Earth and rockfill dam material space autocorrelation detection and sampling point optimization method and system based on sampling simulation

Abstract

The invention discloses an earth and rockfill dam material space autocorrelation detection and sampling point optimization method and system based on sampling simulation, and the method comprises the steps: collecting the whole-section surface basic data of the compaction quality of a test bin surface: selecting a construction bin surface which is similar to the working condition needing the compaction quality evaluation as the test bin surface, discrete random sampling points are arranged in a uniform point distribution mode, the compaction quality of the test bin surface is detected, and related basic data are collected; detecting the spatial autocorrelation of the earth and rockfill dam material by using a Monte Carlo method; detecting the significance of the spatial autocorrelation of the earth and rockfill dam material by comparing the difference between the accuracy rates of no statistical misjudgment when the spatial autocorrelation of the earth and rockfill dam material is considered and not considered; and reasonably optimizing discrete random sampling point arrangement by utilizing an aggregation state fuzzy comprehensive evaluation model considering multi-source information fusion. The method provided by the invention is more visual, and can truly reflect the obvious influence of the earth and rockfill dam material space autocorrelation on the characterization precision of the compaction quality.

Description

Technical Field

[0001] The invention belongs to the technical field of water conservancy and hydropower engineering, and in particular relates to a method and system for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation. Background Art

[0002] The compaction quality of earth-rock dam materials is an important factor in determining the structural performance of earth-rock dams and controlling uneven settlement after construction. Under the current specification system, the compaction quality control of roller-compacted earth-rock dams mainly relies on density detection based on discrete random sampling points. However, in actual engineering applications, due to the limited detection cost and its interference with the continuous construction process, the sample capacity available on site is very limited (generally <1%), which essentially belongs to the category of finite discrete small sample sampling detection problems.

[0003] Usually, the dam materials used in the construction bin of the earth-rock dam are taken from the same source, and the rolling construction process is mostly carried out under the same working conditions, so that the earth-rock dam materials will show a certain spatial autocorrelation after compaction. Therefore, in terms of the spatial distribution pattern of compaction quality, they may no longer be independent of each other, but are related, and the closer the distance, the stronger the spatial correlation. In practical applications, the following problems may exist:

[0004] 1. The spatial distribution pattern of discrete random sampling points may have a significant impact on the characterization accuracy of compaction quality, resulting in sampling statistical errors such as "discarding true" and "accepting false", which is not conducive to the acceptance and assurance of compaction quality of earth-rock dam materials.

[0005] 2. There are many indicators for determining the spatial distribution pattern of discrete random sampling points, and because of their different representational meanings, conflicting results may occur.

[0006] Therefore, it is necessary to propose a method and system for spatial autocorrelation detection and sampling point optimization of earth-rock dam materials based on sampling simulation. Summary of the invention

[0007] In view of the shortcomings of the prior art, the present invention proposes a method and system for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation, which are used to test the spatial autocorrelation of earth-rock dam materials, and for sampling schemes in which the spatial autocorrelation is too obvious and the spatial distribution pattern of discrete random sampling points will have a significant impact on the characterization accuracy of the compaction quality. The spatial arrangement of discrete random sampling points is reasonably optimized, so as to effectively ensure the accuracy of the compaction quality acceptance of earth-rock dam materials and reduce unnecessary rework and misjudgment of qualified quality.

[0008] To achieve the above object, the present invention proposes a method for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation, which is particularly characterized in that it comprises the following steps:

[0009] S1) Collecting basic data of the whole section of the compaction quality of the test bin surface: Select a construction bin surface with similar working conditions to the working conditions for which compaction quality assessment is required as the test bin surface, set discrete random sampling points in a uniform distribution method, test the compaction quality of the test bin surface, and collect relevant basic data;

[0010] S2) Using the Monte Carlo method to detect the spatial autocorrelation of earth-rock dam materials: By comparing the difference between the accuracy rates without statistical misjudgment when the spatial autocorrelation of earth-rock dam materials is considered and when the spatial autocorrelation of earth-rock dam materials is not considered, the significance of the spatial autocorrelation of earth-rock dam materials is detected;

[0011] S3) The clustered state fuzzy comprehensive evaluation model considering multi-source information fusion is used to rationally optimize the layout of discrete random sampling points, effectively ensuring the accuracy of the compaction quality acceptance of earth-rock dam materials and reducing unnecessary rework and misjudgment of qualified quality.

[0012] As a further optimization scheme of the present invention, the step S2) is specifically as follows:

[0013] S2.1) Use the Kriging spatial interpolation method to effectively estimate the compaction quality of the entire cross-section of the test bin surface, and obtain a compaction quality spatial distribution complete matrix M containing N detection units;

[0014] S2.2) Taking the minimum compaction quality value in the compaction quality spatial distribution complete matrix M of the test bin surface as a reference, normalize the compaction quality spatial distribution complete matrix M to obtain a normalized compaction quality matrix M0;

[0015] S2.3) sequentially numbering the N detection units of the normalized compaction mass matrix M0;

[0016] S2.4) Set the inspection group, assume that the defect rate of compaction quality in the construction chamber surface is p, and consider that there are T = N × p units in the normalized compaction quality matrix M0 as unqualified units; wherein the specific distribution position of the unqualified units is the corresponding position of the first T units in the normalized compaction quality element value of each inspection unit in the normalized compaction quality matrix M0 from small to large;

[0017] S2.5) When it is assumed that the defect rate p of the compaction quality in the construction bin is less than the acceptable compaction quality acceptance control pass rate q β When the compaction quality acceptance standard is met, it should be judged as qualified in theory and can pass the inspection; when it is assumed that the defect rate p of the compaction quality in the construction bin surface is greater than or equal to the acceptable compaction quality acceptance control qualified rate q β When the compaction quality acceptance standard is reached, it should be judged as unqualified in theory and cannot pass the inspection;

[0018] S2.6) randomly select n discrete random samples from N independent units according to the sample size n determined by the proposed sampling plan;

[0019] S2.7) Use the nearest neighbor index R to judge the spatial aggregation state of the n discrete random samples obtained by random sampling, and select discrete random samples that meet the concentrated distribution characteristics for the spatial autocorrelation test of earth-rock dam materials;

[0020] S2.8) For the situation that should be judged as qualified in theory, when the proportion q of the number of assumed unqualified units drawn in the n discrete random samples to the sample capacity is less than or equal to the acceptable compaction quality acceptance control qualified rate q β When , the evaluation result is judged to be consistent with the expected theoretical result, and no "true rejection" error occurs; for the situation that should be judged as unqualified in theory, when the proportion q of the number of assumed unqualified units drawn in the n discrete random samples to the sample capacity is greater than the acceptable compaction quality acceptance control pass rate q β When , the evaluation result is judged to be consistent with the expected theoretical result, and no "false positive" error occurs;

[0021] S2.9) Repeat steps S2.6) to S2.8) for no less than 10,000 times. The proportion of the number of times the statistical evaluation results are consistent with the expected theoretical results and no sampling misjudgment occurs in the total number of tests is taken as the accuracy rate P when considering the spatial autocorrelation of earth-rock dam materials. g ;

[0022] S2.10) Assume the defect rate p of the internal pressure quality of the construction bin surface in the inspection group from small to large, repeat steps S2.4) to S2.9), and obtain the accuracy rate P considering the spatial autocorrelation of earth-rock dam materials g The judgment curve S g ;

[0023] S2.11) Set a control group, assume that the defect rate of the compaction quality in the construction warehouse surface is p, and consider that there are T = N × p units in the construction warehouse surface as unqualified units; among which, the specific distribution positions of the unqualified units are randomly distributed, and the positions of T unqualified units are randomly determined;

[0024] S2.12) Repeat step S2.6) for no less than 10,000 times, and the proportion P of the number of times when the statistical evaluation results are consistent with the expected theoretical results and no sampling misjudgment occurs in the total number of tests is taken as the accuracy rate P0 when the spatial autocorrelation of earth-rock dam materials is not considered;

[0025] S2.13) Assume the defect rate p of the internal pressure quality of the construction bin surface in the control group from small to large, repeat steps S2.11) to S2.12), and obtain the judgment curve S0 of the accuracy rate P0 without considering the spatial autocorrelation of earth-rock dam materials;

[0026] S2.14) Compare the accuracy of considering and not considering the spatial autocorrelation of earth-rock dam materials in the test group and the experimental group P g and P0 judgment curve S g and S0; when the difference between the two is too large, that is, relative to the same assumed construction bin surface internal pressure quality defect rate p, the accuracy rate P when considering and not considering the spatial autocorrelation of earth-rock dam materials g When the difference between P0 and P0 is greater than the established threshold ΔP, it can be determined that the earth-rock dam material has significant spatial autocorrelation and the spatial distribution pattern of discrete random sampling points will have a significant impact on the characterization accuracy of the compaction quality; when designing a sampling scheme to determine the location of discrete random sampling points, its spatial distribution pattern should be controlled to prevent the discrete random sampling points determined by random sampling from having clustered distribution characteristics, which may lead to misjudgment of the compaction quality acceptance of earth-rock dam materials.

[0027] As a further optimization scheme of the present invention, the spatial interpolation method in step 2.1) can be Kriging method or other spatial interpolation methods, and the selection of the method does not affect the implementation of step 2.1).

[0028] As a further optimization scheme of the present invention, the data normalization method in step 2.2) can select the minimum benchmark normalization method or other data normalization methods, and the selection of the method does not affect the implementation of step 2.2).

[0029] As a further optimization scheme of the present invention, the discrete random sample screening method of concentrated distribution characteristics in step 2.7) can select the nearest index R method or other spatial aggregation state identification methods, and the selection of the method does not affect the implementation of step 2.7).

[0030] As a further optimization scheme of the present invention, in step 2.7), the criterion for the nearest neighbor index R is: if the nearest neighbor index R≤0.5, the n discrete sampling points are concentratedly distributed; otherwise, they are randomly or uniformly distributed. The nearest neighbor index R can be obtained by the following formula:

[0031]

[0032] In the formula, r e is the average distance between n discrete sampling points, r T is the theoretical standard distance considering the regional boundary effect, r iis the distance between the ith discrete sampling point and the nearest discrete sampling point, n is the sample capacity in the optimal sampling scheme, N is the number of independent units to be inspected discretely in the construction warehouse, D is the theoretical density of the average distribution of n discrete sampling points in the construction warehouse, B is the perimeter of the construction warehouse, and A is the area of ​​the construction warehouse.

[0033] As a further optimization scheme of the present invention, the step S3) is specifically as follows:

[0034] S3.1) For the case where the earth-rock dam material has significant spatial autocorrelation, the nearest neighbor index R and the agglomeration index PAI of the discrete random sampling points extracted in step S2.6) are calculated respectively; wherein the nearest neighbor index R is used to judge the spatial distribution pattern according to the criteria described in step S2.7); the agglomeration index PAI is judged as follows: if the agglomeration index PAI>1, the n discrete sampling points are concentratedly distributed; otherwise, they are randomly or uniformly distributed.

[0035] S3.2) Construct the fuzzy set of evaluation criteria S = {R, PAI} and the support set of uncertainty degree W = {w1, w2, w3}; where w1, w2, w3 represent concentrated distribution, uncertainty, random or uniform distribution state respectively, and w1+w2+w3=1;

[0036] S3.3) According to the definition of different evaluation criteria, the basic credibility function m(C) can be determined by using a semi-trapezoidal distribution type membership function; wherein, the corresponding parameter values ​​of the inflection points a, b, c, d of the semi-trapezoidal distribution type membership function can be determined on the basis of the control value of the evaluation criteria with a certain variation margin (0%, 10%, 20%, 30%);

[0037] S3.4) Effectively integrate multi-source evaluation information;

[0038] S3.5) After information fusion, according to the principle of maximum basic credibility distribution function, the spatial distribution state with the maximum basic credibility distribution function is taken as the final evaluation result;

[0039] S3.6) When the evaluation result of the fuzzy comprehensive evaluation model of the aggregation state considering the fusion of multi-source information is concentratedly distributed or uncertain, it is necessary to re-extract discrete random sampling points according to step S2.6) until the evaluation result obtained by the fuzzy comprehensive evaluation model of the aggregation state considering the fusion of multi-source information is random or uniformly distributed; at this time, the obtained sampling points after reasonable optimization can be used for the compaction quality acceptance of earth-rock dam materials.

[0040] As a further optimization scheme of the present invention, in step S3.1), the agglomeration index PAI is specifically calculated by the following formula:

[0041]

[0042] In the formula, r e is the average distance between n discrete sampling points, r V is the variance of the distances between n discrete sampling points.

[0043] As a further optimization scheme of the present invention, in step S3.3), for the index R, which is a "smaller, more significant" index, the following basic credibility function m can be used: A (C); For the PAI indicator, which is a “larger, more significant” indicator, the following basic confidence function m can be used. B (C):

[0044]

[0045]

[0046] Where C is the corresponding indicator evaluation value obtained according to the evaluation criteria.

[0047] As a further optimization solution of the present invention, in step S3.4), the multi-source evaluation information is effectively fused according to the following criteria:

[0048]

[0049] Where m1(C) and m2(C) are the basic reliability distribution functions of evaluation information 1 and 2 for evaluation result C respectively; m1(C1) is the basic reliability distribution function of evaluation information 1 for evaluation result C1; m2(C2) is the basic reliability distribution function of evaluation information 2 for evaluation result C2.

[0050] As a further optimization scheme of the present invention, the spatial distribution pattern evaluation method in steps 3.1) to 3.2) can select the nearest neighbor index R and the clustering index PAI or adopt other spatial aggregation state identification methods to construct a fuzzy set of evaluation criteria, and the selection of the method does not affect the implementation of steps 3.1) to 3.2).

[0051] As a further optimization scheme of the present invention, the determination of the basic credibility function m(C) in step 3.3) can be carried out by using a semi-trapezoidal distribution membership function method or other fuzzy uncertainty allocation methods, and the selection of the method does not affect the implementation of step 3.3).

[0052] As a further optimization scheme of the present invention, the steps S2) to S3) can be implemented by programming with Matlab program.

[0053] The present invention also provides a system for executing the above-mentioned method for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation, which is special in that it includes:

[0054] The data acquisition module is used to select a construction bin surface with similar working conditions to those for which compaction quality assessment is required as a test bin surface, set discrete random sampling points in a uniform distribution method, detect the compaction quality of the test bin surface, and collect relevant basic data;

[0055] The spatial autocorrelation detection module uses the Monte Carlo method to detect the significance of the spatial autocorrelation of earth-rock dam materials by comparing the difference between the accuracy rates without statistical misjudgment when the spatial autocorrelation of earth-rock dam materials is considered and when the spatial autocorrelation of earth-rock dam materials is not considered;

[0056] The sampling point optimization module uses the aggregation state fuzzy comprehensive evaluation model considering multi-source information fusion to reasonably optimize the layout of discrete random sampling points, effectively ensure the accuracy of the compaction quality acceptance of earth-rock dam materials, and reduce unnecessary rework and misjudgment of qualified quality.

[0057] The present invention also provides a computer-readable storage medium on which a computer program is stored. The special feature of the computer program is that when the program is executed by a processor, the steps of the method for spatial autocorrelation detection and sampling point optimization of earth-rock dam materials based on sampling simulation are implemented.

[0058] The present invention also provides a computer device, including a processor and a memory, wherein the memory stores a computer program, and when the processor executes the computer program, the method for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation is implemented.

[0059] The method and system for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation of the present invention have the following advantages and beneficial effects:

[0060] 1. The present invention is guided by the target results, and takes the spatial distribution pattern of discrete random sampling points, which will have a significant impact on the characterization accuracy of the compaction quality and lead to frequent sampling statistical errors such as "discarding the true" and "accepting the false", as the direct basis for determining that the earth-rock dam material has significant spatial autocorrelation, which has a more intuitive guiding role in the acceptance and guarantee of the compaction quality of earth-rock dam materials.

[0061] 2. The present invention can fully integrate the discriminant information of indicators of various spatial distribution patterns, establish a reasonable multi-source information conflict resolution mechanism, and optimize the spatial pattern of test sampling points through fuzzy comprehensive evaluation, so as to obtain more reliable compaction quality acceptance accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 It is an implementation flow chart of the method for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation of the present invention;

[0063] Figure 2It is a uniformly distributed point map of the rolling test sampling points of a detection group in a specific implementation manner;

[0064] Figure 3 It is a cloud diagram of the compaction quality of the whole section of earth-rock dam materials in a specific implementation method;

[0065] Figure 4 is a set of accuracy evaluation curves in a specific implementation method with and without considering the spatial autocorrelation of earth-rock dam materials;

[0066] Figure 5 Construct a graph for the membership function of a fuzzy set in a specific implementation;

[0067] Figure 6 A spatial position distribution diagram of a sampling point obtained by random sampling in a specific implementation manner;

[0068] Figure 7 It is a spatial position distribution diagram of sampling points obtained after optimization in a specific implementation method. DETAILED DESCRIPTION

[0069] The specific implementation modes of the present invention are described below to facilitate those skilled in the art to understand the present invention. Obviously, the present invention is not limited to the scope of the specific implementation modes.

[0070] like Figure 1 As shown, the spatial autocorrelation detection and sampling point optimization method of earth-rock dam materials based on sampling simulation of the present invention includes the following steps:

[0071] S1) Collect basic data of the entire section of the test bin surface compaction quality. Select a construction bin surface with similar working conditions to the one that needs to be evaluated for compaction quality as the test bin surface. Set discrete random sampling points in a uniform distribution method to test the compaction quality of the test bin surface and collect relevant basic data.

[0072] In this embodiment, the construction surface that needs to be tested for compaction quality is located in the weathered material area on the downstream side of a certain earth-rock dam project in southwest my country. According to the principle of similar working conditions, the previous compacted construction surface of the construction surface can be taken as the test surface.

[0073] In this embodiment, 24 sampling points are evenly arranged in the test chamber to collect relevant basic data, and 3 additional sampling points are set to verify the interpolation accuracy of the spatial interpolation method. Figure 2 shown.

[0074] S2) The spatial autocorrelation of earth-rock dam materials is tested using the Monte Carlo method. The significance of the spatial autocorrelation of earth-rock dam materials is tested by comparing the difference in accuracy between considering and ignoring the spatial autocorrelation of earth-rock dam materials without statistical misjudgment.

[0075] Specifically, the specific steps of using the Monte Carlo method to detect the spatial autocorrelation of earth-rock dam materials are as follows:

[0076] S2.1) Use the Kriging spatial interpolation method to effectively estimate the compaction quality of the entire cross-section of the test bin surface, and obtain a complete compaction quality spatial distribution matrix M containing N detection units.

[0077] S2.2) Taking the minimum compaction quality value in the compaction quality spatial distribution complete matrix M of the test bin surface as a reference, the compaction quality spatial distribution complete matrix M is normalized to obtain a normalized compaction quality matrix M0.

[0078] S2.3) Number the N detection units of the normalized compaction mass matrix M0 in sequence.

[0079] S2.4) Set the inspection group, assume that the defect rate of compaction quality in the construction bin surface is p, and consider that there are T = N × p units in the normalized compaction quality matrix M0 as unqualified units. Among them, the specific distribution position of the unqualified units is the corresponding position of the first T units in the normalized compaction quality matrix M0 of the normalized element value of the compaction quality of each inspection unit in the order from small to large.

[0080] S2.5) When it is assumed that the defect rate p of the compaction quality in the construction bin is less than the acceptable compaction quality acceptance control pass rate q β When the compaction quality acceptance standard is met, it should be judged as qualified in theory and can pass the inspection; when it is assumed that the defect rate p of the compaction quality in the construction bin surface is greater than or equal to the acceptable compaction quality acceptance control qualified rate q β When it fails, it should be judged as unqualified according to the compaction quality acceptance standard in theory and cannot pass the inspection.

[0081] S2.6) According to the sample size n determined by the proposed sampling plan, n discrete random samples are randomly selected from N independent units.

[0082] S2.7) Use the nearest neighbor index R to judge the spatial aggregation state of the n discrete random samples obtained by random sampling. Discrete random samples that meet the characteristics of concentrated distribution are selected for the spatial autocorrelation test of earth-rock dam materials. Among them, the judgment criterion of the nearest neighbor index R is: if the nearest neighbor point index R≤0.5, the n discrete sampling points belong to concentrated distribution; otherwise, they belong to random or uniform distribution. The nearest neighbor index R can be obtained by the following formula:

[0083]

[0084] In the formula,

[0085] r e is the average distance between n discrete sampling points,

[0086] r T is the theoretical standard distance considering the regional boundary effect,

[0087] r i is the distance between the ith discrete sampling point and the nearest discrete sampling point,

[0088] n is the sample size in the optimal sampling scheme,

[0089] N is the number of independent inspection units discretely separated from the construction warehouse surface.

[0090] D is the theoretical density of n discrete sampling points evenly distributed within the construction warehouse.

[0091] B is the perimeter of the construction warehouse surface,

[0092] A is the area of ​​the construction warehouse.

[0093] S2.8) For the situation that should be judged as qualified in theory, when the proportion q of the number of assumed unqualified units drawn in the n discrete random samples to the sample capacity is less than or equal to the acceptable compaction quality acceptance control qualified rate q β When , the evaluation result is judged to be consistent with the expected theoretical result, and no "discarding true" error occurs. For the situation that should be judged as unqualified in theory, when the proportion q of the number of assumed unqualified units drawn in the n discrete random samples to the sample capacity is greater than the acceptable compaction quality acceptance control pass rate q β , then it is judged that the evaluation result is consistent with the expected theoretical result, and no "false positive" error occurs.

[0094] S2.9) Repeat steps S2.6) to S2.8) for no less than 10,000 times. The proportion of the number of times the statistical evaluation results are consistent with the expected theoretical results and no sampling misjudgment occurs in the total number of tests is taken as the accuracy rate P when considering the spatial autocorrelation of earth-rock dam materials. g .

[0095] S2.10) Assume the defect rate p of the internal pressure quality of the construction bin surface in the inspection group from small to large, repeat steps S2.4) to S2.9), and obtain the accuracy rate P considering the spatial autocorrelation of earth-rock dam materials g The judgment curve S g .

[0096] S2.11) Set a control group, assume that the defect rate of the compaction quality in the construction warehouse surface is p, and consider that there are T = N × p units in the construction warehouse surface as unqualified units. Among them, the specific distribution position of the unqualified units is randomly distributed, and the positions of T unqualified units are randomly determined.

[0097] S2.12) Repeat step S2.6) for no less than 10,000 times. The proportion P of the number of times when the statistical evaluation results are consistent with the expected theoretical results and no sampling misjudgment occurs to the total number of tests is taken as the accuracy rate P0 when the spatial autocorrelation of earth-rock dam materials is not considered.

[0098] S2.13) assumes the defect rate p of the internal pressure quality of the construction bin surface in the control group from small to large, repeats steps S2.11) to S2.12), and obtains the judgment curve S0 of the accuracy rate P0 without considering the spatial autocorrelation of earth-rock dam materials.

[0099] S2.14) Compare the accuracy of considering and not considering the spatial autocorrelation of earth-rock dam materials in the test group and the experimental group P g and P0 judgment curve S g and S0. When the difference between the two is too large, that is, relative to the defect rate p of the same assumed construction bin surface internal pressure quality, the accuracy P when considering and not considering the spatial autocorrelation of earth-rock dam materials g When the difference between P0 and P0 is greater than the established threshold ΔP (generally 5%), it can be determined that the earth-rock dam material has significant spatial autocorrelation and the spatial distribution pattern of discrete random sampling points will have a significant impact on the characterization accuracy of the compaction quality; when designing the sampling scheme to determine the location of discrete random sampling points, the spatial distribution pattern should be controlled to prevent the discrete random sampling points determined by random sampling from having clustered distribution characteristics, which may lead to misjudgment of the compaction quality acceptance of earth-rock dam materials.

[0100] Preferably, the spatial interpolation method in step 2.1) can be Kriging or other spatial interpolation methods, and the selection of the method does not affect the implementation of step 2.1);

[0101] Preferably, the data normalization method in step 2.2) can be selected from the minimum benchmark normalization method or other data normalization methods, and the selection of the method does not affect the implementation of step 2.2);

[0102] Preferably, the discrete random sample screening method of the concentrated distribution characteristics in step 2.7) can use the nearest neighbor index R method or other spatial aggregation state identification methods, and the selection of the method does not affect the implementation of step 2.7);

[0103] In this embodiment, the size of the construction warehouse surface that needs to be tested for compaction quality is 80m×20m. According to the current specifications, the water filling method is used for compaction quality testing. The test pit size is a cylindrical pit with a diameter and depth of 1m. The acceptable compaction quality acceptance control rate q β =10%. Referring to the test pit size, the entire construction area can be 2 The unit area is divided into N = 1600 independent square grid units, which are numbered sequentially with the length direction as the main direction.

[0104] In this embodiment, the Kriging method is used to effectively estimate the full-section compaction quality of the test bin surface. The full-section compaction quality cloud map of the earth-rock dam material is as follows: Figure 3 As shown in the figure, after cross-precision verification, the interpolation error accuracy indicators are mean absolute error MAE = 5.28% and root mean square error RMSE = 6.68%, which can meet the engineering needs of full-section compaction quality estimation.

[0105] In this embodiment, considering the acceptable compaction quality acceptance control pass rate q β It is generally small. For situations that should be judged as qualified in theory but may cause "discarding true" errors, it is assumed that the interval of the defect rate p of the compaction quality of the construction bin surface should be appropriately increased and can be set to 1.25%; for situations that should be judged as qualified in theory but may cause "accepting false" errors, it is assumed that the interval of the defect rate p of the compaction quality of the construction bin surface should be appropriately relaxed and can be set to 5%.

[0106] In this embodiment, a sampling scheme with a sample size of n=22 is adopted, and Matlab software is used to randomly generate numbers of discrete random sampling points to simulate the random sampling process.

[0107] In this embodiment, the accuracy P of the test group and the experimental group with and without considering the spatial autocorrelation of earth-rock dam materials is g and P0 judgment curve S g and S0 Figure 4 As shown. Among them, the maximum difference between the two is 11.1%, which is greater than the established threshold value ΔP=5%. Therefore, it can be determined that the earth-rock dam material has significant spatial autocorrelation and the spatial distribution pattern of discrete random sampling points will have a significant impact on the characterization accuracy of compaction quality. When designing a sampling scheme to determine the location of discrete random sampling points, its spatial distribution pattern should be controlled according to step 3) to prevent the discrete random sampling points determined by random sampling from having clustered distribution characteristics.

[0108] S3) The clustered state fuzzy comprehensive evaluation model considering multi-source information fusion is used to rationally optimize the layout of discrete random sampling points, effectively ensuring the accuracy of the compaction quality acceptance of earth-rock dam materials and reducing unnecessary rework and misjudgment of qualified quality.

[0109] Specifically, the specific steps of using the aggregation state fuzzy comprehensive evaluation model considering multi-source information fusion to reasonably optimize the layout of discrete random sampling points are as follows:

[0110] S3.1) For the case where the earth-rock dam material has significant spatial autocorrelation, the nearest neighbor index R and the agglomeration index PAI of the discrete random sampling points extracted in step S2.6) are calculated respectively. Among them, the nearest neighbor index R is used to judge the spatial distribution pattern according to the criteria described in step S2.7); the agglomeration index PAI is judged by the following criteria: if the agglomeration index PAI>1, then the n discrete sampling points belong to concentrated distribution; otherwise, they belong to random or uniform distribution, which is specifically calculated by the following formula:

[0111]

[0112] In the formula,

[0113] r e is the average distance between n discrete sampling points,

[0114] r V is the variance of the distance between n discrete sampling points;

[0115] S3.2) Construct the fuzzy set of evaluation criteria S = {R, PAI} and the support set of uncertainty degree W = {w1, w2, w3}, where w1, w2, w3 represent concentrated distribution, uncertainty, random or uniform distribution, respectively, and w1+w2+w3=1.

[0116] S3.3) According to the definition of different evaluation criteria, the basic credibility function m(C) can be determined by using the semi-trapezoidal distribution type membership function. Among them, the corresponding parameter values ​​of the inflection points a, b, c, and d of the semi-trapezoidal distribution type membership function can be determined based on the control value of the evaluation standard and given a certain variation margin (0%, 10%, 20%, 30%). For the indicator R, which is a "smaller, more significant" indicator, the following basic credibility function m can be used A (C); For the PAI indicator, which is a “larger, more significant” indicator, the following basic confidence function m can be used. B (C).

[0117]

[0118] m Av2 (C) = 1-m Av1 (C)-m Av3 (C)

[0119]

[0120] m Bv2 (C) = 1-m Bv1 (C)-m Bv3 (C)

[0121] In the formula, C is the corresponding indicator evaluation value obtained according to the evaluation criteria;

[0122] S3.4) Effectively integrate multi-source evaluation information according to the following criteria:

[0123]

[0124] In the formula,

[0125] m1(C) and m2(C) are the basic reliability allocation functions of evaluation information 1 and 2 for evaluation result C respectively; m1(C1) is the basic reliability allocation function of evaluation information 1 for evaluation result C1; m2(C2) is the basic reliability allocation function of evaluation information 2 for evaluation result C2.

[0126] S3.5) After information fusion, according to the principle of maximum basic credibility distribution function, the spatial distribution state with the maximum basic credibility distribution function is taken as the final evaluation result.

[0127] S3.6) When the evaluation result of the fuzzy comprehensive evaluation model of the aggregation state considering multi-source information fusion is concentrated distribution or uncertain, in order to control the risk of statistical misjudgment, it is necessary to re-extract discrete random sampling points according to step S2.6) until the evaluation result obtained by the fuzzy comprehensive evaluation model of the aggregation state considering multi-source information fusion is random or uniformly distributed. At this time, the reasonably optimized sampling points can be used for the compaction quality acceptance of earth-rock dam materials.

[0128] Preferably, the spatial distribution pattern evaluation method in steps 3.1) to 3.2) can use the nearest neighbor index R and the agglomeration index PAI or other spatial aggregation state identification methods to construct a fuzzy set of evaluation criteria, and the selection of the method does not affect the implementation of steps 3.1) to 3.2);

[0129] Preferably, the determination of the basic credibility function m(C) in step 3.3) can be carried out by using a semi-trapezoidal distribution membership function method or other fuzzy uncertainty allocation method, and the selection of the method does not affect the implementation of step 3.3);

[0130] Preferably, the steps S2) to S3) can be implemented by programming with Matlab program.

[0131] In this embodiment, according to the above principle, the spatial distribution of n = 22 discrete random sampling points randomly generated in the construction warehouse surface is as follows: Figure 6 As shown. Among them, the nearest neighbor index R of the spatial distribution pattern is 0.485, and the agglomeration index PAI is 0.8359. According to their respective criteria, the support sets of the uncertainty degree of the nearest neighbor index R and the agglomeration index PAI are W A ={1,0,0},W B={0.18,0.5,0.32}. According to the multi-source information fusion criterion, the support set of uncertainty degree after fusion is W = {0.22,0,0}. According to the principle of maximum basic credibility distribution function, the final evaluation result is concentrated distribution. In order to control the risk of statistical misjudgment, it is necessary to re-extract discrete random sampling points according to step S2.6).

[0132] In this embodiment, after sampling optimization, the spatial distribution of n = 22 discrete random sampling points randomly generated in the construction warehouse surface is as follows: Figure 7 As shown. Among them, the nearest neighbor index R of the spatial distribution pattern is 1.0187, and the agglomeration index PAI is 0.8288. According to their respective criteria, the support sets of the uncertainty degree of the nearest neighbor index R and the agglomeration index PAI are W A ={0,0,1},W B ={0.14,0.5,0.356}. According to the multi-source information fusion criterion, the support set of uncertainty degree after fusion is W = {0,0,0.22}. According to the principle of maximum basic credibility distribution function, the final evaluation result is random or uniformly distributed. At this point, the reasonably optimized sampling points can be used for the compaction quality acceptance of earth-rock dam materials.

[0133] The method for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation is more intuitive than the existing method for judging spatial autocorrelation based on the basis ratio of the semivariogram function, and can truly reflect the significant influence of spatial autocorrelation of earth-rock dam materials on the characterization accuracy of compaction quality. At the same time, by considering the fuzzy comprehensive evaluation model of multi-source information conflict resolution mechanism, a variety of different indicators are used to judge the spatial aggregation state of discrete random samples, which is used as the basis for optimizing the spatial distribution pattern of sampling points, so as to obtain more reliable compaction quality acceptance accuracy. In addition, the implementation method of the present invention is flexible and convenient to apply, has strong applicability, is easy to implement on-site, and can effectively reduce unnecessary rework and misjudgment of qualified quality.

[0134] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0135] The contents not described in detail in the specification of the present invention belong to the prior art known to the professional and technical personnel in this field.

Claims

1. A method for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation, characterized in that: The steps include: S1) Collecting basic data of the whole section of the compaction quality of the test bin surface: Select a construction bin surface with similar working conditions to the working conditions for which compaction quality assessment is required as the test bin surface, set discrete random sampling points in a uniform distribution method, test the compaction quality of the test bin surface, and collect relevant basic data; S2) Using the Monte Carlo method to detect the spatial autocorrelation of earth-rock dam materials: By comparing the difference between the accuracy rates without statistical misjudgment when the spatial autocorrelation of earth-rock dam materials is considered and when the spatial autocorrelation of earth-rock dam materials is not considered, the significance of the spatial autocorrelation of earth-rock dam materials is detected; S3) The arrangement of discrete random sampling points is reasonably optimized using the fuzzy comprehensive evaluation model of aggregation state considering multi-source information fusion.

2. The method for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation according to claim 1 is characterized in that: The step S2) is specifically as follows: S2.1) Use the Kriging spatial interpolation method to effectively estimate the compaction quality of the entire cross-section of the test bin surface, and obtain a compaction quality spatial distribution complete matrix M containing N detection units; S2.2) Taking the minimum compaction quality value in the compaction quality spatial distribution complete matrix M of the test bin surface as a reference, normalize the compaction quality spatial distribution complete matrix M to obtain a normalized compaction quality matrix M0; S2.3) sequentially numbering the N detection units of the normalized compaction mass matrix M0; S2.4) Set the inspection group, assume that the defect rate of compaction quality in the construction chamber surface is p, and consider that there are T = N × p units in the normalized compaction quality matrix M0 as unqualified units; wherein the specific distribution position of the unqualified units is the corresponding position of the first T units in the normalized compaction quality element value of each inspection unit in the normalized compaction quality matrix M0 from small to large; S2.5) When it is assumed that the defect rate p of the compaction quality in the construction bin is less than the acceptable compaction quality acceptance control pass rate q β When the compaction quality acceptance standard is met, it should be judged as qualified in theory and can pass the inspection; when it is assumed that the defect rate p of the compaction quality in the construction bin surface is greater than or equal to the acceptable compaction quality acceptance control qualified rate q β When the compaction quality acceptance standard is reached, it should be judged as unqualified in theory and cannot pass the inspection; S2.6) randomly select n discrete random samples from N independent units according to the sample size n determined by the proposed sampling plan; S2.7) Use the nearest neighbor index R to judge the spatial aggregation state of the n discrete random samples obtained by random sampling, and select discrete random samples that meet the concentrated distribution characteristics for the spatial autocorrelation test of earth-rock dam materials; S2.8) For the situation that should be judged as qualified in theory, when the proportion q of the number of assumed unqualified units drawn in the n discrete random samples to the sample capacity is less than or equal to the acceptable compaction quality acceptance control qualified rate q β When , the evaluation result is judged to be consistent with the expected theoretical result, and no "true rejection" error occurs; for the situation that should be judged as unqualified in theory, when the proportion q of the number of assumed unqualified units drawn in the n discrete random samples to the sample capacity is greater than the acceptable compaction quality acceptance control pass rate q β When , it is determined that the evaluation result is consistent with the expected theoretical result, and no "false positive" error occurs; S2.9) Repeat steps S2.6) to S2.8) for no less than 10,000 times. The proportion of the number of times the statistical evaluation results are consistent with the expected theoretical results and no sampling misjudgment occurs in the total number of tests is taken as the accuracy rate P when considering the spatial autocorrelation of earth-rock dam materials. g ; S2.10) Assume the defect rate p of the internal pressure quality of the construction bin surface in the inspection group from small to large, repeat steps S2.4) to S2.9), and obtain the accuracy rate P considering the spatial autocorrelation of earth-rock dam materials g The judgment curve S g ; S2.11) Set a control group, assume that the defect rate of the compaction quality in the construction warehouse surface is p, and consider that there are T = N × p units in the construction warehouse surface as unqualified units; among which, the specific distribution positions of the unqualified units are randomly distributed, and the positions of T unqualified units are randomly determined; S2.12) Repeat step S2.6) for no less than 10,000 times, and the proportion P of the number of times when the statistical evaluation results are consistent with the expected theoretical results and no sampling misjudgment occurs in the total number of tests is taken as the accuracy rate P0 when the spatial autocorrelation of earth-rock dam materials is not considered; S2.13) Assume the defect rate p of the internal pressure quality of the construction bin surface in the control group from small to large, repeat steps S2.11) to S2.12), and obtain the judgment curve S0 of the accuracy rate P0 without considering the spatial autocorrelation of earth-rock dam materials; S2.14) Compare the accuracy of considering and not considering the spatial autocorrelation of earth-rock dam materials in the test group and the experimental group P g and P0 judgment curve S g and S0; when the difference between the two is too large, that is, relative to the same assumed construction bin surface internal pressure quality defect rate p, the accuracy rate P when considering and not considering the spatial autocorrelation of earth-rock dam materials g When the difference between P0 and P0 is greater than the established threshold ΔP, it can be determined that the earth-rock dam material has significant spatial autocorrelation and the spatial distribution pattern of discrete random sampling points will have a significant impact on the characterization accuracy of the compaction quality; when designing a sampling scheme to determine the location of discrete random sampling points, its spatial distribution pattern should be controlled to prevent the discrete random sampling points determined by random sampling from having clustered distribution characteristics, which may lead to misjudgment of the compaction quality acceptance of earth-rock dam materials.

3. The method for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation according to claim 2 is characterized in that: In step 2.7), the criterion for the nearest neighbor index R is: if the nearest neighbor index R≤0.5, the n discrete sampling points are concentratedly distributed; otherwise, they are randomly or uniformly distributed. The nearest neighbor index R can be obtained by the following formula: In the formula, r e is the average distance between n discrete sampling points, r T is the theoretical standard distance considering the regional boundary effect, r i is the distance between the ith discrete sampling point and the nearest discrete sampling point, n is the sample capacity in the optimal sampling scheme, N is the number of independent units to be inspected discretely in the construction warehouse, D is the theoretical density of the average distribution of n discrete sampling points in the construction warehouse, B is the perimeter of the construction warehouse, and A is the area of ​​the construction warehouse.

4. The method for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation according to claim 1, characterized in that: The step S3) is specifically as follows: S3.1) For the case where the earth-rock dam material has significant spatial autocorrelation, the nearest neighbor index R and the agglomeration index PAI of the discrete random sampling points extracted in step S2.6) are calculated respectively; wherein the nearest neighbor index R is used to judge the spatial distribution pattern according to the criteria described in step S2.7); the agglomeration index PAI is judged as follows: if the agglomeration index PAI>1, the n discrete sampling points are concentratedly distributed; otherwise, they are randomly or uniformly distributed. S3.2) Construct the fuzzy set of evaluation criteria S = {R, PAI} and the support set of uncertainty degree W = {w1, w2, w3}; where w1, w2, w3 represent concentrated distribution, uncertainty, random or uniform distribution state respectively, and w1+w2+w3=1; S3.3) According to the definition of different evaluation criteria, the basic credibility function m(C) can be determined by using a semi-trapezoidal distribution type membership function; wherein, the corresponding parameter values ​​of the inflection points a, b, c, d of the semi-trapezoidal distribution type membership function can be determined on the basis of the control value of the evaluation criteria with a certain variation margin (0%, 10%, 20%, 30%); S3.4) Effectively integrate multi-source evaluation information; S3.5) After information fusion, according to the principle of maximum basic credibility distribution function, the spatial distribution state with the maximum basic credibility distribution function is taken as the final evaluation result; S3.6) When the evaluation result of the fuzzy comprehensive evaluation model of the aggregation state considering the fusion of multi-source information is concentratedly distributed or uncertain, it is necessary to re-extract discrete random sampling points according to step S2.6) until the evaluation result obtained by the fuzzy comprehensive evaluation model of the aggregation state considering the fusion of multi-source information is random or uniformly distributed; at this time, the obtained sampling points after reasonable optimization can be used for the compaction quality acceptance of earth-rock dam materials.

5. The method for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation according to claim 4 is characterized in that: In the step S3.1), the agglomeration index PAI is specifically calculated by the following formula: In the formula, r e is the average distance between n discrete sampling points, r V is the variance of the distances between n discrete sampling points.

6. The method for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation according to claim 4 is characterized in that: In step S3.3), for the indicator R, which is a "smaller, more significant" indicator, the following basic reliability function m can be used: A (C); For the PAI indicator, which is a "larger, more significant" indicator, the following basic reliability function m can be used. B (C): Where C is the corresponding indicator evaluation value obtained according to the evaluation criteria.

7. The method for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation according to claim 1, characterized in that: In step S3.4), the multi-source evaluation information is effectively fused according to the following criteria: Where m1(C) and m2(C) are the basic reliability distribution functions of evaluation information 1 and 2 for evaluation result C respectively; m1(C1) is the basic reliability distribution function of evaluation information 1 for evaluation result C1; m2(C2) is the basic reliability distribution function of evaluation information 2 for evaluation result C2.

8. A system for executing the method for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation according to claim 1, characterized in that: include: The data acquisition module is used to select a construction bin surface with similar working conditions to those for which compaction quality assessment is required as a test bin surface, set discrete random sampling points in a uniform distribution method, detect the compaction quality of the test bin surface, and collect relevant basic data; The spatial autocorrelation detection module uses the Monte Carlo method to detect the significance of the spatial autocorrelation of earth-rock dam materials by comparing the difference between the accuracy rates without statistical misjudgment when the spatial autocorrelation of earth-rock dam materials is considered and when the spatial autocorrelation of earth-rock dam materials is not considered; The sampling point optimization module uses the aggregation state fuzzy comprehensive evaluation model considering multi-source information fusion to reasonably optimize the layout of discrete random sampling points, effectively ensure the accuracy of the compaction quality acceptance of earth-rock dam materials, and reduce unnecessary rework and misjudgment of qualified quality.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation described in any one of claims 1 to 7 are implemented.

10. A computer device, comprising a processor and a memory, wherein the memory stores a computer program, and when the processor executes the computer program, the method for detecting spatial autocorrelation of earth-rock dam materials and optimizing sampling points based on sampling simulation as described in any one of claims 1 to 7 is implemented.

Images