Ground stress field three-dimensional dynamic inversion method based on multi-scale adaptive algorithm

Abstract

The invention relates to the technical field of crustal stress field data processing, in particular to a crustal stress field three-dimensional dynamic inversion method based on a multi-scale adaptive algorithm. The method comprises the following steps: acquiring a geological data set of a target area; constructing a crustal stress field three-dimensional initial model based on the geological data set, and performing geologic body space division and mesh generation to obtain crustal stress field three-dimensional mesh model data; performing multi-scale region division on the crustal stress field three-dimensional grid model data, and establishing a multi-scale weighting function to obtain multi-scale partition mapping information; and constructing a cross-scale boundary adaptive transmission mechanism, and establishing a stress tensor continuity constraint model at a multi-scale partition boundary to obtain cross-scale stress boundary coupling data. Through a multi-scale adaptive algorithm and dynamic closed-loop optimization, high-precision, dynamic and continuous inversion of a crustal stress field in a complex geologic structure is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of geostress field data processing, and in particular to a three-dimensional dynamic inversion method of geostress field based on a multi-scale adaptive algorithm. Background Art

[0002] In the fields of geological engineering, rock and soil mechanics, and energy development, accurate inversion of the geostress field is of key significance for engineering safety assessment, fracture prediction, and fracturing optimization. Traditional geostress inversion methods are mainly based on single-scale theories and homogeneous assumptions, relying on limited measurement point data and simplified models for inversion. Although they have achieved quantitative calculation of the stress field to a certain extent, they have problems such as poor adaptability to complex geological structures, local distortion of inversion results, and limited spatial resolution. At present, multi-scale geostress modeling methods have shown certain advantages in dealing with formation heterogeneity and sudden changes in stress gradients, but there are still major technical bottlenecks in terms of scale boundary continuity processing, parameter initialization, and inversion convergence efficiency. On the one hand, traditional methods often use fixed-scale grids or manual region division, resulting in inversion results that are highly subject to human intervention and prominent boundary discontinuities; on the other hand, existing methods lack a dynamic adjustment mechanism, making it difficult to optimize the model in real time based on residual feedback, resulting in a lack of adaptability and low iteration efficiency in the inversion process. Summary of the Invention

[0003] Based on this, it is necessary for the present invention to provide a three-dimensional dynamic inversion method of the geostress field based on a multi-scale adaptive algorithm to solve at least one of the above technical problems.

[0004] To achieve the above objectives, a three-dimensional dynamic inversion method for geostress field based on a multi-scale adaptive algorithm is proposed, comprising the following steps: Step S1: Acquire a geological data set of the target area; construct an initial three-dimensional model of the geostress field based on the geological data set, and perform geological body spatial division and grid subdivision to obtain three-dimensional grid model data of the geostress field; Step S2: Perform multi-scale regional division on the three-dimensional grid model data of the geostress field, establish a multi-scale weight function, and obtain multi-scale partition mapping information; construct a cross-scale boundary adaptive transfer mechanism, establish a stress tensor continuity constraint model at the multi-scale partition boundary, and obtain cross-scale stress boundary coupling data; Step S3: Based on the multi-scale partition mapping information and cross-scale stress boundary coupling data, the adaptive initial inversion parameters of each scale region are set, the multi-scale adaptive inversion objective function is constructed, and a joint inversion is performed to obtain the preliminary three-dimensional distribution results of the geostress field; Step S4: Perform dynamic error evaluation and residual distribution analysis on the preliminary three-dimensional distribution results of the geostress field, identify local abnormal areas, and perform high-resolution intensified inversion on the local areas based on the multi-scale adaptive inversion objective function to obtain the optimized three-dimensional dynamic distribution results of the geostress field; Step S5: Acquire actual microseismic response data; perform coupled comparative analysis on the optimized three-dimensional dynamic distribution results of the geostress field and the actual microseismic response data to generate a dynamic fit evaluation index; if the fit is lower than the preset threshold, update the multi-scale weight function and inversion parameters, and iterate steps S2 to S5 until the convergence condition is met to obtain the final three-dimensional dynamic inversion data of the geostress field.

[0005] The present invention improves the adaptability and resolution of the model to complex geological structures by introducing multi-scale regional division and weight control mechanism; constructs a cross-scale boundary stress tensor continuity constraint model to achieve the physical consistency of the multi-scale inversion boundary and the stability of parameter transfer; adopts adaptive initial parameter setting and dynamic error feedback mechanism to enhance the convergence efficiency and model adjustment capability of the inversion process; combines the residual-driven local encryption inversion strategy to perform high-resolution optimization of abnormal areas and significantly improve the overall inversion accuracy; finally, by introducing actual microseismic response data for fitting verification and closed-loop iterative optimization, it ensures that the inversion results have high geological consistency and physical credibility, thereby realizing the refinement, dynamicization and high-reliability inversion of the stress field in complex tectonic areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0006] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments thereof made with reference to the following drawings: Figure 1 Schematic diagram of the steps of a three-dimensional dynamic inversion method of geostress field based on a multi-scale adaptive algorithm of the present invention; Figure 2 for Figure 1 Detailed step flow diagram of step S1; Figure 3 for Figure 1 Detailed step flow diagram of step S2; Figure 4 This is a navigation diagram of the implementation process of an embodiment of the present invention. DETAILED DESCRIPTION

[0007] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. It is obvious that the embodiments described are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts are within the scope of protection of the present invention.

[0008] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0009] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0010] To achieve this, please refer to Figures 1 to 4 The present invention provides a three-dimensional dynamic inversion method of geostress field based on a multi-scale adaptive algorithm, the method comprising the following steps: Step S1: Acquire a geological data set of the target area; construct an initial three-dimensional model of the geostress field based on the geological data set, and perform geological body spatial division and grid subdivision to obtain three-dimensional grid model data of the geostress field; Step S2: Perform multi-scale regional division on the three-dimensional grid model data of the geostress field, establish a multi-scale weight function, and obtain multi-scale partition mapping information; construct a cross-scale boundary adaptive transfer mechanism, establish a stress tensor continuity constraint model at the multi-scale partition boundary, and obtain cross-scale stress boundary coupling data; Step S3: Based on the multi-scale partition mapping information and cross-scale stress boundary coupling data, the adaptive initial inversion parameters of each scale region are set, the multi-scale adaptive inversion objective function is constructed, and a joint inversion is performed to obtain the preliminary three-dimensional distribution results of the geostress field; Step S4: Perform dynamic error evaluation and residual distribution analysis on the preliminary three-dimensional distribution results of the geostress field, identify local abnormal areas, and perform high-resolution intensified inversion on the local areas based on the multi-scale adaptive inversion objective function to obtain the optimized three-dimensional dynamic distribution results of the geostress field; Step S5: Acquire actual microseismic response data; perform coupled comparative analysis on the optimized three-dimensional dynamic distribution results of the geostress field and the actual microseismic response data to generate a dynamic fit evaluation index; if the fit is lower than the preset threshold, update the multi-scale weight function and inversion parameters, and iterate steps S2 to S5 until the convergence condition is met to obtain the final three-dimensional dynamic inversion data of the geostress field.

[0011] Preferably, step S1 includes the following steps: Step S11: Acquire drilling data, well logging data, seismic profiles, geological mapping and structural interpretation data of the target area to obtain a geological data set of the target area; Step S12: extracting the main faults, stratigraphic interfaces and lithologic boundary information based on the geological data set, establishing a structural boundary framework, and obtaining a geological structural framework model; Step S13: spatially dividing different lithologic units in the geological structure framework model to obtain geological body spatial partition data; Step S14: performing irregular spatial grid division based on the geological body spatial partition data to obtain a three-dimensional initial grid division model of the geostress field; Step S15: applying boundary conditions and initial stress states to the three-dimensional initial mesh model to establish a three-dimensional initial model of the geostress field; Step S16: Bind and integrate the three-dimensional initial model of the geostress field with the three-dimensional initial grid subdivision model to generate three-dimensional grid model data of the geostress field.

[0012] In the embodiment of the present invention, first, by arranging standardized drilling holes in the target area, using a diameter of not less than The 91mm diamond coring drilling method is used to obtain continuous core samples in the depth range of 0 to 2000m, and the corresponding drilling histograms and lithologic comparison data are collected; the logging data obtained by the direct wave and refracted wave joint processing method, including the formation P-wave velocity, density curve and elastic modulus data, the logging depth is not less than 1000m, and the well spacing is less than 200m to ensure the resolution of geological information; then, the two-dimensional seismic reflection profile data is used, and the structural boundary enhancement technology based on amplitude attributes is adopted to extract the main fault position, stratigraphic unconformity surface and sedimentary bedding characteristics, and the 1:5000 geological mapping results and structural interpretation data are integrated to establish a structural boundary framework with the fault strike, dip and extension depth accuracy controlled within ±5°, and generate a geological structure framework model under three-dimensional coordinates; under this framework, according to the differences in the physical and mechanical parameters of the lithologic units, the model is divided into no less than 10 lithologic units. The spatial partitioning data of the geological body are generated by a spatial partitioning method based on the principle of geometric consistency. Subsequently, the Delaunay tetrahedron partitioning method for irregular three-dimensional structures is used to partition the spatial partitioning data of the geological body, and a three-dimensional initial grid partitioning model of the geostress field is generated with the side length of the partitioning unit controlled in the range of 10m~50m. After comparing the grid structure with the actual geological boundary, a linear stress gradient boundary condition is imposed according to the main direction of the regional tectonic stress, and the boundary tensor value range is set to 0~20MPa. The initial stress state is constructed according to the geological time scale. On this basis, the structured data integration interface is used to bind the initial stress state results with the three-dimensional grid partitioning data at the unit level, and the node number consistency verification method is used to ensure the accuracy of the model coupling, and finally a three-dimensional grid model data of the geostress field with a one-to-one correspondence between node, unit and stress tensor is generated.

[0013] The present invention can fully integrate multi-source heterogeneous data such as drilling, logging, seismic and geological interpretation, and comprehensively improve the integrity and accuracy of geological information; realize the precise extraction and structural modeling of major structural elements and lithologic boundaries, which helps to truly reflect the complexity of underground geological structures; enhance the model's ability to resolve heterogeneous geological bodies through fine spatial division of different lithologic units; adopt irregular spatial grid partitioning method to improve the adaptability and computational efficiency of numerical simulation while maintaining the true geometric form of geological structures; establish physical constraints on initial stress state and boundary conditions to ensure that the subsequent inversion process has a reasonable initial mechanical foundation; and finally, realize the organic unity of structural information and stress field through model integration operation, providing solid data support and physical foundation for multi-scale high-precision ground stress inversion.

[0014] Preferably, step S2 includes the following steps: Step S21: performing cluster analysis on the spatial characteristic scales in the three-dimensional grid model data of the geostress field to obtain scale sensitivity classification results; Step S22: Based on the scale sensitivity classification results, the three-dimensional grid model data of the geostress field is divided into regions to obtain preliminary multi-scale partitioning results; Step S23: constructing a spatial mapping relationship between the multi-scale partitioning and the original model based on the preliminary result of the multi-scale partitioning to obtain multi-scale partitioning mapping information; Step S24: construct a cross-scale boundary adaptive transfer mechanism based on the multi-scale partition mapping information, establish a stress tensor continuity constraint model at the multi-scale partition boundary, and obtain cross-scale stress boundary coupling data.

[0015] In the embodiment of the present invention, first, based on the obtained three-dimensional grid model data of the geostress field, a joint feature description method based on the center point position of the grid unit and the unit side length is adopted to extract the spatial scale feature vector, and a hierarchical clustering processing method based on the Euclidean distance metric is used to perform scale sensitivity analysis. The clustering threshold is set to 0.15, and the maximum number of classifications is limited to 6 categories. The scale sensitivity classification label to which each grid unit belongs, that is, the scale sensitivity classification result, is obtained; then, the grid units with the same label are regionally merged according to the spatial adjacency relationship, and the region extraction method based on the spatial connected domain is adopted to eliminate the grid units with an area less than 100. Or discrete areas with less than 20 units are used to complete the generation of preliminary multi-scale partitioning results; on this basis, for each multi-scale regional unit, its corresponding number, node coordinates and topological relationship in the original three-dimensional grid model are recorded, and a mapping index table from the multi-scale region to the original model is constructed. The index table includes the region number, starting unit number, node number mapping relationship and boundary unit identification field to form multi-scale partition mapping information; at the multi-scale partition boundary position, the node pair set on the shared grid surface between adjacent regions is extracted, and the validity screening is carried out based on the criteria of surface normal vector consistency and unit volume difference ratio less than 0.2 to obtain the cross-scale contact boundary node set; Furthermore, based on this set and mapping information, the tensor projection direction and adjacent path length between each pair of boundary nodes are defined, and the normalized weighted method is used to calculate the stress weight adjustment coefficient of each pair of nodes to construct a stress weight adjustment parameter group. Based on this parameter group and the relationship between the principal axis directions of the tensors at the boundary nodes, the tensor interpolation function is used to derive the tensor transition function expression between the nodes, and then the stress tensor continuity constraint formula is established; the constraint formula is discretized to form a stress tensor coordination matrix, and the matrix is ​​projected into the boundary data structure of the original three-dimensional grid model of the stress field to realize the encoding and embedding of cross-scale coupled tensor data, and finally generate cross-scale stress boundary coupling data.

[0016] By performing cluster analysis on spatial characteristic scales, the present invention can automatically identify scale difference characteristics in the geostress field, realize the objectivity and intelligence of the regional division process, and avoid the errors and subjectivity caused by artificial division; multi-scale regional division is carried out in combination with scale sensitivity results, thereby enhancing the adaptability and expression ability of the model to structural units of different scales; by constructing a spatial mapping relationship between the partition and the original model, the consistency and coordination between the regional division and the overall geological structure are ensured; further introduction of the cross-scale boundary adaptive transmission mechanism and the stress tensor continuity constraint model not only effectively eliminates the physical discontinuity problem at the multi-scale boundary, but also ensures the coordinated transmission of stress information between units of different scales, thereby providing key support and physical consistency guarantee for high-precision, multi-scale linkage geostress field inversion.

[0017] Preferably, step S24 includes the following steps: Step S241: constructing a scale response weight function based on the multi-scale partition mapping information to obtain multi-scale stress weight distribution data; Step S242: performing pairing identification on the boundary grid nodes of each region in the multi-scale partition mapping information to obtain a cross-scale contact boundary node set; Step S243: constructing a cross-scale boundary adaptive transfer mechanism based on the cross-scale contact boundary node set and the multi-scale stress weight distribution data to obtain a stress weight adjustment parameter group; Step S244: constructing a stress tensor continuity constraint model based on the stress weight adjustment parameter group and the preset boundary node tensor state to obtain a multi-scale boundary stress tensor coordination matrix; Step S245: Projecting the multi-scale boundary stress tensor coordination matrix into the boundary data structure of the three-dimensional grid model data of the geostress field to generate cross-scale stress boundary coupling data.

[0018] In the embodiment of the present invention, first, based on the established multi-scale partition mapping information, the spatial coordinate range, grid density, unit volume and regional structural complexity index corresponding to each scale region are extracted, and the scale response weight function is defined by a combination of exponential functions, where the average unit volume in the region is recorded as , the node density is , the function form is ,in =1.0, =0.02, =1.5, the weight result is assigned to the grid nodes in each region to generate multi-scale stress weight distribution data; then, the grid node set identified as the region boundary in the multi-scale mapping information is paired according to the adjacency relationship and the surface normal consistency, and the node pairs located at the junction of different scale regions and the Euclidean distance between nodes is less than 1.5 times the average unit side length are identified by face-by-face screening to construct a cross-scale contact boundary node set; further, based on this set and the stress weight distribution data, the weight value of the unit in which each pair of paired nodes is located is extracted, and the double-node weighted average formula is used to calculate the initial stress transfer coefficient to form an unnormalized transfer parameter matrix, and combined with the boundary node neighbor matrix, the initial stress transfer coefficient is calculated by the double-node weighted average formula to form an unnormalized transfer parameter matrix. The relationship topology information is connected, and the sum normalization processing is performed to generate a stress weight adjustment parameter group; on this basis, the preset tensor state of each boundary node is referenced, which is given by the static boundary load condition and the principal stress direction constraint. The linear tensor interpolation formula is used to construct the tensor continuity expression between nodes, and the expressions of all paired nodes are matrixed to assemble and generate a multi-scale boundary stress tensor coordination matrix; finally, the coordination matrix is ​​projected to the boundary node subset of the three-dimensional grid model data structure of the ground stress field according to the regional index and node index rules, and the sparse matrix interpolation slot structure is used to complete the tensor data embedding operation, unify the numbering and coding specifications, and generate cross-scale stress boundary coupling data.

[0019] The present invention realizes weight adjustment and response characteristic modeling of stress information in different scale areas by constructing a scale response weight function, thereby enhancing the adaptability of the model to the complexity of geological structures; adopts a boundary node pairing identification mechanism to accurately extract cross-scale contact interfaces, thereby improving the integrity and accuracy of boundary processing; introduces a stress weight adjustment parameter group in the boundary information integration process, thereby realizing dynamic balance and coordinated adjustment of stress information among multiple scales; by constructing a stress tensor continuity constraint model, ensures the numerical and physical continuity of the stress tensor at the boundary, and eliminates the boundary tensor mutation problem commonly seen in traditional methods; finally, the coordinated tensor information is effectively embedded in the original three-dimensional grid data structure, thereby ensuring the physical consistency and numerical stability of the entire model under multi-scale coupling conditions, and providing key boundary condition support for subsequent high-precision inversion.

[0020] Preferably, step S243 includes the following steps: Step S2431: performing topological adjacency analysis on the cross-scale contact boundary node set to obtain boundary node adjacency structure data; Step S2432: extracting a cross-scale mapping path sequence based on the boundary node adjacency structure data and the multi-scale partition mapping information to obtain a path index dataset; Step S2433: Calculating the initial stress transfer coefficient based on the path index data set and the multi-scale stress weight distribution data to obtain a path transfer coefficient matrix; Step S2434: performing normalization adjustment processing on the path transfer coefficient matrix to obtain normalized weight distribution data; Step S2435: Generate a stress weight adjustment parameter group based on the normalized weight distribution data and the cross-scale contact boundary node set.

[0021] In an embodiment of the present invention, first, a grid connection table retrieval method based on node index number is adopted to perform topological adjacency analysis on a set of cross-scale contact boundary nodes, identify common nodes on the same grid surface by traversing node by node, construct an adjacency dictionary structure with nodes as indexes, and output boundary node adjacency structure data, which records the first-order adjacency information between nodes. Subsequently, based on the obtained adjacency structure data and multi-scale partition mapping information, all continuous mapping paths composed of boundary nodes are retrieved along the boundary direction in a breadth-first traversal manner, and the node sequence numbers are recorded in the order of the path starting point, end point and intermediate jump points. The multi-scale region sequence traversed by each path is extracted, a cross-scale mapping path sequence is generated, and a path index dataset is constructed. Subsequently, based on the path index dataset and the multi-scale stress weight distribution data associated with each node in the corresponding path, the weighted geometric mean method is used to calculate the initial stress transfer coefficient. Let There are nodes, and the corresponding weight is , then the path transfer coefficient is defined as , all path coefficients are used to form a path transfer coefficient matrix; then the matrix is ​​normalized and adjusted, using row-by-row normalization to divide the transfer coefficient of each path by the sum of all path transfer coefficients, ensuring that the value of each path coefficient after normalization falls within the interval , and keep the sum to 1 to generate normalized weight distribution data; finally, a one-to-one correspondence is established between the normalized weight distribution data and the cross-scale contact boundary node set, and the proportion of each pair of nodes in the path is extracted. Combined with their normalized weight values, the stress influence proportion factor is formed, which is defined as the stress weight adjustment coefficient. The stress weight adjustment coefficients of all node pairs are summarized to generate the stress weight adjustment parameter group.

[0022] By analyzing the topological adjacency relationship of contact boundary nodes, the present invention can accurately reflect the structural connectivity and spatial adjacency characteristics between boundary nodes, providing a topological basis for constructing a continuous stress transfer path; combining multi-scale mapping relationships to extract mapping path sequences, realizing the systematic expression of stress transfer paths between units of different scales; using the path transfer coefficient matrix to quantify the stress transmission capacity of each path, enhancing the model's ability to describe complex boundary stress transfer behavior; through normalization processing, the numerical stability and physical interpretability of weight distribution data are improved, effectively avoiding the transfer imbalance problem caused by excessive weight differences between paths; the final generated stress weight adjustment parameter group can serve as the core basis for fine-tuning cross-scale boundary stress distribution, providing key support for achieving boundary continuity, adaptive adjustment and multi-scale coordination in the inversion of geostress fields.

[0023] Preferably, step S3 includes the following steps: Step S31: performing regional index analysis on the multi-scale partition mapping information to obtain structural configuration data of each scale region; Step S32: setting adaptive initial inversion parameters for each scale region based on the structural configuration data to obtain inversion parameter setting data; Step S33: constructing a multi-scale adaptive inversion objective function based on the inversion parameter setting data and the cross-scale stress boundary coupling data to obtain an inversion objective function model; Step S34: performing a joint inversion solution on the inversion objective function model to obtain preliminary three-dimensional distribution results of the geostress field.

[0024] In the embodiment of the present invention, the multi-scale partition mapping information is first subjected to regional index parsing processing, and the region number, grid node set, cell number set and their corresponding topological relationship recorded in the multi-scale mapping index table are called. The structural configuration data corresponding to each scale region is generated by traversing the index structure. The configuration data includes the spatial position of the region, the number of cells, the average cell volume, the number of boundary nodes, the scale level to which it belongs, and the boundary connection information with the adjacent region; then, the adaptive initial inversion parameters are set for each scale region according to the structural configuration data, where the cell volume is greater than 200. The low-resolution parameter group is set in the region, and the inversion step size is set to , the stress update threshold is MPa, unit volume less than 100 The high-resolution parameter group is set in the region, and the inversion step size is set to , the stress update threshold is MPa, a cross-scale weight adjustment parameter is introduced as a constraint factor at the regional boundary according to the adjacency information, and inversion parameter setting data including step size, constraint coefficient, update criterion and upper limit of iteration number is constructed for each region; then the above-mentioned inversion parameter setting data is fused with the cross-scale stress boundary coupling data generated in step S24, and an energy functional expression for minimizing the tensor difference is established for each scale region, and a tensor continuity constraint term, a boundary coordination term and a scale adaptation term are introduced to form a multi-scale adaptive inversion objective function, which is uniformly written into the objective function database as an inversion objective function model; finally, a joint inversion solution operation is performed on the objective function model, and a block iterative solution process is adopted to split the objective function corresponding to each scale region by region, and the tensor distribution values ​​of all regions are jointly solved by the residual update method, and the maximum number of iterations is set to 100 times or the overall tensor difference is less than MPa is used as the stopping criterion, and the set of unit tensor values ​​in all regions is output. The tensor reconstruction operation is performed to generate the preliminary three-dimensional distribution results of the geostress field.

[0025] The present invention accurately grasps the structural characteristics of regions at various scales through regional index analysis, providing a scientific basis for parameter setting; based on the dynamic adjustment of adaptive initial inversion parameters, the adaptability of the inversion model to regions of different scales is improved, and the flexibility and accuracy of the inversion process are enhanced; the construction of the multi-scale adaptive inversion objective function integrates cross-scale boundary coupling information, ensuring the continuity and physical consistency of the model in spatial scale; the joint inversion solution method effectively integrates multi-scale information, realizes the coordinated optimization of the inversion process, and significantly improves the accuracy and stability of the three-dimensional distribution results of the ground stress field, providing a reliable technical guarantee for stress field inversion under complex geological conditions.

[0026] Preferably, step S34 includes the following steps: Step S341: performing model dimension expansion processing on the inversion objective function model to obtain inversion dimension configuration data of each scale area; Step S342: setting a joint inversion parameter set based on the inversion dimension configuration data of each scale region to obtain initial joint inversion parameters; Step S343: constructing a joint optimization solution process based on the joint inversion initial parameters to obtain joint inversion process configuration data; Step S344: performing multiple rounds of numerical optimization iterations on the joint inversion process configuration data to obtain a multi-scale stress response numerical result set; Step S345: reconstructing the three-dimensional geostress tensor distribution based on the multi-scale stress response numerical result set to obtain a preliminary three-dimensional geostress field distribution result.

[0027] In the embodiment of the present invention, first, a dimensional expansion process is performed on the inversion objective function model, and the entire ground stress field is divided into multiple independent scale regions using a regional block numbering method. The number of nodes, the number of tensor components, the proportion of boundary nodes, and the number of stress coupling constraint items are extracted for each region. The regional number is used as an index to form inversion dimension configuration data, which clearly specifies the stress tensor solution dimension, the number of constraints, and the calculation block size for each region. Based on this dimensional configuration data, a joint inversion parameter set is set, and the solution variables are set according to the independence of the tensor distribution characteristics in the three principal stress directions. The tensor update step size is defined as MPa, the constraint penalty coefficient is 0.8, the tensor boundary pull-through factor is 1.2, and the error tolerance upper limit is set to MPa, and uniformly write the initial parameter set of the joint inversion; then build a joint optimization solution process, perform local iteration and boundary cooperative transfer processing in sequence according to the region number, judge the stability of the local solution by the least squares residual convergence criterion in each iteration, and find its cross-scale contact node pair according to the boundary node number, correct the boundary node tensor value by weighted average, and generate the joint inversion process configuration data, including the iteration sequence table, boundary cooperative transfer path, residual evaluation standard and intermediate result buffer configuration; perform multiple rounds of numerical optimization iterations based on the configuration data, and adopt a combination of region-by-region synchronous update and cross-region staggered update to adjust the internal balance of the stress response of each region in each iteration, and use the boundary tensor consistency error less than MPa is used as the round termination standard, and the stress response result set corresponding to each scale area is finally output; the stress response numerical result set is input into the three-dimensional space reconstruction module, and the complete stress tensor data of each grid node is restored through node interpolation. The tensor reconstruction process adopts the tensor principal value direction coincidence interpolation method to ensure the physical consistency between the tensor components, and writes them into the three-dimensional data structure in sequence according to the grid number, completing the preliminary tensor reconstruction of the three-dimensional distribution of the entire ground stress field and obtaining the preliminary three-dimensional distribution results of the ground stress field.

[0028] The present invention dimensionalizes the inversion objective function, divides the inversion tasks into different scale areas, and improves the pertinence and rationality of parameter configuration; the setting of the joint inversion parameter set and the process construction realize the system integration and collaborative optimization of multi-scale information, and enhances the overall consistency of the inversion process; multiple rounds of numerical optimization iterations effectively improve the convergence speed of the inversion model and the stability of the solution, and reduce the influence of local extreme value traps; finally, the three-dimensional geostress tensor reconstruction based on the response numerical results ensures the spatial continuity and physical authenticity of the inversion results, significantly improves the accuracy and reliability of the three-dimensional distribution of the geostress field, and provides solid technical support for geostress assessment under complex geological conditions.

[0029] Preferably, step S4 includes the following steps: Step S41: performing global error field calculation on the preliminary three-dimensional distribution result of the geostress field to obtain error distribution data; Step S42: performing residual statistics and spatial fitting analysis on each scale region based on the error distribution data to obtain residual distribution information; Step S43: performing anomaly detection and cluster identification on the residual distribution information to obtain local abnormal area identification data; Step S44: extracting a corresponding three-dimensional grid model of the local stress field based on the local abnormal area identification data to obtain local area model data; Step S45: Refining the grid and resetting the boundary conditions of the local area model data to obtain a high-resolution inversion input model; Step S46: performing local intensified inversion based on the high-resolution inversion input model and the multi-scale adaptive inversion objective function to obtain optimized local distribution data of the geostress field; Step S47: performing weighted fusion on the optimized local distribution data of the geostress field and the preliminary three-dimensional distribution result of the geostress field to obtain the optimized three-dimensional dynamic distribution result of the geostress field.

[0030] In the embodiment of the present invention, first, the global error field is calculated for the preliminary three-dimensional distribution result of the ground stress field, and the tensor difference between the node-level stress tensor and the boundary continuity constraint is used as the basis for error estimation. By constructing the error tensor norm function and calculating the error modulus on the grid nodes of the whole field, three-dimensional error distribution data is formed, and its value range is limited to 0 to 5 MPa, and the resolution is one level per 10m grid. Then, based on the error distribution data, the residual statistics in the region are performed according to the multi-scale regional index, and the maximum residual value, mean, standard deviation and residual direction distribution characteristics are extracted. The spatial fitting method is used to construct the residual isosurface for each region and analyze its distribution trend to form residual distribution information containing numerical value, position and structural information. The residual information is further subjected to anomaly detection and cluster recognition, and the high residual dense area is extracted by using the density threshold discrimination method. The clustering threshold is set to 2 MPa and the minimum cluster volume is 100. , screen out the local stress incoordination area, mark its spatial index and boundary unit node, and output the local abnormal area identification data; extract the relevant units and their connection structure in the three-dimensional grid of the corresponding ground stress field according to the identification data, construct the local area model data, and perform 10% boundary expansion processing on the extraction range to ensure boundary continuity; after obtaining the local area model data, use the tetrahedron subdivision method to refine the grid, reduce the unit volume to 1 / 4 of the original volume, and control the grid subdivision error to 1%. At the same time, re-apply the tensor boundary constraint to the regional boundary, and perform linear interpolation of the tensor boundary value according to the average stress value of the adjacent area to generate a high-resolution inversion input model; based on the model and the constructed multi-scale adaptive inversion objective function, limit the inversion iteration round upper limit to 50 rounds, and use the tensor residual update method to perform local area inversion calculation, output the optimized local distribution data of the ground stress field, and the tensor accuracy is controlled at MPa; finally, the local distribution data of the optimized geostress field are weightedly fused with the original preliminary three-dimensional distribution results of the geostress field. The fusion weight is determined according to the standard deviation of the residuals in the local area. The higher the standard deviation, the greater the weight of the optimization result. The fusion method adopts the form of tensor weighted superposition, and the weighted average operation is performed on the node tensor components, which are written into a unified three-dimensional spatial data structure to finally generate the optimized three-dimensional dynamic distribution results of the geostress field.

[0031] The present invention realizes the precise identification and quantification of the error distribution in the inversion results by comprehensively calculating the error field and combining residual statistics with spatial fitting analysis; adopts anomaly detection and clustering identification methods to accurately locate local abnormal areas, providing a scientific basis for targeted optimization; refines the grid and resets the boundary conditions of the abnormal area model, significantly improving the spatial resolution and physical reality of the local model; performs local encrypted inversion based on the high-resolution input model, effectively corrects the local inversion error, and improves the model's detail depiction capability; finally, optimizes the local and overall inversion results through weighted fusion, ensuring the continuity and stability of the three-dimensional dynamic distribution of the ground stress field, realizing high-precision, dynamically adaptive ground stress inversion, and providing a solid guarantee for engineering applications in complex geological environments.

[0032] Preferably, in step S5, coupling and comparing the optimized three-dimensional dynamic distribution results of the geostress field with the actual microseismic response data to generate a dynamic fit evaluation index includes the following steps: The actual microseismic response data is filtered, de-noised and time-windowed to obtain standardized microseismic response data. Based on the standardized microseismic response data, a microseismic response spatial distribution model is constructed to obtain microseismic response field data; The optimized three-dimensional dynamic distribution results of the ground stress field are spatially coupled and compared with the microseismic response field data to obtain the coupling residual distribution results; Multi-scale weighted statistical analysis is performed on the coupling residual distribution results to obtain dynamic fit evaluation indicators.

[0033] In the embodiment of the present invention, first, the original microseismic waveform data obtained by the same three-component seismic geophone array in the target area within the past three months are collected with a sampling frequency of 1000 Hz and a duration of not less than 60 seconds. The microseismic waveform data are filtered in the frequency band of 20-250 Hz using a bandpass filter, and the signal is time-windowed and resampled in units of 10 seconds using a smoothing window function. The processed data are mean-normalized to obtain unitless standardized microseismic response data. Secondly, the spatial coordinates and response intensity information of the microseismic event are extracted based on the maximum amplitude value and trigger time of each measuring point in the standardized microseismic response data. A three-dimensional regular grid is constructed in combination with the spatial position of the geophone layout. The microseismic response intensity value of each grid node is calculated at the point using the three-dimensional inverse distance weighted interpolation method to generate a spatial resolution of 20 m of microseismic response field data; subsequently, the optimized three-dimensional dynamic distribution results of the geostress field are interpolated according to the same spatial grid to ensure that the two sets of data are consistent in spatial resolution and grid structure, and the square of the difference between the microseismic response intensity value and the principal stress modulus of the geostress tensor is calculated at each corresponding grid node. The square of the difference results of all grid nodes are summarized to obtain the coupling residual distribution results in the three-dimensional spatial domain; then, the coupling residual distribution results are subjected to spatial partitioning statistical processing based on the geological structure partitioning and scale sensitivity classification results, and the mean, standard deviation and coefficient of variation of the residual values ​​in each partition are calculated respectively. The statistical indicators of each partition are weighted and accumulated according to the scale response weight, and finally a unified dimension dynamic fit evaluation index is generated. This index is limited to [0, 1]. The closer the value is to 1, the higher the coupling degree. If the evaluation index is less than 0.75, the initial geostress parameter configuration is updated according to the residual distribution characteristics and the geostress inversion process is iteratively executed until the evaluation index meets the threshold requirement.

[0034] The present invention effectively improves the signal-to-noise ratio and time consistency of microseismic data through filtering denoising and time window resampling processing, ensuring the reliability and standardization of the data; constructs a spatial distribution model of microseismic response to achieve precise spatial positioning and quantitative expression of microseismic events; spatially couples and compares the stress field data with the microseismic response field, accurately revealing the correlation characteristics between stress field changes and microseismic activities; utilizes a multi-scale weighted statistical analysis method to comprehensively reflect the fitting quality at different spatial scales, forming a dynamic fitting evaluation index, which provides a scientific basis for the dynamic correction and parameter optimization of the inversion model, and effectively promotes the dynamic adaptive iteration and refinement of the inversion results of the stress field.

[0035] Preferably, if the degree of fit is lower than a preset threshold in step S5, the multi-scale weight function and the inversion parameters are updated, and steps S2 to S5 are iteratively performed until the convergence condition is met, and the final three-dimensional dynamic inversion data of the geostress field is obtained, including: Based on the dynamic fit evaluation index, determine whether it is lower than the preset threshold and obtain the fitting result judgment label; If the fitting result determines that the label is not satisfied, the multi-scale weight function is updated based on the coupling residual distribution result to obtain updated multi-scale weight function data; Adjusting the inverse constraint strategy and the optimization parameter set of the geostress based on the updated multi-scale weight function data, generating an updated inversion parameter configuration, and using the updated inversion parameter configuration to iteratively execute steps S2 to S5; If the fitting result judges that the label is satisfied, the current optimized three-dimensional dynamic distribution result of the geostress field is output as the final three-dimensional dynamic inversion data of the geostress field.

[0036] In the embodiment of the present invention, first, the preset threshold of the dynamic fit evaluation index is set to 0.75. When the evaluation index is calculated, a logical judgment is immediately performed. If the value is lower than 0.75, the current judgment result is assigned a "not satisfied" state label; in this state, the original multi-scale partition mapping information is reconstructed according to the coupled residual distribution result obtained above, and the scale weight coefficient corresponding to the area with the residual value standard deviation greater than 0.3 MPa is selected for adjustment processing. The original scale response weight function is updated using the weighted correction factor to obtain updated multi-scale weight function data; the updated weight function data is used to reset the regional priority in the geostress inversion process, specifically, a higher partition index value is assigned to the area with a weight greater than 0.7, and the upper limit of the inversion iteration number is increased by 20%; then, combined with the updated multi-scale weight function data, the constraint strategy and optimization parameter set in the geostress inversion process are adjusted, wherein the boundary condition forced coefficient is limited to [0.8, 1.2], and the initial principal stress value is linearly perturbed by ±10% to form an updated Inversion parameter configuration; input the updated inversion parameter configuration into the three-dimensional dynamic inversion process of the geostress field, and iterate in the order of steps S2 to S5, that is, starting from multi-scale regional division and boundary coupling data reconstruction, the inversion objective function is updated, the joint inversion solution is solved, the error identification and local encrypted inversion are performed, and the coupling comparison with the microseismic response data is performed again. The above process is cyclically executed until the dynamic fit evaluation index is greater than or equal to 0.75 and lasts for two iteration cycles, and the residual mean change rate is less than 1% for two consecutive times. After the convergence criterion is met, the current optimized three-dimensional dynamic distribution result of the geostress field is directly output as the final three-dimensional dynamic inversion data of the geostress field.

[0037] The present invention can judge the degree of match between the inversion results and the actual microseismic response data in real time, and effectively identify situations where the inversion accuracy is insufficient; dynamically update the multi-scale weight function based on the coupled residual distribution, realize adaptive adjustment of geological heterogeneity and local anomalies, and enhance the flexibility and responsiveness of the model; systematically optimize the inversion process by adjusting the inversion constraint strategy and the optimization parameter set, and improve the efficiency and convergence speed of the numerical solution; iteratively execute the multi-scale adaptive inversion process to ensure that the inversion results continue to approach the true ground stress distribution; the final output three-dimensional dynamic inversion data of the ground stress field has high precision, high consistency and dynamic adaptability, meeting the stringent requirements of engineering applications under complex geological conditions.

[0038] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is not limited by the above description. Therefore, it is intended that all changes that fall within the meaning and scope of the equivalent elements of the application documents are included in the present invention.

[0039] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A three-dimensional dynamic inversion method for geostress field based on a multi-scale adaptive algorithm, characterized in that: The following steps are involved: Step S1: Acquire a geological dataset of the target area; Based on the geological data set, a 3D initial model of the geostress field is constructed, and the geological body space is divided and meshed to obtain the 3D grid model data of the geostress field; Step S2: Perform multi-scale regional division on the three-dimensional grid model data of the geostress field, establish a multi-scale weight function, and obtain multi-scale partition mapping information; construct a cross-scale boundary adaptive transfer mechanism, establish a stress tensor continuity constraint model at the multi-scale partition boundary, and obtain cross-scale stress boundary coupling data; Step S3: Based on the multi-scale partition mapping information and cross-scale stress boundary coupling data, the adaptive initial inversion parameters of each scale region are set, the multi-scale adaptive inversion objective function is constructed, and a joint inversion is performed to obtain the preliminary three-dimensional distribution results of the geostress field; Step S4: Perform dynamic error evaluation and residual distribution analysis on the preliminary three-dimensional distribution results of the geostress field, identify local abnormal areas, and perform high-resolution intensified inversion on the local areas based on the multi-scale adaptive inversion objective function to obtain the optimized three-dimensional dynamic distribution results of the geostress field; Step S5: Acquire actual microseismic response data; perform coupled comparative analysis on the optimized three-dimensional dynamic distribution results of the geostress field and the actual microseismic response data to generate a dynamic fit evaluation index; if the fit is lower than the preset threshold, update the multi-scale weight function and inversion parameters, and iterate steps S2 to S5 until the convergence condition is met to obtain the final three-dimensional dynamic inversion data of the geostress field.

2. The three-dimensional dynamic inversion method of geostress field based on a multi-scale adaptive algorithm according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: Acquire drilling data, well logging data, seismic profiles, geological mapping and structural interpretation data of the target area to obtain a geological data set of the target area; Step S12: extracting the main faults, stratigraphic interfaces and lithologic boundary information based on the geological data set, establishing a structural boundary framework, and obtaining a geological structural framework model; Step S13: spatially dividing different lithologic units in the geological structure framework model to obtain geological body spatial partition data; Step S14: performing irregular spatial grid division based on the geological body spatial partition data to obtain a three-dimensional initial grid division model of the geostress field; Step S15: applying boundary conditions and initial stress states to the three-dimensional initial mesh model to establish a three-dimensional initial model of the geostress field; Step S16: Bind and integrate the three-dimensional initial model of the geostress field with the three-dimensional initial grid subdivision model to generate three-dimensional grid model data of the geostress field.

3. The three-dimensional dynamic inversion method of geostress field based on a multi-scale adaptive algorithm according to claim 1 is characterized in that: Step S2 includes the following steps: Step S21: performing cluster analysis on the spatial characteristic scales in the three-dimensional grid model data of the geostress field to obtain scale sensitivity classification results; Step S22: Based on the scale sensitivity classification results, the three-dimensional grid model data of the geostress field is divided into regions to obtain preliminary multi-scale partitioning results; Step S23: constructing a spatial mapping relationship between the multi-scale partitioning and the original model based on the preliminary result of the multi-scale partitioning to obtain multi-scale partitioning mapping information; Step S24: construct a cross-scale boundary adaptive transfer mechanism based on the multi-scale partition mapping information, establish a stress tensor continuity constraint model at the multi-scale partition boundary, and obtain cross-scale stress boundary coupling data.

4. The three-dimensional dynamic inversion method of geostress field based on a multi-scale adaptive algorithm according to claim 3 is characterized in that: Step S24 includes the following steps: Step S241: constructing a scale response weight function based on the multi-scale partition mapping information to obtain multi-scale stress weight distribution data; Step S242: performing pairing identification on the boundary grid nodes of each region in the multi-scale partition mapping information to obtain a cross-scale contact boundary node set; Step S243: constructing a cross-scale boundary adaptive transfer mechanism based on the cross-scale contact boundary node set and the multi-scale stress weight distribution data to obtain a stress weight adjustment parameter group; Step S244: constructing a stress tensor continuity constraint model based on the stress weight adjustment parameter group and the preset boundary node tensor state to obtain a multi-scale boundary stress tensor coordination matrix; Step S245: Projecting the multi-scale boundary stress tensor coordination matrix into the boundary data structure of the three-dimensional grid model data of the geostress field to generate cross-scale stress boundary coupling data.

5. The three-dimensional dynamic inversion method of geostress field based on a multi-scale adaptive algorithm according to claim 4 is characterized in that: Step S243 includes the following steps: Step S2431: performing topological adjacency analysis on the cross-scale contact boundary node set to obtain boundary node adjacency structure data; Step S2432: extracting a cross-scale mapping path sequence based on the boundary node adjacency structure data and the multi-scale partition mapping information to obtain a path index dataset; Step S2433: Calculating the initial stress transfer coefficient based on the path index data set and the multi-scale stress weight distribution data to obtain a path transfer coefficient matrix; Step S2434: performing normalization adjustment processing on the path transfer coefficient matrix to obtain normalized weight distribution data; Step S2435: Generate a stress weight adjustment parameter group based on the normalized weight distribution data and the cross-scale contact boundary node set.

6. The three-dimensional dynamic inversion method of geostress field based on a multi-scale adaptive algorithm according to claim 1 is characterized in that: Step S3 includes the following steps: Step S31: performing regional index analysis on the multi-scale partition mapping information to obtain structural configuration data of each scale region; Step S32: setting adaptive initial inversion parameters for each scale region based on the structural configuration data to obtain inversion parameter setting data; Step S33: constructing a multi-scale adaptive inversion objective function based on the inversion parameter setting data and the cross-scale stress boundary coupling data to obtain an inversion objective function model; Step S34: performing a joint inversion solution on the inversion objective function model to obtain preliminary three-dimensional distribution results of the geostress field.

7. The three-dimensional dynamic inversion method of geostress field based on a multi-scale adaptive algorithm according to claim 6, characterized in that: Step S34 includes the following steps: Step S341: performing model dimension expansion processing on the inversion objective function model to obtain inversion dimension configuration data of each scale area; Step S342: setting a joint inversion parameter set based on the inversion dimension configuration data of each scale region to obtain initial joint inversion parameters; Step S343: constructing a joint optimization solution process based on the joint inversion initial parameters to obtain joint inversion process configuration data; Step S344: performing multiple rounds of numerical optimization iterations on the joint inversion process configuration data to obtain a multi-scale stress response numerical result set; Step S345: reconstructing the three-dimensional geostress tensor distribution based on the multi-scale stress response numerical result set to obtain a preliminary three-dimensional geostress field distribution result.

8. The three-dimensional dynamic inversion method of geostress field based on a multi-scale adaptive algorithm according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: performing global error field calculation on the preliminary three-dimensional distribution result of the geostress field to obtain error distribution data; Step S42: performing residual statistics and spatial fitting analysis on each scale region based on the error distribution data to obtain residual distribution information; Step S43: performing anomaly detection and cluster identification on the residual distribution information to obtain local abnormal area identification data; Step S44: extracting a corresponding three-dimensional grid model of the local stress field based on the local abnormal area identification data to obtain local area model data; Step S45: Refining the grid and resetting the boundary conditions of the local area model data to obtain a high-resolution inversion input model; Step S46: performing local intensified inversion based on the high-resolution inversion input model and the multi-scale adaptive inversion objective function to obtain optimized local distribution data of the geostress field; Step S47: performing weighted fusion on the optimized local distribution data of the geostress field and the preliminary three-dimensional distribution result of the geostress field to obtain the optimized three-dimensional dynamic distribution result of the geostress field.

9. The three-dimensional dynamic inversion method of geostress field based on a multi-scale adaptive algorithm according to claim 1, characterized in that: In step S5, the optimized three-dimensional dynamic distribution results of the geostress field are coupled with the actual microseismic response data for comparative analysis to generate a dynamic fit evaluation index, which includes the following steps: The actual microseismic response data is filtered, de-noised and time-windowed to obtain standardized microseismic response data. Based on the standardized microseismic response data, a microseismic response spatial distribution model is constructed to obtain microseismic response field data; The optimized three-dimensional dynamic distribution results of the ground stress field are spatially coupled and compared with the microseismic response field data to obtain the coupling residual distribution results; Multi-scale weighted statistical analysis is performed on the coupling residual distribution results to obtain dynamic fit evaluation indicators.

10. The three-dimensional dynamic inversion method of geostress field based on a multi-scale adaptive algorithm according to claim 1, characterized in that: If the fitting degree is lower than the preset threshold in step S5, the multi-scale weight function and inversion parameters are updated, and steps S2 to S5 are iteratively executed until the convergence condition is met. The final three-dimensional dynamic inversion data of the geostress field is obtained, including: Based on the dynamic fit evaluation index, determine whether it is lower than the preset threshold and obtain the fitting result judgment label; If the fitting result determines that the label is not satisfied, the multi-scale weight function is updated based on the coupling residual distribution result to obtain updated multi-scale weight function data; Adjusting the inverse constraint strategy and the optimization parameter set of the geostress based on the updated multi-scale weight function data, generating an updated inversion parameter configuration, and using the updated inversion parameter configuration to iteratively execute steps S2 to S5; If the fitting result judges that the label is satisfied, the current optimized three-dimensional dynamic distribution result of the geostress field is output as the final three-dimensional dynamic inversion data of the geostress field.

Images