robust local weighted regression algorithm, and the trend line is used as the main axis of the three-dimensional slice. (2) Calculate the length of the small earthquake cluster trend line, starting from the initial step size and increasing it incrementally, and find the maximum step size that satisfies the minimum overlap ratio as the calculation step size; (3) Based on the obtained step size and trend line, calculate the center of the three-dimensional slice and the slice direction T and normal direction N of each center, and normalize them (N= cross( T, [0, 0, 1] )). Each slice is nearly perpendicular to the trend of the small earthquake cluster. (4) Based on the obtained slice direction and normal direction, the query range area of ​​the slice is obtained based on the minimum overlap design. Based on the boundary of the range area, the point index of each slice is calculated. (5) Project the small earthquake vertically onto the plane of each slice, and use the least squares method to fit the fault interpretation line on each slice; (6) Integrate the obtained multi-segment fault interpretation lines and surface rupture data, and use the discrete smooth interpolation algorithm to construct the three-dimensional fault initial model. The core of this algorithm is to discretize the fault interpretation lines into a series of nodes, and solve a constrained least squares optimization problem to simultaneously satisfy the hard constraints of control points and the soft constraints of smoothness, thereby realizing the continuity and smooth optimization of the fault interpretation lines.

[0038] For example, suppose the fault trajectory is discretized as Nodes, Nodes The coordinates are Its coordinates are The core formula of the algorithm is: ; in, It is the first A roughness criterion, typically expressed in a third-order discrete difference form (similar to discrete thin-plate spline energy), is as follows: ; in, These are the control point weighting coefficients, used to balance smoothness and fitting accuracy; It is a set of control points; The location of the known control points (which may be from the initial fault point or the centroid of the earthquake cluster); This represents the number of roughness measurement items.

[0039] The least squares optimization problem is ultimately transformed into a large system of sparse linear equations: ; in, It is the roughness matrix (symmetric positive definite sparse matrix). For the control point constraint matrix, This is the right-hand term.

[0040] Solving this system of equations yields the optimal locations of all nodes, enabling the continuity and smoothing optimization of the initial fault interpretation lines, and constructing an initial seismic active fault model.

[0041] See Figure 3 , Figure 3 This is a schematic diagram illustrating the seismic cluster slice division and trend line fitting results provided in an embodiment of the present invention. Figure 3As shown, for the seismic clusters corresponding to the main active fault, a trend line (purple line) for the small earthquake clusters is obtained by fitting the robust local weighted regression algorithm. Multiple three-dimensional slices (black vertical lines) perpendicular to the fault strike are adaptively divided along the trend line. The red dots are the centers of each slice. The colored dots in the figure are the effective seismic clusters corresponding to the fault, and the gray dots are the noise data that has been removed. This shows the processing process from discrete seismic clusters to structured slice data, which can provide a spatial reference for subsequent fault interpretation line fitting and three-dimensional model construction.

[0042] Step S4: Supplement and correct the initial seismic active fault model according to geological constraints to obtain a three-dimensional fine model of the seismic active fault in the seismic zone.

[0043] In one alternative embodiment, the geological constraints include at least one of surface fault traces, focal mechanism solutions, and surface rupture zones.

[0044] In an optional embodiment, step S4 includes steps S401 to S403: Step S401: Determine the activity nature of each active fault based on geological constraints; Step S402: Construct fault interpretation lines and three-dimensional seismic active fault models for each branch active fault to determine the hierarchical relationship between the branch active fault and the main active fault structure; Step S403: Based on the activity properties and the hierarchical association, supplement and correct the fault structure of the initial seismic active fault model to obtain a refined seismic active fault model of the seismic zone.

[0045] For example, based on geological constraints such as surface fault traces, focal mechanism solutions, and surface rupture zones in the seismic zone, the activity nature of each fault is determined, distinguishing between exposed surface faults and concealed active faults, and thus determining the top boundary of the fault in the initial seismic active fault model. For complex fault zones containing multiple branch faults, branch fault interpretation lines are independently generated for each branch fault, and corresponding three-dimensional seismic active fault models are constructed. For fault zones containing multiple branch faults, the cascading relationships between branch faults are analyzed by combining seismic activity data and geological information such as surface ruptures, including spatial combinations such as intersection, conjugate, and parallelism. Spatial interpolation methods are then used to further construct the complete three-dimensional fault structure of each branch fault, ensuring that branch faults are spatially matched and connected with the main fault. Finally, the three-dimensional geometric and attitude attributes of each fault are calculated, including geometric parameters such as fault length, cutting depth, and distribution scale, as well as attitude information such as strike, dip, and dip angle. Based on the above parameters, the main fault and branch fault structures are integrated to finally construct a refined seismic active fault model.

[0046] In an optional embodiment, after step S4, the method further includes: The detailed model of active faults in the earthquake zone was spatially registered and fused with the Digital Earth Engine.

[0047] For example, the seismic fault model is first standardized and sliced ​​to eliminate redundant data and optimize the model structure. The processed model is then output in a standardized 3dtile format for efficient transmission. After standardization, the detailed 3D fault model in 3dtile format can be loaded using the WebGL framework, enabling rapid loading and smooth rendering. Simultaneously, core interactive functions such as rotation, scaling, and transparency adjustment are provided, allowing users to intuitively browse the 3D spatial morphology and spatial distribution characteristics of the fault, facilitating a quick understanding of its overall distribution and detailed features. Based on the successful rendering of the seismic fault model, multi-source related data, including fault surface data, seismic event data, surface topography data, and satellite imagery data, are further integrated. Registration technology is used to achieve the overlay, fusion, and synchronous display of data from various dimensions.

[0048] In summary, the seismic fault model construction method based on spatial data mining and geological constraints provided by this invention firstly extracts the minimum complete sub-directory of seismic data and accurately selects the minimum complete sub-directory with the characteristics of three-dimensional fault structure clusters by combining the spatial nearest neighbor index algorithm. This can replace traditional manual interpretation, reducing subjective errors caused by human judgment from the source, and significantly improving the processing efficiency of massive seismic data. Secondly, the mean drift algorithm is used to cluster seismic events without pre-specifying the number of clusters and key parameters, which can effectively avoid the problem of over-segmentation or omission of seismic clusters and accurately identify the small earthquake clusters corresponding to each active fault. Then, by automatically slicing the small earthquake clusters in three dimensions and fitting the fault interpretation lines using the least squares method, the continuity of the fault interpretation lines and their fit with the actual seismic activity distribution can be improved. The initial three-dimensional model constructed in this way is more in line with the real spatial distribution characteristics of the fault. Finally, the initial model is supplemented and corrected by combining geological constraints to further optimize the model structure and accuracy. This invention enables full automation and quantification of the entire process from seismic data analysis to the construction and application of a detailed three-dimensional model of active faults, improving the objectivity, efficiency, accuracy and reliability of model construction, and providing important technical support for earthquake disaster risk assessment and active fault detection.

[0049] Based on the above method items, the present invention provides corresponding system items embodiments.

[0050] See Figure 4 , Figure 4This is a structural block diagram of a seismic active fault model construction system based on spatial data mining and geological constraints, provided in an embodiment of the present invention. The seismic active fault model construction system based on spatial data mining and geological constraints includes: The data processing module 21 is used to organize and analyze the seismic data of the seismic zone to extract the minimum complete subdirectory; the spatial nearest neighbor index algorithm is used to determine whether the minimum complete subdirectory has the characteristics of fault three-dimensional structure cluster, and the minimum complete subdirectory with the characteristics of fault three-dimensional structure cluster is obtained. Clustering module 22 is used to cluster the minimal complete subdirectory with fault three-dimensional structure cluster features using an improved mean drift algorithm to obtain multiple earthquake clusters; and to identify the multiple earthquake clusters to obtain the small earthquake clusters corresponding to each active fault. The initial fault model construction module 23 is used to perform three-dimensional automatic slicing of the small earthquake clusters corresponding to each active fault and to fit the fault interpretation line using the least squares method, and to construct an initial three-dimensional model of the active earthquake fault based on the fault interpretation line. The fault model correction module 24 is used to supplement and correct the initial seismic active fault model according to geological constraints, so as to obtain a three-dimensional fine model of the seismic active fault in the seismic zone.

[0051] In one alternative embodiment, the clustering module 22 includes clustering units for: A smooth estimation is performed on the minimum complete subdirectory that possesses the characteristics of a fault three-dimensional structural cluster. An improved mean-drift algorithm is used to cluster the minimal complete subdirectories with fault 3D structural cluster characteristics after smooth estimation. The cluster center points are iteratively updated according to the mean-drift vector until the preset convergence condition is met, resulting in multiple seismic clusters.

[0052] In an optional embodiment, the clustering module 22 includes a tomography identification unit, used for: Calculate the spatial distance between earthquake clusters and select earthquake clusters whose spatial distance is lower than a preset geological scale threshold as candidate earthquake clusters of the same active fault. Based on surface geological evidence, the candidate earthquake clusters are assessed to identify active faults and the corresponding small earthquake clusters.

[0053] In one alternative embodiment, the initial fault model construction module 23 is configured to: A robust local weighted regression algorithm is used to fit the three-dimensional spatial point set of the small earthquake clusters corresponding to the active fault, and a trend line of the small earthquake clusters reflecting the extension direction of the small earthquake clusters is obtained. The positions of multiple three-dimensional slices are determined along the small earthquake cluster trend line, and the fault interpretation lines on each three-dimensional slice are obtained by fitting with the least squares method. Based on surface rupture data and fault interpretation lines fitted on all three-dimensional slices, a discrete smoothing interpolation algorithm is used to spatially continuum the discrete fault interpretation lines, thereby constructing an initial seismic active fault model.

[0054] It should be noted that the seismic active fault model construction system based on spatial data mining and geological constraints provided in this embodiment of the invention is used to execute all the process steps of the seismic active fault model construction method based on spatial data mining and geological constraints in the above embodiment. The working principles and beneficial effects of the two are one-to-one, so they will not be described again.

[0055] This invention also provides a terminal device, such as... Figure 5 The diagram shown is a structural block diagram of a preferred embodiment of a terminal device provided by the present invention. The terminal device includes a processor 31, a memory 32, and a computer program stored in the memory 32 and configured to be executed by the processor 31. When the processor 31 executes the computer program, it implements the method for constructing seismic active fault models based on spatial data mining and geological constraints as described in any of the above embodiments.

[0056] In addition, embodiments of the present invention also provide a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the seismic active fault model construction method based on spatial data mining and geological constraints as described in any of the above embodiments.

[0057] When the processor 31 executes the computer program, it implements the steps in the above-described embodiment of the method for constructing seismic active fault models based on spatial data mining and geological constraints, for example... Figure 1 The method for constructing seismic active fault models based on spatial data mining and geological constraints, as shown, encompasses all steps. Alternatively, when the processor 31 executes the computer program, it implements the functions of each module in the above-described embodiment of the seismic active fault model construction system based on spatial data mining and geological constraints, for example... Figure 4 The diagram shows the functions of each module in the seismic active fault model construction system based on spatial data mining and geological constraints.

[0058] Preferably, the computer program can be divided into one or more modules / units, which are stored in the memory 32 and executed by the processor 31 to complete the present invention. The one or more modules / units can be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in the terminal device.

[0059] The processor 31 can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor, or the processor 31 can be any conventional processor. The processor 31 is the control center of the terminal device, connecting various parts of the terminal device through various interfaces and lines.

[0060] The memory 32 mainly includes a program storage area and a data storage area. The program storage area can store the operating system, applications required for at least one function, etc., and the data storage area can store related data, etc. In addition, the memory 32 can be a high-speed random access memory, or a non-volatile memory, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, and a flash card, etc., or the memory 32 can also be other volatile solid-state storage devices.

[0061] It should be noted that the aforementioned terminal devices may include, but are not limited to, processors and memory, as will be understood by those skilled in the art. Figure 5 The structural block diagram shown is merely a structural example of the terminal device described above and does not constitute a limitation on the structure of the terminal device. The terminal device may include more or fewer components than shown, or combine certain components, or use different components.

[0062] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for constructing a seismic active fault model based on spatial data mining and geological constraints, characterized in that, include: Seismic data from the earthquake zone are organized and analyzed to extract a minimal complete subdirectory; The spatial nearest neighbor index algorithm is used to determine whether the minimum complete subdirectory has fault three-dimensional structure cluster features, and the minimum complete subdirectory with fault three-dimensional structure cluster features is obtained. The mean-shift algorithm is used to cluster the minimal complete subdirectory with fault three-dimensional structure cluster characteristics to obtain multiple seismic clusters; By identifying the multiple earthquake families, the cluster of minor earthquakes corresponding to each active fault is obtained; Three-dimensional automatic slicing is performed on the small earthquake clusters corresponding to each active fault, and the fault interpretation line is fitted using the least squares method. Based on the fault interpretation line, an initial three-dimensional model of the active earthquake fault is constructed. The initial active seismic fault model was supplemented and corrected based on geological constraints to obtain a fine three-dimensional model of the active seismic faults in the seismic zone.

2. The method for constructing a seismic active fault model based on spatial data mining and geological constraints as described in claim 1, characterized in that, The mean-shift algorithm is used to cluster the minimal complete subdirectory with fault 3D structural cluster characteristics, resulting in multiple seismic clusters, including: A smooth estimation is performed on the minimum complete subdirectory that possesses the characteristics of a fault three-dimensional structural cluster. The mean-drift algorithm is used to cluster the minimal complete subdirectories with fault 3D structural cluster characteristics after smooth estimation. The cluster center points are iteratively updated according to the mean-drift vector until the preset convergence condition is met, resulting in multiple seismic clusters.

3. The method for constructing a seismic active fault model based on spatial data mining and geological constraints as described in claim 2, characterized in that, The identification of the multiple earthquake families, resulting in the cluster of small earthquakes corresponding to each active fault, includes: Calculate the spatial distance between earthquake clusters and select earthquake clusters whose spatial distance is lower than a preset geological scale threshold as candidate earthquake clusters of the same active fault. Based on surface geological evidence, the candidate earthquake clusters are assessed to identify active faults and the corresponding small earthquake clusters.

4. The method for constructing a seismic active fault model based on spatial data mining and geological constraints as described in claim 1, characterized in that, The process of automatically slicing the small earthquake clusters corresponding to each active fault in three dimensions and fitting the fault interpretation lines using the least squares method, and constructing an initial three-dimensional model of the active earthquake fault based on the fault interpretation lines, includes: A robust local weighted regression algorithm is used to fit the three-dimensional spatial point set of the small earthquake clusters corresponding to the active fault, and a trend line of the small earthquake clusters reflecting the extension direction of the small earthquake clusters is obtained. The positions of multiple three-dimensional slices are determined along the small earthquake cluster trend line, and the fault interpretation lines on each three-dimensional slice are obtained by fitting with the least squares method. Based on surface rupture data and fault interpretation lines fitted on all three-dimensional slices, a discrete smoothing interpolation algorithm is used to spatially continuum the discrete fault interpretation lines, thereby constructing an initial seismic active fault model.

5. The method for constructing a seismic active fault model based on spatial data mining and geological constraints as described in claim 1, characterized in that, The process of supplementing and correcting the initial seismic active fault model based on geological constraints to obtain a refined three-dimensional model of the seismic active faults in the seismic zone includes: Determine the activity nature of each active fault based on geological constraints; For each branch active fault, fault interpretation lines and three-dimensional seismic active fault models are constructed to determine the hierarchical relationship between the branch active fault and the main active fault structure. Based on the activity characteristics and the hierarchical correlation, the fault structure of the initial seismic active fault model is supplemented and corrected to obtain a refined seismic active fault model of the seismic zone.

6. The method for constructing a seismic active fault model based on spatial data mining and geological constraints as described in claim 1, characterized in that, After supplementing and correcting the initial seismic active fault model according to geological constraints to obtain a detailed three-dimensional model of the seismic active fault in the seismic zone, the method further includes: The detailed model of active faults in the earthquake zone was spatially registered and fused with the Digital Earth Engine.

7. A system for constructing seismic active fault models based on spatial data mining and geological constraints, characterized in that, include: The data processing module is used to organize and analyze seismic data from the seismic zone in order to extract the minimum complete subdirectories. The spatial nearest neighbor index algorithm is used to determine whether the minimum complete subdirectory has fault three-dimensional structure cluster features, and the minimum complete subdirectory with fault three-dimensional structure cluster features is obtained. The clustering module is used to cluster the minimal complete subdirectory with fault three-dimensional structure clustering characteristics using an improved mean drift algorithm to obtain multiple seismic clusters; By identifying the multiple earthquake families, the cluster of minor earthquakes corresponding to each active fault is obtained; The initial fault model construction module is used to automatically slice the small earthquake clusters corresponding to each active fault in three dimensions and fit the fault interpretation line using the least squares method, and construct the initial three-dimensional model of the active earthquake fault based on the fault interpretation line. The fault model correction module is used to supplement and correct the initial seismic active fault model according to geological constraints, so as to obtain a three-dimensional fine model of the seismic active fault in the seismic zone.

8. The seismic active fault model construction system based on spatial data mining and geological constraints as described in claim 7, characterized in that, The initial fault model construction module is used for: A robust local weighted regression algorithm is used to fit the three-dimensional spatial point set of the small earthquake clusters corresponding to the active fault, and a trend line of the small earthquake clusters reflecting the extension direction of the small earthquake clusters is obtained. The positions of multiple three-dimensional slices are determined along the small earthquake cluster trend line, and the fault interpretation lines on each three-dimensional slice are obtained by fitting with the least squares method. Based on surface rupture data and fault interpretation lines fitted on all three-dimensional slices, a discrete smoothing interpolation algorithm is used to spatially continuum the discrete fault interpretation lines, thereby constructing an initial seismic active fault model.

9. A terminal device, characterized in that, The method includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the method for constructing seismic active fault models based on spatial data mining and geological constraints as described in any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the seismic active fault model construction method based on spatial data mining and geological constraints as described in any one of claims 1 to 6.

Images