Dam foundation engineering rock mass classification method, system and device based on firefly gradient lifting tree intelligent algorithm and medium

Abstract

The invention provides a dam foundation engineering rock mass classification method, system and device based on a firefly gradient lifting tree intelligent algorithm and a medium, and the method comprises the following steps: S1, determining a dam foundation engineering rock mass evaluation criterion according to four basic features of an engineering rock mass; s2, according to dam foundation engineering rock mass classification criteria, dam foundation rock mass engineering quality evaluation indexes are determined; s3, the dam foundation rock mass engineering quality evaluation indexes are imported into the firefly optimization gradient boosting tree model, and a rock mass classification model is obtained through training; s4, judging a rock mass structure, a rock mass hydraulic structure and basic mechanical properties of the rock mass through the rock mass classification model; and S5, comprehensively obtaining a test set dam foundation engineering rock mass classification result. According to the method, the intelligent algorithm is applied to dam foundation rock mass quality classification, qualitative analysis and quantitative analysis are combined, the scientificity and accuracy of rock mass classification can be improved, and the influence of human subjective factors is reduced. According to the method, the selected parameters aim at the rock mass characteristics, and the rock mass quality classification has better accuracy.

Description

Technical Field

[0001] The invention belongs to the technical field of hydraulic and hydroelectric dam foundation rock mass engineering, and particularly relates to a classification method, system, device and medium for dam foundation engineering rock mass based on a firefly gradient boosting tree intelligent algorithm. Background Art

[0002] Since the development of rock mass engineering problems to date, due to the complex geological characteristics of engineering rock masses, limited in-situ test data and monitoring data, the development of discontinuous surfaces in rock mass structures, the existence of bedding, joints, fracture zones, etc., and the fact that rock mass engineering is located in a certain geological and physical environment, the influence of groundwater, in-situ stress, ground temperature, etc. makes the properties of rock masses not only controlled by themselves, making rock masses more complex than any other engineering material, and more intelligent methods are needed to be applied to rock mass engineering analysis.

[0003] A large number of joints, fissures, faults and fracture zones are widely distributed in the dam foundation rock masses in the hydropower development area. Under the action of external loads, the mechanical environment and properties of the rock masses are changed. The evaluation and classification of rock mass quality can not only objectively reflect the inherent physical and mechanical properties of the rock mass itself, but also play a decisive control role in the selection of the dam foundation surface and the stability analysis of dam foundation engineering. The existing methods for evaluating and classifying rock mass quality are relatively numerous, and different classification methods will result in rock masses with different quality grades. This problem urgently needs to be solved. Summary of the Invention

[0004] The first object of the invention is to provide a classification method for dam foundation engineering rock mass based on a firefly gradient boosting tree intelligent algorithm in order to accurately classify the quality of dam foundation rock mass and improve the accuracy of dam foundation rock mass quality classification by using an intelligent algorithm model.

[0005] To achieve the above object, the following technical solutions are adopted in the invention:

[0006] A classification method for dam foundation engineering rock mass based on a firefly gradient boosting tree intelligent algorithm, characterized in that the method comprises the following steps:

[0007] The method comprises the following steps:

[0008] S1. Determine the evaluation criteria for dam foundation engineering rock mass according to four basic characteristics of engineering rock mass;

[0009] S2. Determine the evaluation index for the quality of dam foundation rock mass engineering according to the classification criteria for dam foundation engineering rock mass determined in step S1;

[0010] S3. Import the evaluation index for the quality of dam foundation rock mass engineering into the firefly-optimized gradient boosting tree model to train and obtain a rock mass classification model;

[0011] S4. Determine the rock mass structure, rock mass hydro-mechanical structure, and basic mechanical properties of the rock mass through the rock mass classification model;

[0012] S5. Comprehensively obtain the classification results of the dam foundation engineering rock mass for the test set.

[0013] While adopting the above technical solutions, the present invention can also adopt or combine the following technical solutions:

[0014] As a preferred technical solution of the present invention: In step S1, the four basic characteristics are engineering geological rock groups, rock mass structure, occurrence geological environment, and engineering attributes; the evaluation criteria for dam foundation engineering rock masses can be divided into primary structures, secondary structures, structural geometric parameters, structural mechanical parameters, in-situ stress, groundwater, and engineering influence coefficients.

[0015] As a preferred technical solution of the present invention: In step S2, determine the engineering quality evaluation index of the dam foundation rock mass according to the dam foundation engineering classification criteria. The specific classification indexes include: ratio of hard and soft rock components, modulus ratio, weathering coefficient, unloading coefficient, cutting rate, overlapping rate, offset, rock mass structure modulus, anisotropy ratio, rock mass integrity coefficient, RQD, rock mass structural plane strength coefficient, rock mass deformation modulus coefficient, ultimate compressive strength of the rock mass, quasi-rock mass strength stress ratio, in-situ stress azimuth coefficient, permeability coefficient, seepage direction, engineering action position, and importance degree.

[0016] As a preferred technical solution of the present invention: In step S3, the implementation process of the firefly optimization gradient boosting tree model is as follows:

[0017] First, initialize the weak learner. Among them, c is the function space, and y i is the sample space and the test set:

[0018]

[0019] For m = 1, 2,..., M:

[0020] For i = 1, 2,..., N, calculate the residual:

[0021]

[0022] Take (x i , r im ) as the true value of the sample, and fit a regression tree to r im to obtain the leaf node region R jm of the m-th tree, where j = 1, 2,..., J; among them, J is the number of leaf nodes of the regression tree t;

[0023] For j = 1, 2,..., J, calculate c mj as the best fitting value of the j-th leaf of the m-th tree:

[0024]

[0025] Update the current model:

[0026]

[0027] Finally, obtain the gradient boosting tree model:

[0028]

[0029] Assume there are N fireflies in the hypothesis space, X i (x i1 , x i2 ,..., x iD ) is the i-th firefly and the parameter set to be optimized. Among them, D represents the dimension of the problem and the number of parameters to be optimized. i = 1, 2,..., N, and N represents the size of the firefly population and the possible value range of the optimization parameters. Define β 0 as the attraction when r = 0. The greater the attraction, the better the effect of the parameter. γ is the light absorption coefficient, and r ij is the distance between any two different fireflies X i and X j . Then the attraction between the two fireflies is:

[0030]

[0031] The darker firefly needs to move towards the brighter firefly. Assume X j is brighter than X i , then X i will move towards X j . The moving method is:

[0032] x id (t + 1) = x id (t) + β(r ij )·(x jd (t) - x id (t)) + αε

[0033] where d = 1, 2,..., D, α ∈ [0, 1] is the step factor, ε is a random number in the interval [-0.5, 0.5], and t is the number of iterations;

[0034] Take the obtained model as the penalty function and optimize to obtain the optimal parameters of the model.

[0035] As a preferred technical solution of the present invention: in step S4, by importing the index into the trained rock mass classification model, the rock mass classification model is used to determine the rock mass structure, rock mass hydro-mechanical structure, and basic mechanical properties of the rock mass;

[0036] Among them, the hydraulic structure classification includes: layered type, fault type, fissure type, and solution cavity type;

[0037] The rock mass structure classification includes: integral structure, massive structure, layered structure, mosaic structure, fragmented structure, and dispersed structure;

[0038] The mechanical property classification includes: hard rock, medium-hard rock, and soft rock.

[0039] As a preferred technical solution of the present invention: in step S5, the final rock mass classification results are divided into five categories, namely categories I, II, III, IV, and V.

[0040] The second object of the present invention is to provide a dam foundation engineering rock mass classification system based on the firefly gradient boosting tree intelligent algorithm.

[0041] For this reason, the above object of the present invention is achieved by the following technical solutions:

[0042] A dam foundation engineering rock mass classification system based on the firefly gradient boosting tree intelligent algorithm, characterized in that: the system includes the following modules:

[0043] - Dam foundation engineering rock mass evaluation criterion module, which is used to determine the dam foundation engineering rock mass evaluation criterion according to the four basic characteristics of the engineering rock mass;

[0044] - Dam foundation rock mass engineering quality evaluation index determination module, which is used to determine the dam foundation rock mass engineering quality evaluation index according to the dam foundation engineering rock mass classification criterion determined by the dam foundation engineering rock mass evaluation criterion module;

[0045] - Rock mass classification model acquisition module, which is used to import the dam foundation rock mass engineering quality evaluation index into the firefly-optimized gradient boosting tree model and train to obtain the rock mass classification model;

[0046] - Rock mass property determination module, which is used to determine the rock mass structure, rock mass hydraulic structure, and basic mechanical properties of the rock mass through the rock mass classification model;

[0047] - Rock mass classification result acquisition module, which is used to comprehensively obtain the dam foundation engineering rock mass classification result of the test set.

[0048] The third object of the present invention is to provide a dam foundation engineering rock mass classification device based on the firefly gradient boosting tree intelligent algorithm.

[0049] For this reason, the above object of the present invention is achieved by the following technical solutions:

[0050] A device for classifying rock masses in dam foundation engineering based on a firefly gradient boosting tree intelligent algorithm, characterized in that the device comprises:

[0051] - At least one processor;

[0052] - At least one memory for storing at least one computer program;

[0053] The processor executes the computer program on the memory to implement the steps of the method for classifying rock masses in dam foundation engineering based on the firefly gradient boosting tree intelligent algorithm as described above.

[0054] Another object of the present invention is to provide a computer storage medium.

[0055] To this end, the above object of the present invention is achieved by the following technical solutions:

[0056] A computer storage medium, characterized in that a computer program is stored on the computer storage medium, and the computer program is run by a computer to implement the steps of the method for classifying rock masses in dam foundation engineering based on the firefly gradient boosting tree intelligent algorithm as described above.

[0057] The present invention provides a method, system, device and medium for classifying rock masses in dam foundation engineering based on a firefly gradient boosting tree intelligent algorithm. The present invention applies the intelligent algorithm to the classification of dam foundation rock mass quality, combining qualitative and quantitative methods, which can improve the scientificity and accuracy of rock mass classification. Compared with traditional classification methods, the present invention has the following beneficial effects:

[0058] 1) The present invention adopts an intelligent algorithm, reducing the influence of human subjective factors, and has higher accuracy and efficiency compared with existing intelligent algorithm models such as support vector machines and artificial neural networks.

[0059] 2) The parameters selected by the present invention are targeted at the characteristics of rock masses, and have better accuracy for classifying rock mass quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 is a flowchart of the method for classifying rock masses in dam foundation engineering based on the firefly gradient boosting tree intelligent algorithm provided by the present invention.

[0061] Figure 2 is a diagram of the test data set provided by the present invention.

[0062] Figure 3 is a diagram of the output result provided by the present invention.

[0063] Figure 4 is a flowchart of the operation of the rock mass classification model.

[0064] DETAILED IMPLEMENTATION METHOD

[0065] The present invention will be further described in detail with reference to the accompanying drawings and specific embodiments.

[0066] A method for classifying rock masses of dam foundation engineering based on a firefly gradient boosting tree intelligent algorithm, the method flow chart is as Figure 1 shown, compiled using the Python language, specifically including the following steps:

[0067] S1. According to the four basic characteristics of engineering rock masses, determine the evaluation criteria for rock masses of dam foundation engineering;

[0068] S2. According to the evaluation criteria for rock masses of dam foundation engineering determined in step S1, determine the evaluation indexes for the engineering quality of rock masses of dam foundation;

[0069] S3. Import the evaluation indexes for the engineering quality of rock masses of dam foundation into the firefly-optimized gradient boosting tree model, and train to obtain a rock mass classification model, as Figure 4 shown;

[0070] S4. Use the rock mass classification model to judge the rock mass structure, rock mass hydro-mechanical structure, and basic mechanical properties of the rock mass;

[0071] S5. Comprehensively obtain the classification results of rock masses of dam foundation engineering in the test set, as Figure 3 shown.

[0072] In step S1, the four basic characteristics are engineering geological rock groups, rock mass structure, occurrence geological environment, and engineering attributes; the evaluation criteria for rock masses of dam foundation engineering can be divided into primary structures, secondary structures, structural geometric parameters, structural mechanical parameters, in-situ stress, groundwater, and engineering influence coefficients.

[0073] In step S2, the evaluation indexes for the engineering quality of rock masses of dam foundation are determined by the dam foundation engineering classification criteria. The specific classification indexes include: ratio of hard and soft rock components, modulus ratio, weathering coefficient, unloading coefficient, cutting rate, overlapping rate, offset, rock mass structure modulus, anisotropy ratio, rock mass integrity coefficient, RQD, rock mass structural plane strength coefficient, rock mass deformation modulus coefficient, ultimate compressive strength of rock mass, quasi-rock mass strength stress ratio, in-situ stress azimuth coefficient, permeability coefficient, seepage direction, engineering action position, and importance degree.

[0074] In step S3, the implementation process of the firefly-optimized gradient boosting tree model is as follows:

[0075] First, initialize the weak learner, where c is the function space, and y i is the sample space and the test set:

[0076]

[0077] For m = 1, 2,..., M:

[0078] For \(i = 1, 2, \ldots, N\), calculate the residuals:

[0079]

[0080] Take \((x i , r im ) as the true value of the sample, and fit a regression tree to \(r im to obtain the leaf node regions \(R jm \) of the \(m\)-th tree, where \(j = 1, 2, \ldots, J\); here, \(J\) is the number of leaf nodes of the regression tree \(t\);

[0081] For \(j = 1, 2, \ldots, J\), calculate \(c mj \) as the best fit value for the \(j\)-th leaf of the \(m\)-th tree:

[0082]

[0083] Update the current model:

[0084]

[0085] Finally, obtain the gradient boosting tree model:

[0086]

[0087] Suppose there are \(N\) fireflies in the hypothesis space, \(X i (x i1 , x i2 , \ldots, x iD ) is the \(i\)-th firefly and the parameter set to be optimized, where \(D\) represents the dimension of the problem and the number of parameters to be optimized, \(i = 1, 2, \ldots, N\), and \(N\) represents the firefly population size and the possible value range of the optimization parameters. Define \(\beta 0 as the attractiveness when \(r = 0\). The greater the attractiveness, the better the effect of the parameter. \(\gamma\) is the light absorption coefficient, and \(r ij is the distance between any two different fireflies \(X i and \(X j . Then the attractiveness between two fireflies is:

[0088]

[0089] The darker firefly needs to move towards the brighter firefly. Suppose \(X j is brighter than \(X i , then \(X i will move towards \(X j . The movement method is:

[0090] x id (t + 1)=x id (t)+\beta(r ij)(x jd (t) - x id (t)) + αε

[0091] where d = 1, 2, …, D, α ∈ [0, 1] is the step factor, ε is a random number in the interval [-0.5, 0.5], and t is the number of iterations;

[0092] Take the obtained model as the penalty function and obtain the optimal parameters of the model through optimization.

[0093] In step S4, by importing the indexes into the trained rock mass classification model, the rock mass structure, the rock mass hydro-mechanical structure, and the basic mechanical properties of the rock mass are determined through the rock mass classification model;

[0094] Among them, the hydro-mechanical structure classification includes: layered type, fault type, fissure type, and solution cavity type;

[0095] The rock mass structure classification includes: integral structure, massive structure, layered structure, mosaic structure, fragmented structure, and granular structure;

[0096] The mechanical property classification includes: hard rock, medium-hard rock, and soft rock.

[0097] In step S5, the final rock mass classification results are divided into five categories, namely categories I, II, III, IV, and V.

[0098] The present invention also provides a rock mass classification system for dam foundation engineering based on the firefly gradient boosting tree intelligent algorithm, including the following modules:

[0099] - Dam foundation engineering rock mass evaluation criterion module, which is used to determine the dam foundation engineering rock mass evaluation criterion according to the four basic characteristics of the engineering rock mass;

[0100] - Dam foundation rock mass engineering quality evaluation index determination module, which is used to determine the dam foundation rock mass engineering quality evaluation index according to the dam foundation engineering rock mass classification criterion determined by the dam foundation engineering rock mass evaluation criterion module;

[0101] - Rock mass classification model acquisition module, which is used to import the dam foundation rock mass engineering quality evaluation index into the firefly-optimized gradient boosting tree model and train to obtain the rock mass classification model;

[0102] - Rock mass property determination module, which is used to determine the rock mass structure, the rock mass hydro-mechanical structure, and the basic mechanical properties of the rock mass through the rock mass classification model;

[0103] - Rock mass classification result acquisition module, which is used to comprehensively obtain the dam foundation engineering rock mass classification result of the test set.

[0104] The present invention also provides a device for classifying rock masses in dam foundation engineering based on the intelligent algorithm of firefly gradient boosting tree, including:

[0105] - At least one processor;

[0106] - At least one memory for storing at least one computer program;

[0107] The processor executes the computer program on the memory to implement the steps of the method for classifying rock masses in dam foundation engineering based on the intelligent algorithm of firefly gradient boosting tree as described above.

[0108] The present invention also provides a computer storage medium, on which a computer program is stored, and the computer program is run by a computer to implement the steps of the method for classifying rock masses in dam foundation engineering based on the intelligent algorithm of firefly gradient boosting tree as described above.

[0109] The above specific implementation cases are used to explain the present invention, which are only the preferred embodiments of the present invention, rather than limiting the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and scope of the protection of the claims of the present invention fall within the protection scope of the present invention.

Claims

1. A classification method for rock masses in dam foundation engineering based on the intelligent algorithm of firefly gradient boosting tree, characterized in that: The method comprises the following steps: S1. Determine the evaluation criteria for rock masses in dam foundation engineering according to four basic characteristics of engineering rock masses; S2. Determine the evaluation index for the engineering quality of dam foundation rock masses according to the classification criteria for dam foundation engineering rock masses determined in step S1; S3. Import the evaluation index for the engineering quality of dam foundation rock masses into the firefly-optimized gradient boosting tree model, and train to obtain a rock mass classification model; S4. Determine the rock mass structure, rock mass hydro-mechanical structure, and basic mechanical properties of the rock mass through the rock mass classification model; S5. Comprehensively obtain the classification results of rock masses in the dam foundation engineering of the test set.

2. The classification method for rock masses in dam foundation engineering based on the intelligent algorithm of firefly gradient boosting tree according to claim 1, characterized in that: In step S1, the four basic characteristics are engineering geological rock groups, rock mass structures, occurrence geological environments, and engineering attributes; the evaluation criteria for rock masses in dam foundation engineering can be divided into primary structures, secondary structures, structural geometric parameters, structural mechanical parameters, in-situ stress, groundwater, and engineering influence coefficients.

3. The classification method for rock masses in dam foundation engineering based on the intelligent algorithm of firefly gradient boosting tree according to claim 1, characterized in that: In step S2, the evaluation index for the engineering quality of dam foundation rock masses is determined by the classification criteria for dam foundation engineering, and the specific classification indexes include: ratio of hard and soft rock components, modulus ratio, weathering coefficient, unloading coefficient, cutting rate, overlapping rate, offset, rock mass structure modulus, anisotropy ratio, rock mass integrity coefficient, RQD, rock mass structural plane strength coefficient, rock mass deformation modulus coefficient, ultimate compressive strength of rock mass, quasi-rock mass strength stress ratio, in-situ stress azimuth coefficient, permeability coefficient, seepage direction, engineering action position, and importance degree.

4. The classification method for rock masses in dam foundation engineering based on the intelligent algorithm of firefly gradient boosting tree according to claim 1, characterized in that: In step S3, the implementation process of the firefly-optimized gradient boosting tree model is as follows: First, initialize the weak learner, where c is the function space, and y i is the sample space and the test set: For m = 1, 2, …, M: For i = 1, 2, …, N, calculate the residual: Taking (x i , r im ) as the true value of the sample, fitting a regression tree to r im , obtaining the leaf node region R jm of the m-th tree, where j = 1, 2, …, J; among them, J is the number of leaf nodes of the regression tree t; For j = 1, 2, …, J, compute c mj as the best fit value for the j-th leaf of the m-th tree: Update the current model: Finally, obtain the gradient boosting tree model: Suppose there are N fireflies in space, X i (x i1 , x i2 , …, x iD ) is the i-th firefly and the parameter set to be optimized. Among them, D represents the dimension of the problem and the number of parameters to be optimized. i = 1, 2, …, N, and N represents the size of the firefly population and the possible value range of the optimization parameters. Define β 0 as the attraction when r = 0. The greater the attraction, the better the effect of the parameter. γ is the light absorption coefficient, and r ij is the distance between any two different fireflies X i and X j . Then the attraction between the two fireflies is: The darker fireflies need to move towards the brighter ones, assuming X j brighter than X i , then X i will move towards X j The moving method is as follows: x id (t + 1) = x id (t) + β(r ij )·(x jd (t) - x id (t)) + αε where d = 1, 2, …, D, α ∈ [0, 1] is the step factor, ε is a random number in the interval [-0.5, 0.5], and t is the number of iterations; Take the obtained model as the penalty function, and optimize to obtain the optimal parameters of the model.

5. The classification method for rock masses in dam foundation engineering based on the intelligent algorithm of firefly gradient boosting tree according to claim 1, characterized in that: In step S4, by importing the indexes into the trained rock mass classification model, determine the rock mass structure, rock mass hydro-mechanical structure, and basic mechanical properties of the rock mass through the rock mass classification model; Among them, the hydro-mechanical structure classification includes: layered type, fault type, fissure type, and solution cavity type; The rock mass structure classification includes: integral structure, massive structure, layered structure, mosaic structure, fragmented structure, and dispersed structure; The mechanical property classification includes: hard rock, medium-hard rock, and soft rock.

6. The classification method for rock masses in dam foundation engineering based on the intelligent algorithm of firefly gradient boosting tree according to claim 1, characterized in that: In step S5, the final rock mass classification results are divided into five categories, namely categories I, II, III, IV, and V.

7. A rock mass classification system for dam foundation engineering based on the intelligent algorithm of firefly gradient boosting tree, characterized in that: The system includes the following modules: - A dam foundation engineering rock mass evaluation criterion module, which is used to determine the dam foundation engineering rock mass evaluation criterion according to the four basic characteristics of the engineering rock mass; - A dam foundation rock mass engineering quality evaluation index determination module, which is used to determine the dam foundation rock mass engineering quality evaluation index according to the dam foundation engineering rock mass classification criterion determined by the dam foundation engineering rock mass evaluation criterion module; - A rock mass classification model acquisition module, which is used to import the dam foundation rock mass engineering quality evaluation index into the firefly optimized gradient boosting tree model and train to obtain a rock mass classification model; - A rock mass property determination module, which is used to determine the rock mass structure, rock mass hydro-mechanical structure, and basic mechanical properties of the rock mass through the rock mass classification model; - A rock mass classification result acquisition module, which is used to comprehensively obtain the dam foundation engineering rock mass classification result of the test set.

8. A rock mass classification device for dam foundation engineering based on the intelligent algorithm of firefly gradient boosting tree, characterized in that: The device includes: - At least one processor; - At least one memory for storing at least one computer program; The processor executes the computer program on the memory to implement the steps of the rock mass classification method for dam foundation engineering based on the intelligent algorithm of firefly gradient boosting tree as described in any one of claims 1-6.

9. A computer storage medium, characterized in that: A computer program is stored on the computer storage medium, and the computer program is run by a computer to implement the steps of the rock mass classification method for dam foundation engineering based on the intelligent algorithm of firefly gradient boosting tree as described in any one of claims 1-6.

Images