Explanatable machine learning method for predicting rock shear strength parameters

Abstract

The invention relates to the field of mechanical property testing, in particular to an interpretable machine learning method for predicting rock shear strength parameters. According to the machine learning method, rock shear strength parameters are predicted mainly based on a random forest algorithm and a litsea rotundifolia optimization algorithm, preliminary prediction is performed through the random forest algorithm to obtain a preliminary prediction value, and a search sample boundary is set for the rock sample group number through the litsea rotundifolia optimization algorithm to obtain a rock sample group number; iterative optimization is carried out on two parameters, namely the number of decision trees in the random forest algorithm and the number of features selected by random sampling when each decision tree is constructed, and the rock shear strength parameter is predicted again through the random forest algorithm according to the obtained optimization parameters. According to the method, high-accuracy prediction is achieved when rock shear strength parameters are predicted, and the problems that in the prior art, when rock shear strength is detected through an instrument detection mode, cost is high, and time consumption is high are solved.

Description

Technical Field

[0001] The present invention relates to the field of mechanical property testing, and in particular to an interpretable machine learning method for predicting rock shear strength parameters. Background Art

[0002] Machine learning models are a powerful tool in the field of artificial intelligence. They analyze and learn from large data sets to discover hidden patterns, trends, and correlations, enabling accurate predictions and decisions. These models are not just simple algorithmic expressions, they are complex mathematical and statistical models that can simulate and extend human cognitive abilities. In recent years, machine learning models have been widely recognized in many fields such as computer vision, financial risk assessment, natural language processing, and medical diagnosis, greatly promoting the development of these fields. Similarly, machine learning models have also been widely used in geotechnical and geological engineering, for example, for the prediction of rock compressive strength, rock crushing, mortar compressive strength, rock deformation, slope stability, rock mass classification, and tunnel boring machine performance. Cohesion and friction angle are widely used parameters to describe the shear strength of rock materials, where cohesion is related to the bonding between rock crystals or particles, and friction angle is related to the internal friction of the rock shear surface. Cohesion and friction angle play a vital role in the design and construction of rock slopes, tunnel excavation, and foundations. Therefore, obtaining these shear strength parameters is of great significance for the evaluation of rocks and rock masses.

[0003] In the prior art, the detection of rock shear strength parameters such as cohesion and friction angle is mostly obtained through rock testing, and the commonly used prior art solution is rock triaxial testing. The disadvantage of triaxial testing with the help of testing instruments is that it has strict requirements on test conditions, requires expensive instruments and is time-consuming, so it is difficult to meet the current research and testing on rock-related aspects. Summary of the invention

[0004] The purpose of the present invention is to solve the problems of high cost and long time consumption faced by the prior art in detecting the shear strength of rocks, and to propose a machine learning method for predicting the shear strength parameters of rocks.

[0005] On the one hand, the present invention provides an interpretable machine learning method for predicting rock shear strength parameters, characterized in that it includes the following steps:

[0006] S1. Preliminary prediction of cohesion and friction angle in rock shear strength parameters by using random forest algorithm;

[0007] S2. Set the size of the search boundary of the golden jackal optimization algorithm, and iteratively optimize the parameters that need to be optimized in the random forest algorithm, namely, the number of decision trees and the number of features randomly sampled when building each decision tree;

[0008] S3, bringing the number of the optimized decision trees and the number of features randomly selected when constructing each decision tree into the random forest algorithm, and predicting the cohesion and the friction angle again,

[0009] Get the final prediction result.

[0010] Preferably, the specific steps of the random forest algorithm in step S1 are as follows:

[0011] S11, taking the cohesion and the friction angle in the rock sample as a data set to form a training set required by a random forest algorithm;

[0012] S12, re-sampling the training set of rock parameters using the Bootstrap method, and randomly generating a number of training subsets and regression trees according to the sampling results;

[0013] S13, when each node of the regression tree splits and grows, randomly extracting input parameters as the split subset of the current node;

[0014] S14, the regression tree recursively branches and grows from top to bottom, and the regression tree stops growing after the segmentation termination condition of the preliminary rock shear strength parameter prediction is met, and the regression trees that have completed growth are aggregated to form a random forest, and a prediction set is obtained and the average value is calculated to obtain the final result;

[0015] Preferably, the setting of the search boundary size in step S2 is determined by the number of rock sample groups, the upper limit of the search boundary is specifically the number of sample groups, and the lower limit of the search boundary is the smallest positive integer.

[0016] Further preferably, the number of features randomly sampled when constructing each decision tree in step S2 is determined by the number of input variables, and the specific input variables include P-wave, uniaxial compressive strength, tensile strength and rock density.

[0017] Preferably, the number of decision trees in the random forest algorithm and the number of features randomly sampled when constructing each decision tree are used as the initial population of the golden jackal optimization algorithm for parameter optimization.

[0018] Preferably, the specific steps of the golden jackal optimization algorithm in step S2 are as follows:

[0019] S21. Define the objective function f(X), set the population size N, and the maximum number of iterations Max i ter, leadership rate α, exploration rate β and randomly generate the initial population X1, X2, …, X N ;

[0020] S22, calculate the fitness f(X) of each solution in the initial population i );

[0021] S23, for each solution X in the initial population i If randn()<α, execute the leadership behavior, otherwise execute the exploration behavior, process the search boundary, and ensure that the new solution In the search space;

[0022] S24, evaluate the new solution, calculate And update the calculation results, if Better than f(X i ),

[0023] Use Replace X i .

[0024] Further preferably, the leadership behavior and the exploration behavior in step S23 are specifically as follows:

[0025] Leadership Behavior:

[0026] Where: X i is the solution in the initial population, X j is the number of the initial population with X i Another different solution, α is the leadership rate, and the value range of α is 0 to 1;

[0027] Exploration Behavior:

[0028] Where: X i new is the new solution, β is the exploration rate, and randn() is a random number from a standard normal distribution.

[0029] Compared with the prior art, the present invention has the following beneficial effects:

[0030] The present invention provides an interpretable machine learning method for predicting rock shear strength parameters, and realizes the prediction of rock shear strength parameters by establishing a random forest algorithm whose parameters are optimized by the golden jackal algorithm. Compared with the prior art, the present invention selects the best prediction effect based on the three algorithm models, and then optimizes the model parameters with the help of the golden jackal algorithm, thereby improving the accuracy of rock shear strength prediction and solving the problem that the instrument detection technology in the prior art is high cost and time consuming and cannot meet the needs of some projects to obtain rock performance. The detailed beneficial effects are shown in the experimental results in the embodiments; BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 14 is a flow chart of an interpretable machine learning method for predicting rock shear strength in an embodiment of the present invention.

[0032] Figure 2 It is a model algorithm diagram of the random forest algorithm in an embodiment of the present invention.

[0033] Figure 3 4 is a flow chart of the golden jackal optimization algorithm in an embodiment of the present invention.

[0034] Figure 4 1 is a graph comparing the predicted value and the actual value of the random forest model in an embodiment of the present invention.

[0035] Figure 5 It is a result diagram of the final prediction set in an embodiment of the present invention. DETAILED DESCRIPTION

[0036] The present invention is further described in detail below in conjunction with test examples and specific implementation methods. However, this should not be understood as the scope of the above subject matter of the present invention being limited to the following embodiments, and all technologies realized based on the content of the present invention belong to the scope of the present invention.

[0037] Example 1

[0038] The purpose of the present invention is to overcome the problems of high cost and high time consumption in instrument detection of rock shear strength in the prior art, and to provide an explainable machine learning method for predicting rock shear strength parameters.

[0039] The present invention is mainly based on a random forest algorithm, which is optimized by a golden jackal optimization algorithm to obtain a random forest algorithm with higher prediction accuracy, and thereby predicts rock shear strength parameters.

[0040] The present invention specifically proposes an interpretable machine learning method for predicting rock shear strength parameters, the flow chart is as follows: Figure 1 As shown, it is characterized in that it includes the following steps:

[0041] S1. Preliminary prediction of cohesion and friction angle in rock shear strength parameters is performed through random forest algorithm to obtain preliminary prediction results;

[0042] S2. Set the search boundary of the golden jackal optimization algorithm, and iteratively optimize the number of decision trees in the random forest algorithm and the number of features randomly sampled when building each decision tree;

[0043] S3. According to the number of decision trees in the optimized random forest algorithm and the number of features randomly sampled when constructing each decision tree, the optimized random forest algorithm is used to predict the cohesion and the friction angle to obtain a final prediction result.

[0044] The step S1 mainly uses the Random Forest algorithm (RF) to make a preliminary prediction on the sample data set, and obtains the parameters to be optimized according to the prediction results. The flowchart of the Random Forest algorithm is as follows: Figure 2 shown.

[0045] Specifically, step S1 includes firstly taking the cohesion and the friction angle in the collected 199 groups of rock samples as data sets to form the training set required by the random forest algorithm; secondly, resampling the training set of rock parameters by using the Bootstrap method, and randomly generating several training subsets and regression trees according to the sampling results; when each node of the regression tree splits and grows, randomly extracting input parameters as the split subset of the current node; the regression tree recursively branches and grows from top to bottom, and the regression tree stops growing after the segmentation termination condition of the preliminary rock shear strength parameter prediction is met, and the regression trees that have completed growth are aggregated to form a random forest, and a prediction set is obtained and the average value is calculated to obtain a preliminary prediction result, and the parameters that need to be optimized in the prediction of rock shear strength parameters are obtained according to the preliminary prediction result.

[0046] The setting of the search boundary size in step S2 is specifically determined by the number of rock sample groups. The upper limit of the boundary is specifically the number of sample groups, and the lower limit is the smallest positive integer. In this embodiment, the specific number of sample groups is 199 groups, and the number of sample groups is used as the upper limit value of the search boundary.

[0047] The number of features randomly selected when constructing each decision tree in step S2 is determined by the number of input variables, and the specific input variables are P-wave, uniaxial compressive strength, tensile strength and rock density. Therefore, the number of features randomly selected when constructing each decision tree is set to 1-4;

[0048] The Golden Jackal Optimization (GJO) algorithm used in this embodiment is a population-based meta-heuristic algorithm. The flowchart of the Golden Jackal algorithm is as follows: Figure 3 The core principle of the algorithm is mainly based on the two main behavior modes of golden jackals: (1) leadership behavior: simulating the behavior of a leader in a group guiding other members; (2) exploration behavior: simulating the behavior of an individual exploring a new area independently;

[0049] Leadership Behavior:

[0050] Where: X i is the solution in the initial population, X j is the number of the initial population with X i Another different solution, α is the leadership rate, and the value range of α is 0 to 1;

[0051] Exploration Behavior:

[0052] Where: is the new solution, β is the exploration rate, and randn() is a random number from a standard normal distribution.

[0053] The number of decision trees in the random forest algorithm and the number of features randomly selected when constructing each decision tree are used as the initial population of the golden jackal optimization algorithm for parameter optimization.

[0054] The specific steps of the golden jackal optimization algorithm in step S2 are as follows:

[0055] S21. Define the objective function f(X), set the population size N, and the maximum number of iterations Max i ter, leadership rate α, exploration rate β and randomly generate the initial population X1, X2, …, X N ;

[0056] S22, calculate the fitness f(X) of each solution in the initial population i );

[0057] S23, for each solution X in the initial population i If randn()<α, execute the leadership behavior, otherwise execute the exploration behavior, process the search boundary, and ensure that the new solution In the search space;

[0058] S24, evaluate the new solution, calculate And update the calculation results, if Better than f(X i ),

[0059] Use Replace X i .

[0060] The step S3 needs to update the parameters optimized by S2 into a new random forest algorithm to obtain a random forest optimized by the golden jackal algorithm (GJO-RF), and use the optimized random forest algorithm to predict the rock shear strength parameters again;

[0061] Specifically, the step S3 is to use the cohesion and friction angle in the 199 groups of rock samples collected as data sets to form a training set required by the random forest algorithm. The number of decision trees in the random forest algorithm is determined by the golden jackal algorithm in the step S2, which is specifically 199; the training set of rock parameters is resampled by the Bootstrap method, and a number of training subsets and regression trees are randomly generated according to the sampling results. The number of features randomly sampled and selected when constructing each decision tree is set to 1-4 after being optimized by the golden jackal algorithm; when each node of the regression tree splits and grows, an input parameter is randomly selected as the split subset of the current node; the regression tree recursively branches and grows from top to bottom, and the regression tree stops growing after the segmentation termination condition of the preliminary rock shear strength parameter prediction is met. The regression trees that have completed growth are aggregated to form a random forest, and a prediction set is obtained and the average value is calculated to obtain the final prediction result. The comparison image between the predicted results of cohesion and friction angle and the actual results is shown in FIG. Figure 4 As shown in the figure, the prediction set results of the optimized random forest algorithm for predicting cohesion and friction angle are as follows Figure 5 shown.

Claims

1. An interpretable machine learning method for predicting rock shear strength parameters, characterized in that: The following steps are involved: S1. Preliminary prediction of cohesion and friction angle in rock shear strength parameters is carried out by using random forest algorithm; S2. Set the size of the search boundary of the golden jackal optimization algorithm, and iteratively optimize the parameters that need to be optimized in the random forest algorithm, namely, the number of decision trees and the number of features randomly sampled when building each decision tree; S3. The number of the optimized decision trees and the number of features randomly selected when constructing each decision tree are brought into the random forest algorithm, and the cohesion and the friction angle are predicted again to obtain a final prediction result.

2. An interpretable machine learning method for predicting rock shear strength parameters according to claim 1, characterized in that: The specific steps of the random forest algorithm in step S1 are as follows: S11, taking the cohesion and the friction angle in the rock sample as a data set to form a training set required by a random forest algorithm; S12, re-sampling the training set of rock parameters using the Bootstrap method, and randomly generating a number of training subsets and regression trees according to the sampling results; S13, when each node of the regression tree splits and grows, randomly extracting input parameters as the split subset of the current node; S14, the regression tree recursively branches and grows from top to bottom, and the regression tree stops growing after the segmentation termination condition of the preliminary rock shear strength parameter prediction is met, and the regression trees that have completed growth are aggregated to form a random forest, and a prediction set is obtained and the average value is calculated to obtain the final result.

3. The interpretable machine learning method for predicting rock shear strength parameters according to claim 1, characterized in that: The setting of the search boundary size in step S2 is determined by the number of rock sample groups, the upper limit of the search boundary is specifically the number of sample groups, and the lower limit of the search boundary is the smallest positive integer.

4. The interpretable machine learning method for predicting rock shear strength parameters according to claim 1, characterized in that: The number of features randomly sampled when constructing each decision tree in step S2 is determined by the number of input variables, and the specific input variables include P-wave, uniaxial compressive strength, tensile strength and rock density.

5. The interpretable machine learning method for predicting rock shear strength parameters according to claim 1, characterized in that: The number of decision trees in the random forest algorithm and the number of features randomly selected when constructing each decision tree are used as the initial population of the golden jackal optimization algorithm for parameter optimization.

6. The interpretable machine learning method for predicting rock shear strength parameters according to claim 1, characterized in that: The specific steps of the golden jackal optimization algorithm in step S2 are as follows: S21. Define the objective function f(X), set the population size N, and the maximum number of iterations Max i ter, leadership rate α, exploration rate β and randomly generate the initial population X1, X2, …, X N ; S22, calculate the fitness f(X) of each solution in the initial population i ); S23, for each solution X in the initial population i If randn()<α, execute the leadership behavior, otherwise execute the exploration behavior, process the search boundary, and ensure that the new solution In the search space; S24, evaluate the new solution, calculate And update the calculation results, if Better than f(X i ), then use Instead of X i .

7. An interpretable machine learning method for predicting rock shear strength parameters according to claim 6, characterized in that: The leadership behavior and the exploration behavior in step S23 are specifically as follows: Leadership Behavior: Where: X i is the solution in the initial population, X j is the number of the initial population with X i Another different solution, α is the leadership rate, and the value range of α is 0 to 1; Exploration Behavior: Where: is the new solution, β is the exploration rate, and randn() is a random number from a standard normal distribution.

8. An interpretable machine learning device for predicting rock shear strength parameters, characterized in that: It includes at least one processor and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute an interpretable machine learning method for predicting rock shear strength parameters as described in any one of claims 1 to 7.

Images