A method for predicting peak strength based on coarse-grained soil triaxial test database
By constructing a triaxial test database for coarse-grained soil and using the random forest method to screen key parameters, a power function formula was established, which solved the problem of the failure to effectively summarize the test data of coarse-grained soil, and realized the accurate assessment of mechanical properties for those lacking test data, thus guiding engineering design and safety evaluation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- DALIAN UNIV OF TECH
- Filing Date
- 2023-08-22
- Publication Date
- 2026-05-12
AI Technical Summary
In the existing technology, triaxial test data of coarse-grained soil cannot be effectively collected and analyzed, which makes it difficult to guarantee the reliability of parameters in engineering design, especially when test data is lacking, the parameters are not accurately obtained.
A triaxial test database for coarse-grained soil was constructed. The importance of physical property parameters was evaluated using the random forest method, key parameters were screened, and a power function formula was established through multiple nonlinear regression to predict peak intensity. Parameter prediction was performed using computer equipment.
It has improved the accuracy and reliability of the assessment of the mechanical properties of coarse-grained soils in the absence of experimental data, and guided the design and safety evaluation of engineering structures.
Smart Images
Figure CN117093955B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of geotechnical engineering testing, and more specifically, relates to a method for predicting peak strength based on a triaxial test database of coarse-grained soil. Background Technology
[0002] Coarse-grained soil is an important material widely used in the construction of earth-rock dams, roadbeds, and other engineering projects. Accurately assessing its mechanical properties is crucial for the safety evaluation of engineering structures. Early research has accumulated a wealth of triaxial test data on coarse-grained soil; however, this data is often limited to studies of specific projects and has not been summarized, analyzed, or expanded for further application, thus failing to explore its potential value and utility. Furthermore, in many projects, especially during the initial design phase or when no coarse-grained soil test data is available, parameters are often obtained through engineering analogies, making reliability difficult to guarantee.
[0003] Therefore, it is urgent to integrate a triaxial test database for coarse-grained soil and to establish a prediction method for key mechanical parameters of coarse-grained soil using big data analysis and artificial intelligence technology, so as to guide the preliminary design of engineering structures as accurately as possible in the absence of test data. Summary of the Invention
[0004] The purpose of this invention is to provide a method for predicting peak strength based on a triaxial test database of coarse-grained soil, providing a relatively scientific and reasonable method for determining material mechanical parameters for the design and safety evaluation of engineering structures involving coarse-grained soil.
[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:
[0006] A method for predicting peak strength based on a coarse-grained soil triaxial test database, which predicts peak strength using physical property parameters of coarse-grained soil, includes the following steps:
[0007] S1. Construct a triaxial test database for coarse-grained soil: Collect test information on coarse-grained soil under conventional triaxial consolidated drained shear test conditions, including the source of coarse-grained soil, physical property parameters and test results, to form a database containing coarse-grained soil with different physical properties and test results.
[0008] The source of the coarse-grained soil includes the project name and geographical location.
[0009] The physical properties of the coarse-grained soil include the properties of the parent rock and the saturated uniaxial compressive strength σ. r Specific gravity G s Dry density ρ d The parameters include void ratio e0, particle shape regularity index ρ, test confining pressure σ3, and gradation parameters; gradation parameters include the coefficient of uniformity C. u curvature coefficient C c Maximum particle size D maxThe particle size values D corresponding to the percentages of soil particles smaller than a certain size on the gradation curve are 80%, 60%, 50%, 30%, and 10%, respectively. 80 D 60 D 50 D 30 and D 10 .
[0010] The test results include stress-strain curves and peak strength values from triaxial consolidated drained shear tests. Peak strength is the maximum stress value in the stress-strain curve.
[0011] S2. Variable Selection: Select physical property parameters of coarse-grained soil as independent variables from the coarse-grained soil triaxial test database constructed in step S1, and select the peak strength q of coarse-grained soil. peak As a dependent variable.
[0012] S3. Ranking of Independent Variable Importance: The importance of independent variables is evaluated (relative importance is determined) and ranked using the random forest method. This invention, through research, ranks the importance of these variables as σ3, σ... r 、e0、ρ、D 80 C u D 30 D 10 C c D 60 D 50 D max G s Among them, the earlier the independent variable is, the more important it is.
[0013] S4. Select key independent variables: Rank the independent variables according to their importance, and select the top few independent variables with a cumulative relative importance exceeding 95% as key independent variables, namely σ3, σ... r 、e0、ρ、D 80 C u .
[0014] S5. Establishing an empirical formula: This invention uses multivariate nonlinear regression analysis on key independent and dependent variables to establish an empirical formula in the form of a power function: q peak =1.702σ3 0.814 σ r 0.097 e0 -0.214 ρ -0.106 D 80 0.043 C u 0.004 Coefficient of determination R 2 It is 0.96.
[0015] S6. Predict peak strength: The physical property parameters (σ3, σ...) of the coarse-grained soil to be tested... r 、e0、ρ、D 80 C u Substituting these values into the formula in step S5, the predicted peak strength can be directly determined. Based on design and calculation requirements, the peak strength is further converted into shear strength parameters of coarse-grained soil (such as the internal friction angle). wait).
[0016] Furthermore, the random forest method in step S3 is a machine learning method, specifically: first, it determines the contribution of each independent variable to each decision tree in the random forest; second, it calculates the average contribution; and finally, it compares the contribution of variables and calculates a weighted normalized relative importance ranking, wherein the contribution is calculated based on the Gini index.
[0017] A device for predicting peak strength based on a triaxial test database of coarse-grained soil, specifically a computer device, mainly comprising a data acquisition module, a memory, and a processor, has the following features:
[0018] The acquisition module is used to collect and construct a triaxial test database for coarse-grained soil;
[0019] The memory is used to store the triaxial test database of coarse-grained soil and the computer program;
[0020] The processor is used to execute the computer program stored in the memory. When the computer program is executed, it can predict the peak intensity using the physical property parameters of coarse-grained soil.
[0021] The beneficial effects of this invention are:
[0022] (1) The importance of numerous physical property parameters of coarse-grained soil was ranked, and the key parameters affecting its strength were identified.
[0023] (2) An empirical relationship between strength parameters and key physical property parameters was proposed, and a method for predicting the peak strength of coarse-grained soil was established. This method has important guiding significance and reference value for the evaluation of the mechanical properties of coarse-grained soil that lacks experimental data. Attached Figure Description
[0024] Figure 1 This is a flowchart of a method for predicting peak strength based on a triaxial test database of coarse-grained soil according to the present invention. Detailed Implementation
[0025] The technical solution of the present invention will now be described in detail with reference to embodiments thereof:
[0026] A method for predicting peak strength based on a coarse-grained soil triaxial test database, which predicts peak strength using physical property parameters of coarse-grained soil, includes the following steps:
[0027] S1. Construct a triaxial test database for coarse-grained soil: Collect test information on coarse-grained soil under conventional triaxial consolidated drained shear test conditions, including the project name and geographical location of the coarse-grained soil; the parent rock properties of the coarse-grained soil; and the saturated uniaxial compressive strength σ. r Specific gravity G s Dry density ρ d The parameters include void ratio e0, particle shape regularity index ρ, test confining pressure σ3, and gradation parameters; gradation parameters include the coefficient of uniformity C. u curvature coefficient C c Maximum particle size D max The particle size values D corresponding to the percentages of soil particles smaller than a certain size on the gradation curve are 80%, 60%, 50%, 30%, and 10%, respectively. 80 D 60 D 50 D 30 and D 10 The stress-strain curves and peak strength values of the triaxial consolidated drained shear test of coarse-grained soil are shown. The peak strength is the maximum value of the stress in the stress-strain curve.
[0028] S2. Variable Selection: Select physical property parameters of coarse-grained soil as independent variables from the coarse-grained soil triaxial test database constructed in step S1, and select the peak strength q of coarse-grained soil. peak As a dependent variable.
[0029] S3. Ranking of Independent Variable Importance: The importance of the independent variables is evaluated (relative importance is determined) and ranked using the random forest method. In this embodiment, the importance is ranked as follows: σ3, σ... r 、e0、ρ、D 80 C u D 30 D 10 C c D 60 D 50 D max G s .
[0030] S4. Select key independent variables: Rank the independent variables according to their importance, and select the top few independent variables with a cumulative relative importance exceeding 95% as key independent variables, namely σ3, σ... r 、e0、ρ、D 80 C u .
[0031] S5. Establish an empirical formula: Perform multiple nonlinear regression analysis on the key independent and dependent variables to establish an empirical formula in the form of a power function, q. peak =1.702σ3 0.814 σ r 0.097 e0 -0.214 ρ -0.106 D 80 0.043 C u 0.004 Coefficient of determination R 2 It is 0.96.
[0032] S6. Predicting Peak Strength: Taking the literature "Experimental Study on the Influence of Gradation on Strength and Deformation of Coarse-grained Soil" [Ling Hua, Fu Hua, Han Huaqiang. Experimental Study on the Influence of Gradation on Strength and Deformation of Coarse-grained Soil [J]. Chinese Journal of Geotechnical Engineering, 2017, 39(S1):12-16] as an example, the key physical property parameters (σ3, σ) of the coarse-grained soil to be tested are... r 、e0、ρ、D 80 C u Substituting the values into the formula in step S5, the predicted values of the peak intensity are obtained (see Table 1). The predicted values obtained by the empirical formula of this invention are compared with the experimental values clearly given in the literature. The relative error of this prediction method is less than 5%, and the accuracy is good.
[0033] In summary, the method for predicting the peak strength of coarse-grained soil proposed in this invention can provide important guidance and reference value for accurately evaluating the mechanical properties of coarse-grained soil in the absence of experimental data.
[0034] Table 1
[0035]
[0036] The above description, in conjunction with specific preferred technical solutions, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.
Claims
1. A method for predicting peak strength based on a triaxial test database of coarse-grained soil, characterized in that, The method can predict peak intensity using physical property parameters of coarse-grained soil; it includes the following steps: S1. Construct a triaxial test database for coarse-grained soils: Collect test information on coarse-grained soils under conventional triaxial consolidated drained shear test conditions, including the source of the coarse-grained soils, physical property parameters, and test results, forming a database containing coarse-grained soils with different physical properties and test results; specifically: The source of the coarse-grained soil includes the project name and geographical location; The physical properties of the coarse-grained soil include the properties of the parent rock and the saturated uniaxial compressive strength σ. r Specific gravity G s Dry density ρ d The parameters include void ratio e0, particle shape regularity index ρ, test confining pressure σ3, and gradation parameters; among which, the gradation parameters include the non-uniformity coefficient C. u curvature coefficient C c Maximum particle size D max The particle size values D corresponding to the percentages of soil particles smaller than a certain size on the gradation curve are 80%, 60%, 50%, 30%, and 10%, respectively. 80 D 60 D 50 D 30 and D 10 ; The test results include stress-strain curves and peak strength values from triaxial consolidated drained shear tests. S2. Variable Selection: Select physical property parameters of coarse-grained soil as independent variables from the coarse-grained soil triaxial test database constructed in step S1, and select the peak strength q of coarse-grained soil. peak As a dependent variable; S3. Ranking of Independent Variable Importance: The importance of the independent variables was evaluated and ranked using the random forest method. The ranking results are σ3, σ... r 、e0、ρ、D 80 C u D 30 D 10 C c D 60 D 50 D max G s The further back in the sequence, the more important it becomes; S4. Select key independent variables: Rank the independent variables according to their importance, and select the top few independent variables with a cumulative relative importance exceeding 95% as key independent variables, namely σ3, σ... r 、e0、ρ、D 80 C u ; S5. Establishing an empirical formula: By performing multivariate nonlinear regression analysis on the key independent and dependent variables, an empirical formula in the form of a power function is established as follows: Coefficient of determination R 2 It is 0.96; S6. Predict peak strength: The physical property parameters (σ3, σ...) of the coarse-grained soil to be tested... r 、e0、ρ、D 80 C u Substituting these values into the formula in step S5, the predicted value of the peak intensity can be directly determined.
2. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the method for predicting peak strength based on a coarse-grained soil triaxial test database as described in claim 1.