A geological three-dimensional model source data evaluation method

Abstract

The application discloses a geological three-dimensional model source data evaluation method, relates to the technical field of geological models, and comprises the following steps: acquiring a drilling data point set and grid system parameters, mapping the drilling data points to the grid system, and obtaining grid coverage statistical results; calculating horizontal and vertical distribution indexes according to the grid coverage statistical results, performing distribution rationality evaluation processing, and obtaining a source data space distribution coefficient; based on the source data space distribution coefficient and the grid system parameters, calculating the shortest distance from a control point to a grid system surface to obtain a root mean square error precision report; according to the root mean square error precision report and the grid system parameters, checking whether there is a blank zone in the geological model, calculating a model integrity coefficient, and obtaining a final quality evaluation result. The application provides a geological three-dimensional model source data evaluation method to solve the problem that the existing technology is one-sided in distribution evaluation, ignores the vertical resolution and spatial correlation, and causes inaccurate model blank zone detection.

Description

Technical Field

[0001] This invention relates to the field of geological modeling technology, and in particular to a method for evaluating source data of a three-dimensional geological model. Background Technology

[0002] Existing technologies for source data assessment often employ statistical distribution analysis (such as borehole density calculation), but these methods are limited to two-dimensional planar assessments and have not been extended to three-dimensional space. Traditional algorithms calculate data coverage through grid division, but do not incorporate formation influence parameters, making it difficult to quantify spatial continuity.

[0003] Accuracy verification relies on manual methods, model fitting accuracy assessment requires manual measurement of point-to-surface distances, RMSE calculation is inefficient and does not integrate spatial indexing for acceleration; integrity checks are not automated: blank zone identification relies on grid traversal but is not associated with geological object attribution, CM coefficient calculation requires manual labeling, which easily introduces subjective errors.

[0004] Existing technologies only consider the horizontal direction and ignore vertical resolution and spatial correlation, resulting in inaccurate detection of blank bands in the model.

[0005] The information disclosed in this background section is intended only to enhance the understanding of the overall background of the invention and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a method for evaluating source data of a geological three-dimensional model, in order to solve the problem that the existing technology has a one-sided distribution evaluation, ignores vertical resolution and spatial correlation, and leads to inaccurate detection of blank zones in the model.

[0007] To achieve the above objectives, the present invention provides a method for evaluating source data of a geological three-dimensional model, comprising the following steps:

[0008] S1: Obtain the set of borehole data points and grid system parameters, map the borehole data points to the grid system, and obtain the grid coverage statistics. Each borehole data point contains three-dimensional coordinates (x, y, z) and formation code. The grid system parameters include horizontal grid resolution and vertical dimension range.

[0009] S2: Based on the grid coverage statistics, calculate the horizontal and vertical distribution indices, perform distribution rationality assessment, and obtain the spatial distribution coefficient of the source data;

[0010] S3: Based on the spatial distribution coefficient of the source data and the grid system parameters, the root mean square error accuracy report is obtained by calculating the shortest distance from the control point to the surface of the grid system;

[0011] S4: Based on the root mean square error accuracy report and grid system parameters, check whether there are blank zones in the geological model, calculate the model integrity coefficient, and obtain the final quality assessment result;

[0012] S5: Test and verify the final quality assessment results, and optimize based on the test results.

[0013] In one embodiment of the present invention, S1: Obtain a set of borehole data points and grid system parameters, map the borehole data points to the grid system, and obtain grid coverage statistics. Each borehole data point contains three-dimensional coordinates (x, y, z) and a formation code. The grid system parameters include horizontal grid resolution and vertical dimension range, including:

[0014] Obtain the set of borehole data points and grid system parameters, and determine the spatial range of the modeling area by calculating the coordinate extreme values;

[0015] Within the spatial range of the modeling area, the horizontal range is divided into regular grids according to the horizontal grid resolution, and the vertical range is divided into vertical layers to obtain the divided three-dimensional grid.

[0016] For each borehole point, the horizontal grid index to which it belongs is calculated based on its coordinates and the vertical assignment is determined. The point is then mapped to the corresponding three-dimensional grid cell to obtain the grid after the borehole point is mapped.

[0017] Traverse all 3D mesh cells to count the number of boreholes in each cell, and count the number of horizontally covered mesh cells according to the strata to generate mesh coverage statistics results.

[0018] In one embodiment of the present invention, S2: Based on the grid coverage statistics, calculate the horizontal and vertical distribution indices, perform distribution rationality assessment processing, and obtain the source data spatial distribution coefficient, including:

[0019] Based on the grid coverage statistics, the ratio of the number of grids covering each layer in the horizontal direction to the total number of grids that can be distributed is calculated as the horizontal distribution index, and the ratio of the number of borehole point distribution layers to the total number of layers in the vertical direction is combined as the vertical distribution index.

[0020] The spatial distribution coefficients of the source data are obtained by weighted fusion of the horizontal and vertical distribution indices.

[0021] The spatial distribution coefficients are graded and evaluated based on preset thresholds to determine the rationality of the data distribution in three-dimensional space, thereby obtaining the spatial distribution coefficients of the source data.

[0022] In one embodiment of the present invention, S2: Based on the grid coverage statistics, calculate the horizontal and vertical distribution indices, perform distribution rationality assessment processing, and obtain the source data spatial distribution coefficient, further comprising:

[0023] Based on the configuration of the stratum influence range, three-dimensional radiation simulation is performed to calculate the coverage effect of each borehole point within the horizontal and vertical influence grids, and obtain the spatial continuity index.

[0024] Based on continuity indicators, model integrity prediction processing is performed, and a report on blank risk areas is output to guide supplementary data collection.

[0025] In one embodiment of the present invention, S3: Based on the spatial distribution coefficient of the source data and the grid system parameters, a root mean square error accuracy report is obtained by calculating the shortest distance from the control point to the surface of the grid system, including:

[0026] Based on the control point data of the grid system parameters and the spatial distribution coefficients of the source data, a spatial index is constructed to accelerate distance queries and obtain a set of candidate triangles.

[0027] Based on the candidate triangle set, adaptive subdivision and distance calculation are performed to obtain the minimum distance of each point and summarize the root mean square error to obtain a root mean square error accuracy report.

[0028] In one embodiment of the present invention, S4: Based on the root mean square error accuracy report and grid system parameters, check whether there are blank zones in the geological model, calculate the model integrity coefficient, and obtain the final quality assessment result, including:

[0029] Based on the grid system parameters, grid affiliation checks are performed, and the geological object correlation of each grid cell is traversed to obtain blank zone marking results.

[0030] Based on the blank band marking results, the volume percentage is calculated to obtain the integrity coefficient.

[0031] In one embodiment of the present invention, S5: The final quality assessment result is tested and verified, and optimization is performed based on the test results, including:

[0032] Based on the final quality assessment results, performance testing is performed, and processing time and memory usage are recorded to obtain test results;

[0033] Based on the test results, the algorithm was adjusted to obtain the optimized algorithm parameters.

[0034] On the other hand, an electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the geological three-dimensional model source data evaluation method.

[0035] On the other hand, a non-transitory computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the geological three-dimensional model source data evaluation method.

[0036] On the other hand, a computer program product includes a computer program that, when executed by a processor, implements the geological three-dimensional model source data evaluation method.

[0037] Compared with the prior art, the geological three-dimensional model source data evaluation method of the present invention introduces the stratigraphic influence range parameter, calculates the coverage grid number statistics and uniformity index, realizes the three-dimensional spatial continuity evaluation, improves the accuracy of distribution rationality analysis, and supports multi-stratum comparison;

[0038] By combining grid indexing and triangle subdivision techniques, the calculation of point-to-surface distance is accelerated, and the RMSE and error distribution are automatically output. It is adaptable to large-scale data and has improved accuracy compared to traditional methods.

[0039] Blank zones are automatically marked by grid assignment, CM coefficients are calculated using formulas, integrity verification is performed, human error is eliminated, and the model is ensured to be complete. Attached Figure Description

[0040] Figure 1 This is a flowchart of a geological three-dimensional model source data evaluation method according to an embodiment of the present invention. Detailed Implementation

[0041] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings, but it should be understood that the scope of protection of the present invention is not limited to the specific embodiments.

[0042] Unless otherwise expressly stated, throughout the specification and claims, the term "comprising" or its variations such as "including" or "comprises" shall be understood to include the stated elements or components without excluding other elements or other components.

[0043] like Figure 1 As shown, this invention provides a method for evaluating source data of a geological three-dimensional model according to a preferred embodiment.

[0044] This invention provides a method for evaluating source data of a geological three-dimensional model, comprising the following steps:

[0045] S1: Obtain the set of borehole data points and grid system parameters, map the borehole data points to the grid system, and obtain grid coverage statistics. Each borehole data point contains three-dimensional coordinates (x, y, z) and formation code. The grid system parameters include horizontal grid resolution and vertical dimension range.

[0046] S2: Based on the grid coverage statistics, calculate the horizontal and vertical distribution indices, perform distribution rationality assessment, and obtain the spatial distribution coefficient of the source data;

[0047] S3: Based on the spatial distribution coefficient of the source data and the grid system parameters, the root mean square error accuracy report is obtained by calculating the shortest distance from the control point to the surface of the grid system;

[0048] S4: Based on the root mean square error accuracy report and grid system parameters, check whether there are blank zones in the geological model, calculate the model integrity coefficient, and obtain the final quality assessment result;

[0049] S5: Test and verify the final quality assessment results, and optimize based on the test results.

[0050] In this embodiment of the invention, a full-process assessment of source data quality is achieved through five interconnected steps: Step S1 transforms the raw data into structured grid statistics, providing a foundation for assessment; Step S2 quantifies the rationality of the distribution and identifies data defects; Step S3 assesses the accuracy of the model and locates fitting deviations; Step S4 verifies the integrity of the model to ensure no blank areas; Step S5 verifies the reliability of the method through testing; subjective quality judgments are transformed into objective quantitative indicators, improving assessment efficiency and accuracy; large-scale data (such as 20,313 borehole points in the Liaocheng City model) can be efficiently processed in a normal PC environment, supporting the promotion and application of geological survey informatization standards, providing a standardized tool for the quality assessment of three-dimensional geological models, and significantly improving the reliability and application value of geological modeling.

[0051] like Figure 1 As shown, S1: Obtain the set of borehole data points and grid system parameters, map the borehole data points to the grid system, and obtain the grid coverage statistics. Each borehole data point contains three-dimensional coordinates (x, y, z) and a formation code. The grid system parameters include the horizontal grid resolution and the vertical dimension range, including:

[0052] S11: Obtain the set of borehole data points and grid system parameters, and determine the spatial range of the modeling area by calculating the coordinate extreme values;

[0053] S12: Within the spatial range of the modeling area, the horizontal range is divided into regular grids according to the horizontal grid resolution, and the vertical range is divided into vertical layers to obtain the divided three-dimensional grid.

[0054] S13: For each borehole point, calculate the horizontal grid index and determine the vertical assignment based on its coordinates, and map the point to the corresponding three-dimensional grid cell to obtain the grid after the borehole point is mapped.

[0055] S14: Traverse all 3D mesh cells to count the number of boreholes in each cell, and count the number of horizontally covered mesh cells according to the formation to generate mesh coverage statistics results.

[0056] In this embodiment of the invention, through systematic spatial mapping processing, discrete borehole data is transformed into structured grid coverage statistics, providing quantitative input for subsequent distribution rationality assessment; S11 reads the borehole data point set and grid system parameters, calculates coordinate extreme values ​​(xmin, xmax, ymin, ymax, zmin, zmax), determines the spatial boundary of the modeling area, establishes a three-dimensional spatial reference frame, ensures that the grid division covers all data points, and avoids boundary omissions. In one embodiment of the invention, the Liaocheng City model case uses this step to determine the range (X: 397116m~423178m, Y: -4,044,030m~-4,025,340m, Z: -70.28m~38.51m), providing an accurate basis for grid division;

[0057] S12 divides the modeling area into regular three-dimensional grids based on the horizontal grid resolution and vertical dimension range, realizing spatial discretization and converting continuous space into grid cells, which facilitates data indexing and statistics. The vertical direction can be flexibly layered according to the stratigraphic code or resolution. In one embodiment of the present invention, a 100×100×500 grid is used to balance the calculation accuracy and efficiency; the grid size is adjustable to adapt to different accuracy requirements.

[0058] S13 maps each borehole point to its corresponding 3D grid cell according to its coordinates, establishes the correspondence between points and grid cells through index calculation, and realizes data positioning; combined with the formation code, it ensures that the point data is correctly classified in the vertical direction. In one embodiment of the present invention, the vertical index is assigned according to the formation code (such as "0-2-2-3"); boundary checks are used to avoid index out-of-bounds and improve mapping accuracy.

[0059] S14 iterates through all grid cells, counts the number of boreholes in each cell, and summarizes the number of horizontally covered grid cells according to the formation code; it generates a coverage matrix (e.g., Count[i][j][k]) to visually display the data distribution density. In one embodiment of the present invention, it outputs the number of covered grid cells for each formation (e.g., formation "0-2-2-3" covers 3,563,209 grid cells); it quickly locates sparse data areas through statistical results, providing a basis for data supplementation; the statistical results are directly used to calculate distribution rationality indicators (e.g., S1, S2). In one embodiment of the present invention, the CV value (75.34%) is calculated based on this.

[0060] By linking four sub-steps, a complete transformation from raw data to grid statistics is achieved: S11 establishes the spatial scope, S12 constructs the grid framework, S13 completes point mapping, and S14 generates statistical results, providing standardized and quantitative input data for distribution rationality assessment.

[0061] like Figure 1As shown, S2: Based on the grid coverage statistics, calculate the horizontal and vertical distribution indices, perform distribution rationality assessment, and obtain the spatial distribution coefficients of the source data, including:

[0062] S21: Based on the grid coverage statistics, the ratio of the number of grids covering each layer in the horizontal direction to the total number of grids that can be distributed is calculated as the horizontal distribution index, and the ratio of the number of borehole points distributed in the vertical direction to the total number of layers is combined as the vertical distribution index.

[0063] S22: Weighted fusion of horizontal and vertical distribution indices to obtain the spatial distribution coefficient of the source data;

[0064] S23: Based on a preset threshold, the spatial distribution coefficient is graded and evaluated to determine the rationality of the data distribution in three-dimensional space, and the spatial distribution coefficient of the source data is obtained.

[0065] In this embodiment of the invention, based on the grid coverage statistics generated in step S1, the spatial distribution coefficient of the source data is finally obtained by calculating and weighting the horizontal and vertical distribution indices; distribution quantification transforms the abstract spatial distribution characteristics into quantifiable numerical indicators, eliminating subjective judgment bias; three-dimensional comprehensive evaluation considers the distribution characteristics in both the horizontal and vertical directions to achieve true three-dimensional spatial assessment; hierarchical guidance is provided by conducting hierarchical evaluation through preset thresholds, providing a clear direction for data quality improvement; and efficiency is improved by rapidly calculating based on the grid statistical results.

[0066] The lateral distribution index quantifies the uniformity of data distribution on the horizontal plane by calculating the ratio of the number of cover grids in each stratigraphic layer to the total number of available grids; the cover assessment directly reflects the completeness of data coverage in the horizontal direction; the sparse area identification quickly locates sparse areas of data by using the ratio to guide subsequent data supplementation work; and the stratigraphic correlation supports horizontal distribution comparison between different stratigraphic layers to discover stratigraphic units with abnormal distribution.

[0067] The vertical distribution index is calculated by comparing the number of borehole points distributed in the vertical direction with the total number of layers to assess the continuity of data distribution in the vertical direction. Vertical continuity assessment reflects the sampling density and distribution range of data in the vertical direction. Stratigraphic integrity check identifies strata or layers with missing vertical data. Three-dimensional distribution balance, combined with the lateral distribution index, avoids the assessment bias of "emphasizing horizontal and neglecting vertical".

[0068] The weighted fusion method obtains the spatial distribution coefficient by fusing the horizontal and vertical distribution indices through weighting, generating a comprehensive spatial distribution coefficient; it comprehensively quantifies and generates a single numerical index, facilitating comparison between different models; the weights are adjustable, allowing adjustment of the horizontal and vertical weight ratios according to actual needs to adapt to different evaluation scenarios; and the weighted fusion method eliminates the influence of dimensions, making the results comparable.

[0069] Lateral distribution index: Longitudinal distribution index: Spatial distribution coefficient: ,in, The horizontal distribution coefficient is... The longitudinal distribution coefficient is... α is the spatial distribution coefficient, and β are the weighting coefficients. α+β=1, and the weights are adjusted according to actual needs.

[0070] The tiered evaluation assesses the spatial distribution coefficients based on preset thresholds to determine the rationality of the data distribution in three-dimensional space; it discretizes the continuous spatial distribution coefficients into levels such as "excellent, good, average, and poor" for easier understanding; it provides a clear basis for whether to accept the current data quality or whether supplementary data is needed; and it ensures the consistency and comparability of evaluation results across different projects.

[0071] like Figure 1 As shown, S2: Based on the grid coverage statistics, calculate the horizontal and vertical distribution indices, perform distribution rationality assessment, and obtain the spatial distribution coefficient of the source data. This also includes:

[0072] S24: Based on the configuration of the stratum influence range, perform three-dimensional radiation simulation processing, calculate the coverage effect of each borehole point in the horizontal and vertical influence grids, and obtain the spatial continuity index;

[0073] S25: Based on the continuity index, perform model integrity prediction processing and output a report on the blank risk area to guide the collection of supplementary data.

[0074] In this embodiment of the invention, the S24 three-dimensional radiation simulation processing, based on the configuration of the stratum influence range (horizontal / vertical grid range), effectively solves the "hard boundary" problem of traditional grid statistics. Each borehole point is no longer regarded as isolated data, but as a radiation source with a specific influence radius, which is more in line with the actual influence characteristics of geological data. By simulating the radiation coverage effect, "pseudo-continuous" areas with surface data but insufficient actual influence are identified. Different influence range parameters can be set for different strata. In one embodiment of the invention, the stratum "0-2-2-3" has a horizontal range of 25 grids and a vertical range of 30 grids, while "0-1-1-5" has only 5 grids, reflecting geological differences and upgrading the spatial continuity assessment from a simple "presence / absence of data" to "effective influence range", significantly improving the assessment accuracy. Based on the original CV value of 75.34% in the Liaocheng City model, an additional 15% of areas with insufficient influence were identified.

[0075] The S25 model integrity prediction processing accurately predicts the risk of blank zones. Based on continuity indicators, it performs model integrity prediction processing, realizing a leap from "current status assessment" to "risk warning". By comparing continuity indicators with integrity thresholds, it quantifies the probability of blank zone generation; it accurately outputs the spatial coordinates and risk level of risk areas to guide targeted supplementary exploration; and it automatically prioritizes processing according to the degree of risk to optimize resource allocation.

[0076] The generated report, including the blank risk area report, contains: a list of high-risk areas (coordinate range, risk value, affected strata), a distribution map of medium and low-risk areas, data supplementation suggestions, and an assessment of expected improvement effects.

[0077] Predictive processing improved data supplementation efficiency by more than 60%, avoiding the waste of resources from blind exploration. Guided by risk reports, the city model supplemented borehole data in three key areas, increasing the model integrity coefficient (CM) from 92% to 98%.

[0078] like Figure 1 As shown, S3: Based on the spatial distribution coefficient of the source data and the grid system parameters, the root mean square error accuracy report is obtained by calculating the shortest distance from the control point to the grid system surface, including:

[0079] S31: Based on the control point data of the grid system parameters and the spatial distribution coefficients of the source data, perform spatial index construction processing to accelerate distance query and obtain a set of candidate triangles;

[0080] S32: Based on the candidate triangle set, perform adaptive subdivision and distance calculation to obtain the minimum distance of each point and summarize the RMSE to obtain the root mean square error accuracy report.

[0081] In this embodiment of the invention, the core step in the accuracy assessment of the source data evaluation method for geological 3D models is to generate a root mean square error (RMSE) accuracy report by accurately calculating the shortest distance from control points to the model surface, providing a key quantitative indicator for model quality; transforming model fitting quality into a quantifiable RMSE indicator, eliminating subjective judgment bias; significantly improving the efficiency of large-scale data processing through spatial indexing and adaptive subdivision technology; not only providing an overall accuracy assessment, but also identifying fitting problems in specific regions or strata; and providing a scientific basis for model optimization and data analysis.

[0082] The S31 spatial index construction process features an efficient query acceleration mechanism. Based on the control point data and the spatial distribution coefficients of the source data, it performs spatial index construction to create a 3D mesh index structure. This results in a revolutionary improvement in computational efficiency, with local search replacing global traversal: reducing the computational complexity of point-triangle distance from O(N×M) to nearly O(NlogM), where N is the number of points and M is the number of triangles. In one embodiment, the computation time for 20,313 control points and 4,921,282 triangles is reduced from hours to minutes. Dynamic memory management optimizes memory usage, keeping memory usage stable below 500MB in one embodiment, avoiding memory overflow issues when processing large models. Parallel computing is supported, as the spatial index inherently supports parallelization and can utilize multi-core processors for accelerated computation. Enhanced quality control ensures precise selection of candidate triangles, quickly locating the set of triangles that may contain the nearest point for each control point, avoiding unnecessary distance calculations. Improved boundary handling automatically processes points located on mesh boundaries to ensure index integrity. A fault-tolerant mechanism provides robustness to abnormal point data, preventing index construction failures.

[0083] S32 adaptive subdivision and distance calculation processing ensures accuracy and optimizes details. Based on the candidate triangle set, adaptive subdivision and distance calculation processing are performed to ensure distance calculation accuracy while optimizing computational efficiency. The core value of adaptive subdivision is improved accuracy for large triangles. Triangles larger than the grid cell size are subdivided to avoid distance calculation errors caused by excessively large triangles. In one embodiment of the invention, a subdivision threshold based on the grid size is used to ensure that each subdivided triangle matches the grid cell scale. Optimized allocation of computational resources means that subdivision is performed only when needed, avoiding the waste of computational resources caused by global subdivision. Geometric feature preservation means that the original triangle shape features are maintained during the subdivision process, without introducing geometric features. Introducing new geometric distortion; leveraging the advantages of distance calculation algorithms, including the Moller-Trumbore algorithm and employing an efficient point-triangle intersection detection algorithm to accurately calculate the shortest distance; multiple verification mechanisms, encompassing distance calculations for various cases such as points within triangles, points to edges, and points to vertices, ensuring result accuracy; real-time progress feedback, supporting progress monitoring during large-scale computations and enhancing user experience; RMSE summarization and report generation, not only calculating the overall RMSE but also outputting stratified accuracy indicators by formation; visualization support, generating error distribution maps to intuitively display the model's fitting quality in each region; historical data comparison, supporting comparative analysis of multiple evaluation results to track model optimization effects.

[0084] like Figure 1 As shown, S4: Based on the root mean square error accuracy report and grid system parameters, check for blank zones in the geological model, calculate the model integrity coefficient, and obtain the final quality assessment results, including:

[0085] S41: Based on the grid system parameters, perform grid attribution checks, traverse the geological object associations of each grid cell, and obtain the blank zone marking results;

[0086] S42: Based on the blank band marking results, perform volume percentage calculation to obtain the integrity coefficient.

[0087] In this embodiment of the invention, a complete technical path from grid data to integrity assessment is established, transforming the abstract concept of "model integrity" into a specific numerical indicator (CM coefficient), realizing the automatic identification and marking of blank bands, avoiding the subjectivity and omissions of manual inspection, providing clear quantitative standards for model acceptance, supporting scientific decision-making, and guiding the direction of data supplementation and model optimization through blank band location.

[0088] The S41 grid attribution check and processing mechanism features a precise blank zone identification mechanism. Based on grid system parameters, it performs grid attribution checks, systematically traversing the geological object associations of each grid unit. This includes unit-level checks, using grid units as the basic unit for attribution judgment to ensure sufficiently fine detection granularity; association verification, checking whether each grid unit establishes an association with a specific geological object (strata, rock mass, etc.); boundary handling, properly handling grid boundary situations to avoid misjudgments caused by boundary effects; quality control value, with systematic traversal ensuring all grid units are checked; real-time marking, immediately marking any blank zones for subsequent visualization; and type recording, distinguishing different types of attribution relationships to provide a foundation for detailed analysis.

[0089] The S42 volume proportion calculation process calculates the volume proportion based on the blank zone marking results, generating a model integrity coefficient. Calculations are performed using grid cell volume as a benchmark, which better reflects the spatial characteristics of geological models. The ratio of blank zone volume to total volume visually reflects the degree of integrity. Standardized coefficients ranging from 0-100% are generated for easy comparison between different models.

[0090] Steps S41 and S42 form a complete "detection-evaluation" closed loop. S41 provides accurate blank band labeling results, and S42 performs scientific quantitative evaluation based on the labeling results. Together, they achieve the leap from qualitative judgment to quantitative evaluation.

[0091] like Figure 1 As shown, S5: The final quality assessment results are tested and verified, and optimizations are made based on the test results, including:

[0092] S51: Based on the final quality assessment results, perform performance testing, record processing time and memory usage to obtain test results;

[0093] S52: Based on the test results, the algorithm is adjusted to obtain the optimized algorithm parameters.

[0094] In this embodiment of the invention, a feedback mechanism from evaluation results to method optimization is established. Reliability verification verifies the universality and stability of the method through multi-scenario testing. Performance optimization is based on test data to specifically optimize algorithm parameters and improve method efficiency. Standardization forms a verified standardized parameter system, which promotes the application of the method and establishes an iterative optimization mechanism to support the continuous evolution of the method.

[0095] S51 performance testing and processing employs a multi-dimensional performance evaluation system. Based on the final quality assessment results, comprehensive performance testing is conducted to establish a scientific performance evaluation system. Multi-data type coverage utilizes five different data types, including 3D coal mine models, ore body models, and urban geological models, to verify the universality of the methods. Multi-scale data testing, ranging from 30MB to 100MB files, evaluates the scalability of the methods. Multiple indicators are monitored and recorded synchronously, including key performance metrics such as processing time, memory usage, and CPU utilization.

[0096] The S52 algorithm is adjusted and processed using data-driven parameter optimization. Based on test results, precise algorithm adjustments are made to achieve continuous optimization of method performance. Tolerance parameter optimization dynamically adjusts the distance tolerance according to the model accuracy requirements, balancing accuracy and efficiency. Adaptive grid resolution automatically adjusts the grid division granularity according to the data scale to optimize the use of computing resources. Index structure optimization selects the optimal spatial index structure (KD-Tree or grid index) for different data characteristics.

[0097] Steps S51 and S52 form a complete "test-optimize" iterative cycle. S51 provides comprehensive performance data collection and locates bottlenecks through performance analysis. S52 optimizes parameters and algorithms for the bottlenecks and retests to verify the optimization effect, thus forming a closed loop.

[0098] The foregoing description of specific exemplary embodiments of the invention is for illustrative and explanatory purposes. These descriptions are not intended to limit the invention to the precise forms disclosed, and it will be apparent that many changes and variations can be made in accordance with the foregoing teachings. The exemplary embodiments were chosen and described in order to explain the specific principles of the invention and its practical application, thereby enabling those skilled in the art to implement and utilize various different exemplary embodiments of the invention, as well as various different choices and variations. The scope of the invention is intended to be defined by the claims and their equivalents.

Claims

1. A method for evaluating source data of a geological three-dimensional model, characterized in that, Includes the following steps: S1: Obtain the set of borehole data points and grid system parameters, map the borehole data points to the grid system, and obtain the grid coverage statistics. Each borehole data point contains three-dimensional coordinates (x, y, z) and formation code. The grid system parameters include horizontal grid resolution and vertical dimension range. S2: Based on the grid coverage statistics, calculate the horizontal and vertical distribution indices, perform distribution rationality assessment, and obtain the spatial distribution coefficient of the source data; S3: Based on the spatial distribution coefficient of the source data and the grid system parameters, the root mean square error accuracy report is obtained by calculating the shortest distance from the control point to the surface of the grid system; S4: Based on the root mean square error accuracy report and grid system parameters, check whether there are blank zones in the geological model, calculate the model integrity coefficient, and obtain the final quality assessment result; S5: Test and verify the final quality assessment results, and optimize based on the test results; S2: Based on the grid coverage statistics, calculate the horizontal and vertical distribution indices, perform distribution rationality assessment, and obtain the spatial distribution coefficients of the source data, including: Based on the grid coverage statistics, the ratio of the number of grids covering each layer in the horizontal direction to the total number of grids that can be distributed is calculated as the horizontal distribution index, and the ratio of the number of borehole point distribution layers to the total number of layers in the vertical direction is combined as the vertical distribution index. The spatial distribution coefficients of the source data are obtained by weighted fusion of the horizontal and vertical distribution indices. The spatial distribution coefficients are graded and evaluated based on preset thresholds to determine the rationality of the data distribution in three-dimensional space, thereby obtaining the spatial distribution coefficients of the source data. S3: Based on the spatial distribution coefficient of the source data and the grid system parameters, the root mean square error accuracy report is obtained by calculating the shortest distance from the control point to the grid system surface. This includes: performing spatial index construction processing on the control point data based on the grid system parameters and the spatial distribution coefficient of the source data to accelerate distance query and obtain a set of candidate triangles. Based on the candidate triangle set, adaptive subdivision and distance calculation are performed to obtain the minimum distance of each point and summarize the root mean square error to obtain a root mean square error accuracy report.

2. The geological three-dimensional model source data evaluation method as described in claim 1, characterized in that, S1: Obtain the set of borehole data points and grid system parameters, map the borehole data points to the grid system, and obtain grid coverage statistics. Each borehole data point contains three-dimensional coordinates (x, y, z) and a formation code. The grid system parameters include horizontal grid resolution and vertical dimension range, including: Obtain the set of borehole data points and grid system parameters, and determine the spatial range of the modeling area by calculating the coordinate extreme values; Within the spatial range of the modeling area, the horizontal range is divided into regular grids according to the horizontal grid resolution, and the vertical range is divided into vertical layers to obtain the divided three-dimensional grid. For each borehole point, the horizontal grid index to which it belongs is calculated based on its coordinates and the vertical assignment is determined. The point is then mapped to the corresponding three-dimensional grid cell to obtain the grid after the borehole point is mapped. Traverse all 3D mesh cells to count the number of boreholes in each cell, and count the number of horizontally covered mesh cells according to the strata to generate mesh coverage statistics results.

3. The geological three-dimensional model source data evaluation method as described in claim 1, characterized in that, S2: Based on the grid coverage statistics, calculate the horizontal and vertical distribution indices, perform distribution rationality assessment, and obtain the spatial distribution coefficients of the source data. This also includes: Based on the configuration of the stratum influence range, three-dimensional radiation simulation is performed to calculate the coverage effect of each borehole point within the horizontal and vertical influence grids, and obtain the spatial continuity index. Based on continuity indicators, model integrity prediction processing is performed, and a report on blank risk areas is output to guide supplementary data collection.

4. The geological three-dimensional model source data evaluation method as described in claim 1, characterized in that, S4: Based on the root mean square error accuracy report and grid system parameters, check for blank zones in the geological model, calculate the model integrity coefficient, and obtain the final quality assessment results, including: Based on the grid system parameters, grid affiliation checks are performed, and the geological object correlation of each grid cell is traversed to obtain blank zone marking results. Based on the blank band marking results, the volume percentage is calculated to obtain the integrity coefficient.

5. The geological three-dimensional model source data evaluation method as described in claim 1, characterized in that, S5: The final quality assessment results are tested and verified, and optimizations are made based on the test results, including: Based on the final quality assessment results, performance testing is performed, and processing time and memory usage are recorded to obtain test results; Based on the test results, the algorithm was adjusted to obtain the optimized algorithm parameters.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the geological three-dimensional model source data evaluation method as described in any one of claims 1 to 5.

7. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the geological three-dimensional model source data evaluation method as described in any one of claims 1 to 5.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the geological three-dimensional model source data evaluation method as described in any one of claims 1 to 5.

Images