Geological model accuracy evaluation method

Abstract

The invention relates to the technical field of accuracy evaluation methods, in particular to a geologic model accuracy evaluation method, which comprises the following steps: S1, randomly extracting a preset proportion of drill holes from all the drill holes used for modeling as verification holes, the total number of the verification holes being as follows; s2, after the geologic model is generated, corresponding virtual drill holes are extracted at the actual coordinate positions of the verification holes; s3, lithology comparison operation is carried out on each verification hole, wherein a plurality of sampling points are arranged from the hole opening to the hole bottom according to a fixed sampling interval, and lithology types of the verification holes and the virtual drill holes are compared point by point; s4, calculating the single-hole accuracy rate based on the proportion of the single-hole sampling total score to the full score; s5, setting the weight of each verification hole, wherein the weight is composed of the product of the basic weight, the position coefficient and the special stratum coefficient; and S6, calculating the overall accuracy through a weighted summation formula. The accuracy of effectively evaluating the geologic model is realized.

Description

Technical Field

[0001] This method relates to the technical field of accuracy evaluation methods, specifically to a method for evaluating the accuracy of geological models. Background Technology

[0002] With the development of geological technology, two-dimensional or three-dimensional geological modeling techniques have been widely used. These techniques, based on local stratigraphic data obtained through drilling, combine computers, visualization tools, and geological analysis methods to construct digital geological models to represent and reproduce geological bodies and their environments. However, there is currently a lack of methods to quantitatively evaluate the accuracy of geological models, making it impossible to verify the true geological conditions of unexplored areas and effectively assess the accuracy of geological models. This is especially true when dealing with geological complexity and locational specificities, where a reliable evaluation mechanism is lacking.

[0003] Method content The purpose of this method is to overcome the shortcomings of the existing technologies and provide a method for evaluating the accuracy of geological models. This method addresses the current lack of quantitative methods for assessing the accuracy of geological models, the inability to verify the true geological conditions of unexplored areas, and the inability to effectively evaluate the accuracy of geological models, especially when dealing with geological complexity and locational uniqueness, where there is a lack of reliable evaluation mechanisms.

[0004] One embodiment of this method provides a method for evaluating the accuracy of a geological model, including the following steps: S1. Randomly select a preset proportion of boreholes from all boreholes used in modeling as verification boreholes. The total number of verification boreholes is [number missing]. ; S2. After the geological model is generated, extract the corresponding virtual boreholes at the actual coordinates of each verification borehole; S3. Perform lithological comparison operation for each verification hole: set multiple sampling points from the hole mouth to the bottom of the hole at a fixed sampling interval, and compare the lithological type of the verification hole with that of the virtual borehole point by point. If the lithology is consistent, score 1 point; if it is inconsistent, score 0 points. S4. Calculate the single-hole accuracy rate based on the ratio of the total score of a single-hole sample to the full score. ; S5. Set the weight of each verification well. This weight is composed of basic weights. Location coefficient and special stratigraphic coefficients The product of all weights is formed, and the sum of all weights is 1; S6. Calculate the overall accuracy using the weighted summation formula: Overall Accuracy = .

[0005] In one embodiment, the extraction ratio of the verification well in step S1 is 5%.

[0006] In one embodiment, the sampling interval in S3 is set to 0.2 meters.

[0007] In one embodiment, the position coefficient The setting rules are as follows: When the distance between the verification borehole and the outer edge of the geological model is ≤30 meters =1; When the distance between the verification borehole and the outer edge of the geological model is greater than 30 meters... =2.

[0008] In one embodiment, the special stratigraphic coefficient The setting rules are as follows: When there are no caves or isolated rocks at the location of the verification hole =1; When there is a cave or isolated rock at the location of the verification hole. =2.

[0009] In one embodiment, the basic weights To uniformly assign baseline weight values ​​to each verification well.

[0010] In one embodiment, the basic weights Through formula Dynamically determined.

[0011] In one embodiment, step S7 is further included: The model's accuracy level is determined based on the range in which the overall accuracy value falls: When the overall score is greater than or equal to 0 and less than 60, it is judged as inaccurate; When the overall score is greater than or equal to 60 and less than 80, it is judged as basically accurate; When the overall score is greater than or equal to 80 and less than 95, it is considered relatively accurate; When the overall score is greater than or equal to 95 and less than or equal to 100, it is considered accurate.

[0012] In one embodiment, the extraction of the virtual borehole in step S2 is achieved through spatial coordinate mapping of the geological model; Single-hole accuracy in step S4 .

[0013] In one embodiment, the weights in step S5 .

[0014] The geological model accuracy evaluation method provided by the above technical solution has the following beneficial effects: 1. Provide quantifiable evaluation standards, establish an objective and accurate evaluation system through lithological comparison of sampling points and score calculation, and avoid subjective bias.

[0015] 2. Enhance adaptability to geological complexity by introducing special stratigraphic coefficients (such as karst caves or isolated boulders) to ensure that model evaluation remains reliable under complex geological conditions.

[0016] 3. Enhance location sensitivity analysis by using location coefficients (such as the distance to the outer edge of the model) to make the evaluation results more in line with actual engineering needs.

[0017] 4. A weighting mechanism that achieves comprehensive coverage is used to distribute weights through the product of basic weights, location coefficients, and special formation coefficients, ensuring a balanced contribution from all verification wells.

[0018] 5. Simplify model quality assessment by setting accuracy level classifications (such as "accurate" or "inaccurate") to help users quickly understand model reliability.

[0019] 6. Supports dynamic well selection and random well extraction design, improving the flexibility and universality of the evaluation method.

[0020] 7. Ensure the feasibility of the technical solution, with detailed sampling intervals and comparison procedures, so that the method can be easily implemented in a computer system. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of this method or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this method. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0022] Figure 1 This is a schematic diagram of a three-dimensional geological model for this method; Figure 2 This is a statistical chart showing the accuracy of one of the verification wells in this method; Figure 3 This is a diagram showing the weight coefficients of each verification well in this method. Detailed Implementation

[0023] The above detailed description is a specific description of a feasible embodiment of the method. This embodiment is not intended to limit the patent scope of the method. Any equivalent implementation or modification that does not depart from the so-called equivalent implementation or modification of the method should be included in the patent scope of this case.

[0024] Combination Figures 1 to 3 As shown, one embodiment of this method provides a method for evaluating the accuracy of a geological model, including the following steps: S1. Randomly select a preset proportion of boreholes from all boreholes used in modeling as verification boreholes. The total number of verification boreholes is [number missing]. ; S2. After the geological model is generated, extract the corresponding virtual boreholes at the actual coordinates of each verification borehole; S3. Perform lithological comparison operation for each verification hole: set multiple sampling points from the hole mouth to the bottom of the hole at a fixed sampling interval, and compare the lithological type of the verification hole with that of the virtual borehole point by point. If the lithology is consistent, score 1 point; if it is inconsistent, score 0 points. S4. Calculate the single-hole accuracy rate based on the ratio of the total score of a single-hole sample to the full score. ; S5. Set the weight of each verification well. This weight is composed of basic weights. Location coefficient and special stratigraphic coefficients The product of all weights is formed, and the sum of all weights is 1; S6. Calculate the overall accuracy using the weighted summation formula: Overall Accuracy = .

[0025] In this embodiment of the invention, the above steps provide a complete evaluation framework and establish a standardized evaluation process. By defining the sequential steps of verification well extraction, virtual well generation, lithological comparison and weight calculation, the problem of lack of quantitative algorithms is solved, making the geological model evaluation systematic and avoiding errors caused by subjective human judgment, and ensuring that the evaluation results are reproducible and consistent. To improve model adaptability, the weighting mechanism (combining location and stratigraphic factors) can dynamically adjust the contribution of different verification wells. For example, verification wells in edge areas have lower weights, while verification wells in complex strata (such as karst caves) have higher weights, making the assessment more consistent with the geological diversity of actual engineering scenarios. To enhance the feasibility of the technology, detailed steps (such as coordinate mapping to extract virtual holes) provide clear instructions for computer programs, making it easier to integrate into geological modeling software and shortening the development cycle; To ensure comprehensive evaluation, from single-hole comparison to overall weighted calculation, it covers the accuracy verification of both local and global aspects of the model, avoiding evaluation blind spots caused by traditional methods that ignore specific areas (such as the outer edge of the model).

[0026] In one embodiment, the extraction ratio of the verification well in step S1 is 5%.

[0027] In this embodiment of the invention, the extraction ratio of verification wells is explicitly limited to 5%, which balances efficiency and reliability. The 5% ratio ensures that the sample size is sufficient to represent the overall model, while avoiding excessive verification wells that would increase the computational burden and optimize resource utilization. This is particularly suitable for large-scale geological engineering projects. To improve the universality of the method, a pre-set ratio provides a unified standard, making the evaluation method easy to apply quickly to datasets of different sizes (such as small mining areas or large basin models) without the need to readjust parameters; To reduce the impact of randomness, the "uniform distribution selection" rule described in the document, combined with random selection at a 5% ratio, can avoid sample bias and ensure that the evaluation results objectively reflect the overall accuracy of the model.

[0028] In one embodiment, the sampling interval in S3 is set to 0.2 meters.

[0029] In this embodiment of the invention, the sampling interval for lithological correlation is limited to 0.2 meters. This fixed interval parameter enhances the accuracy and operability of the method. To ensure the ability to capture details, the 0.2-meter spacing can densely cover the changes in borehole depth, avoiding the omission of thin-layer lithology (such as shale interlayers) due to sparse sampling, thereby improving the accuracy of single-hole correlation; Standardized operating procedures and preset spacing simplify the implementation steps. Users do not need to dynamically adjust parameters, reducing the risk of human error and facilitating integration with automated tools (such as borehole scanning software). To improve the consistency of results, uniform spacing ensures that the comparative data of different verification wells are comparable. For example, in complex strata (such as karst areas), small spacing can more sensitively detect lithological inconsistencies. Sampling interval is directly related to the lithological correlation process, such as Figure 2 The lithological samples are shown at 0.2-meter intervals.

[0030] In one embodiment, the position coefficient The setting rules are as follows: When the distance between the verification borehole and the outer edge of the geological model is ≤30 meters =1; When the distance between the verification borehole and the outer edge of the geological model is greater than 30 meters... =2.

[0031] In this embodiment of the invention, the introduction of the location coefficient enhances the spatial sensitivity of the evaluation. The influence range of the outermost strata of the model is small, and its weight is reduced, which is more in line with reality. The model structure is dynamically adapted, and the coefficient rules automatically adjust the weights to make the evaluation results more in line with actual geological needs. Simplify parameter configuration, clearly define the threshold (30 meters) to provide users with an intuitive standard, eliminate the need for additional calculations of distance influence, improve the ease of use of the method, and the position coefficient depends on the spatial distribution of the model.

[0032] In one embodiment, the special stratigraphic coefficient The setting rules are as follows: When there are no caves or isolated rocks at the location of the verification hole =1; When there is a cave or isolated rock at the location of the verification hole. =2.

[0033] In this embodiment of the invention, the above coefficients are optimized for geological complexity: highlighting the influence of key strata, with karst caves or isolated boulders as high-risk factors, and doubling their weight to ensure that the assessment results focus more on unstable areas and improve the reliability of the model under complex conditions. To address the issue of geological diversity, the coefficient rules automatically identify special strata (without manual annotation), making the method adaptable to complex environments such as karst or fracture zones, thus overcoming the deficiency of "lack of complex strata processing" in the background technology. To enhance the robustness of the assessment, a binary rule (existence / non-existence) is used to simplify the judgment logic and avoid the ambiguity of coefficients caused by stratigraphic gradual changes. Special stratigraphic coefficients are directly correlated with anomalies in lithological correlation, such as... Figure 3 This shows the lithological differences between the caves and the isolated boulders.

[0034] In one embodiment, the basic weights To uniformly assign baseline weight values ​​to each verification well.

[0035] In this embodiment of the invention, the weight calculation formula is explicitly defined as the product of the basic weight, the location coefficient, and the special stratum coefficient. The design of the weight formula realizes the synergy of multiple factors: By comprehensively adjusting the evaluation focus and integrating location and stratigraphic factors through the product of three coefficients, the bias of a single indicator is avoided, resulting in a more comprehensive overall accuracy. Maintaining computational flexibility, basic weights As a baseline value, it allows for dynamic scaling based on model size, adapting to different application scenarios; Simplify implementation complexity; the product form (unweighted sum or other complex operations) reduces computational resource requirements and facilitates real-time evaluation.

[0036] In one embodiment, the basic weights Through formula Dynamically determined.

[0037] In this embodiment of the invention, the weight normalization mechanism improves the mathematical rigor of the evaluation: To ensure comparability of results, a weight and constraint of 1 eliminate the influence of sample size, allowing for direct comparison of the accuracy of different models. To avoid the risk of weight overflow, normalization prevents individual high-weight holes (such as those with position coefficients) from being affected. and (Time) guides the outcome, ensuring fairness in the evaluation; Supports automated solving, constraints (such as...) ) is the basic weight Provides a basis for calculation, reducing human intervention.

[0038] In one embodiment, step S7 is further included: The model's accuracy level is determined based on the range in which the overall accuracy value falls: When the overall score is greater than or equal to 0 and less than 60, it is judged as inaccurate; When the overall score is greater than or equal to 60 and less than 80, it is judged as basically accurate; When the overall score is greater than or equal to 80 and less than 95, it is considered relatively accurate; When the overall score is greater than or equal to 95 and less than or equal to 100, it is considered accurate.

[0039] In this embodiment of the invention, an overall accuracy calculation formula is defined, and the formula structure is optimized for evaluation output: Achieve weighted average fusion. and Sum of products and divide by Integrating single-well data with weights improves the representativeness of the results; Enhancing the transparency of the method and clarifying the formula (non-black box algorithm) makes it easier for users to verify the calculation process and improves credibility; Simplify data aggregation; division normalization directly outputs the range of 0-100%, requiring no additional conversion.

[0040] In one embodiment, the extraction of the virtual borehole in step S2 is achieved through spatial coordinate mapping of the geological model; Single-hole accuracy in step S4 .

[0041] In this embodiment of the invention, the accuracy rate is classified into levels as follows: ≥0 and <60 is inaccurate; ≥60 and <80 is basically accurate; ≥80 and <95 is relatively accurate; ≥95 and ≤100 is accurate. The rating system provides intuitive decision support and simplifies the interpretation of results. The four-level classification (such as "accurate" or "inaccurate") replaces the original numerical values, helping non-professional users to quickly judge the reliability of the model and solve the problem of lack of evaluation standards. It supports tiered responses, with different action recommendations corresponding to different levels (e.g., "inaccurate" requires model correction), thus optimizing the engineering decision-making process. Enhanced method compatibility; fixed thresholds adapt to most geological scenarios, eliminating the need for dynamic adjustments. The detailed formula ensures the accuracy of single-hole evaluation, standardizes single-hole scoring, and eliminates the influence of borehole depth differences by calculating the ratio of total score to sampling points, making boreholes of different lengths comparable, improving data granularity, and providing a basis for local model corrections (such as low depth). (The area corresponding to the hole needs to be carefully verified), simplifying the implementation. Multiply by 100% to directly output the percentage format, making it easy to integrate into visualization tools.

[0042] In one embodiment, the weights in step S5 .

[0043] In this embodiment of the invention, the basic weights Assign datum importance and position coefficient to all verification wells. Enhance the contribution of the model's edge regions, special stratigraphic coefficients Emphasizing the weight of complex strata such as karst caves / isolated boulders, the overall assessment avoids relying on a single coefficient to dominate the results, ensuring a comprehensive evaluation.

[0044] The above description is only a preferred embodiment of the method and does not limit the patent scope of the method. All equivalent structural transformations made using the paper documents and drawings of the method under the method concept, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the method.

Claims

1. A method for evaluating the accuracy of a geological model, characterized in that, Includes the following steps: S1. Randomly select a preset proportion of boreholes from all boreholes used in modeling as verification boreholes. The total number of verification boreholes is [number missing]. ; S2. After the geological model is generated, extract the corresponding virtual boreholes at the actual coordinates of each verification borehole; S3. Perform lithological comparison operation for each verification hole: set multiple sampling points from the hole mouth to the bottom of the hole at a fixed sampling interval, and compare the lithological type of the verification hole with that of the virtual borehole point by point. If the lithology is consistent, score 1 point; if it is inconsistent, score 0 points. S4. Calculate the single-hole accuracy rate based on the ratio of the total score of a single-hole sample to the full score. ; S5. Set the weight of each verification well. This weight is composed of basic weights. Location coefficient and special stratigraphic coefficients The product of all weights is formed, and the sum of all weights is 1; S6. Calculate the overall accuracy using the weighted summation formula: Overall Accuracy = .

2. The method for evaluating the accuracy of a geological model according to claim 1, characterized in that, The extraction ratio of the verification wells in step S1 is 5%.

3. The method for evaluating the accuracy of a geological model according to claim 1, characterized in that, The sampling interval in S3 is set to 0.2 meters.

4. The method for evaluating the accuracy of a geological model according to claim 2, characterized in that, The position coefficient The setting rules are as follows: When the distance between the verification borehole and the outer edge of the geological model is ≤30 meters =1; When the distance between the verification borehole and the outer edge of the geological model is greater than 30 meters... =2.

5. The method for evaluating the accuracy of a geological model according to claim 2, characterized in that, The special stratigraphic coefficient The setting rules are as follows: When there are no caves or isolated rocks at the location of the verification hole =1; When there is a cave or isolated rock at the location of the verification hole. =2.

6. The method for evaluating the accuracy of a geological model according to claim 5, characterized in that, The basic weights To uniformly assign baseline weight values ​​to each verification well.

7. The method for evaluating the accuracy of a geological model according to claim 5, characterized in that, The basic weights Through formula Dynamically determined.

8. The method for evaluating the accuracy of a geological model according to claim 5, characterized in that, It also includes step S7: The model's accuracy level is determined based on the range in which the overall accuracy value falls: When the overall score is greater than or equal to 0 and less than 60, it is judged as inaccurate; When the overall score is greater than or equal to 60 and less than 80, it is judged as basically accurate; When the overall score is greater than or equal to 80 and less than 95, it is considered relatively accurate; When the overall score is greater than or equal to 95 and less than or equal to 100, it is considered accurate.

9. The method for evaluating the accuracy of a geological model according to claim 5, characterized in that, In step S2, the extraction of the virtual borehole is achieved through spatial coordinate mapping of the geological model; Single-hole accuracy in step S4 .

10. The method for evaluating the accuracy of a geological model according to claim 5, characterized in that, The weights in step S5 .

Images