Tunnel excavation plastic zone evolution characteristic prediction method based on data driving

Abstract

The invention discloses a tunnel excavation plastic zone evolution characteristic prediction method based on data driving, and the method comprises the steps: 1, determining an input variable and a value range for the prediction of the evolution characteristic of a tunnel excavation plastic zone based on main influence factors of the evolution characteristic of the tunnel excavation plastic zone; 2, parameterizing and representing an evolution characteristic curve of a tunnel excavation plastic zone; step 3, establishing a tunnel excavation plastic zone evolution characteristic curve parameter prediction model based on data driving; 4, on the basis of the tunnel excavation plastic zone evolution characteristic curve parameter prediction model, tunnel excavation plastic zone evolution characteristic curve parameters under the given input variable condition are predicted; and 5, based on the predicted parameters of the evolution characteristic curve of the tunnel excavation plastic zone, drawing the evolution characteristic curve of the tunnel excavation plastic zone. The method can break through the bottleneck that a traditional simulation analysis method cannot rapidly predict the evolution characteristics of the tunnel excavation plastic zone, and has wide application prospects in tunnel surrounding rock stability safety control and support optimization design.

Description

Technical Field

[0001] This invention relates to the field of tunnel structural behavior prediction and analysis, and in particular to a data-driven method for predicting the evolution characteristics of the plastic zone during tunnel excavation. Background Technology

[0002] Hydraulic tunnels are core infrastructure components in water conservancy and hydropower projects, responsible for water conveyance and diversion. They must be excavated and constructed within mountains or underground. During tunnel excavation, the original stress balance of the surrounding rock is disrupted, leading to stress redistribution. When the local stress exceeds the yield strength of the surrounding rock, a plastic zone is formed. If the plastic zone expands excessively, it may cause safety accidents such as rock collapse and support structure failure. Therefore, accurately predicting the evolution characteristics of the plastic zone during tunnel excavation is the prerequisite and foundation for achieving safe control of tunnel surrounding rock stability and optimizing the design of excavation and support schemes.

[0003] Currently, the prediction of the evolution characteristics of the plastic zone during tunnel excavation mainly employs numerical simulation analysis, which uses numerical analysis tools such as the finite element method to simulate the tunnel excavation process and predict the evolution characteristics of the plastic zone of the surrounding rock during the advancement of the excavation face. However, existing technologies still have the following shortcomings:

[0004] 1. In the existing technology, predicting the evolution characteristics of the plastic zone in tunnel excavation requires the establishment of a complex finite element simulation model and the corresponding calculation and analysis, which is labor-intensive, time-consuming, and has low prediction efficiency.

[0005] 2. In the existing technology, when the design scheme is adjusted or the surrounding rock mechanical parameters change, the corresponding finite element simulation modeling and analysis need to be carried out again. It is impossible to obtain the evolution characteristics of the plastic zone of tunnel excavation under different schemes and parameters in a timely manner, which seriously affects the design and construction progress. Summary of the Invention

[0006] The technical problem to be solved by this invention is to address the shortcomings of the prior art by providing a data-driven method for predicting the evolution characteristics of the plastic zone in tunnel excavation. This data-driven method for predicting the evolution characteristics of the plastic zone in tunnel excavation breaks through the bottleneck of traditional simulation analysis and prediction methods being unable to quickly predict the evolution characteristics of the plastic zone in tunnel excavation by parametrically characterizing the evolution characteristic curve of the plastic zone in tunnel excavation and establishing a parameter prediction model for the evolution characteristic curve of the plastic zone in tunnel excavation. It has broad application prospects in the stability and safety control of tunnel surrounding rock and the optimization design of support.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0008] A data-driven method for predicting the evolution characteristics of the plastic zone in tunnel excavation includes the following steps.

[0009] Step 1: Based on the main influencing factors of the evolution characteristics of the plastic zone in tunnel excavation, determine several input variables for predicting the evolution characteristics of the plastic zone in tunnel excavation and the value range of each input variable.

[0010] Step 2: Parametrically characterize the evolution curve of the plastic zone during tunnel excavation.

[0011] Step 3: Establish a data-driven model for predicting the parameters of the evolution characteristic curve of the plastic zone in tunnel excavation.

[0012] Step 4: Based on the prediction model of the evolution characteristic curve parameters of the plastic zone of tunnel excavation established in Step 3, predict the parameters of the evolution characteristic curve of the plastic zone of tunnel excavation under given input variable conditions.

[0013] Step 5: Substitute the parameters of the tunnel excavation plastic zone evolution characteristic curve obtained in Step 4 into Step 2 to obtain the tunnel excavation plastic zone evolution characteristic curve.

[0014] In step 1, the main influencing factors on the evolution characteristics of the plastic zone during tunnel excavation include lithology, burial depth, tunnel diameter, initial geostress field, and excavation scheme.

[0015] In step 1, the evolution characteristics of the plastic zone in tunnel excavation refer to the evolution process of the distribution characteristics of the plastic zone in the surrounding rock at the predicted profile during tunnel excavation, as the distance D between the predicted profile and the excavation face changes. Specifically, the distribution characteristics of the plastic zone in the surrounding rock are the radial average distribution depth T of the plastic zone in the surrounding rock. p .

[0016] In step 2, the evolution characteristic curve T of the plastic zone during tunnel excavation is shown. p ~D is represented by a three-parameter exponential function, and its specific expression is:

[0017]

[0018] In the formula, T p To predict the radial average distribution depth of the plastic zone of tunnel excavation in the surrounding rock at the profile.

[0019] D is the distance between the predicted profile and the excavation face.

[0020] The maximum radial average depth of the excavated plastic zone in the surrounding rock is denoted as , and is an unknown parameter to be predicted.

[0021] α1 is T when D=0 p Values ​​and The ratio of to represents the unknown parameter to be predicted.

[0022] α2 is the rate of change of Tp, which is an unknown parameter to be predicted.

[0023] Step 3, the method for establishing the prediction model of the evolution characteristic curve of the plastic zone in tunnel excavation, includes the following steps.

[0024] Step 3-1: Select the basic model: The input to the basic model is N determined in Step 1. p The input variables are the unknown parameters in the evolution characteristic curve of the plastic zone of tunnel excavation in step 2.

[0025] Step 3-2: Construct an input sample library: Sample within the sampling range of each input variable to obtain N input samples.

[0026] Step 3-3: For each input sample, construct a finite element model of tunnel excavation.

[0027] Steps 3-4: For each tunnel excavation finite element model, analyze its T value under different D values. p Values, and all resulting T values p The values ​​are fitted to the corresponding D values ​​to obtain all the unknown parameter values ​​of the corresponding input samples; thus, all the unknown parameter values ​​corresponding to the N input samples constitute the output sample library.

[0028] Step 3-5: Using the input sample library constructed in Step 3-2 and the output sample library obtained in Step 3-4, the basic model in Step 3-1 is trained to obtain the parameter prediction model for the evolution characteristic curve of the plastic zone in tunnel excavation.

[0029] In step 3-1, the base model is a machine learning model.

[0030] In step 3-3, the tunnel excavation finite element model uses the predicted profile as the middle cross section of the model, with an axial length of not less than 15 times the tunnel diameter, the center of the cross section located at the center of the tunnel, and the vertical distance between the center of the cross section and the surrounding boundaries not less than 5.5 times the tunnel diameter.

[0031] In steps 3-4, the selection method for different D values ​​is as follows: the maximum value of D is not less than 4 times the tunnel diameter, the minimum value of D is the opposite of its maximum value, and the difference between adjacent D values ​​is taken as the excavation cycle advance.

[0032] In steps 3-4, for each tunnel excavation finite element model, T is calculated under different D values. p The method for calculating the value includes the following steps:

[0033] Step 3-4A: Calculate the area A of the plastic zone in the tunnel excavation at each predicted profile under each D value.

[0034] Step 3-4B: Calculate T for each D value p The value is calculated using the following formula:

[0035]

[0036] In the formula, R is the tunnel diameter.

[0037] In steps 3-4A, the calculation method for the area A of the tunnel excavation plastic zone at each predicted profile under each D value is as follows: For any value of D, based on the three-dimensional finite element analysis results of the excavation process, obtain the equivalent plastic strain distribution cloud map of the surrounding rock at the predicted profile, and calculate the sum of the areas of the regions with equivalent plastic strain greater than 0 in the equivalent plastic strain distribution cloud map, which is the desired A value.

[0038] The present invention has the following beneficial effects:

[0039] 1. Based on the parametric characterization of the evolution characteristic curve of the plastic zone in tunnel excavation, this invention establishes a data-driven prediction model for the parameters of the evolution characteristic curve of the plastic zone in tunnel excavation. After training, this prediction model can directly predict the parameters of the corresponding evolution characteristic curve of the plastic zone in tunnel excavation for a given set of input data without conducting corresponding finite element simulation analysis of tunnel excavation construction, thereby quickly plotting the evolution characteristic curve of the plastic zone in tunnel excavation.

[0040] 2. This invention breaks through the bottleneck of traditional simulation analysis and prediction methods being unable to quickly predict the evolution characteristics of the plastic zone during tunnel excavation, and provides an effective solution for efficiently carrying out the stability and safety control of tunnel surrounding rock and optimizing the design of excavation and support schemes. Attached Figure Description

[0041] Figure 1 This is a flowchart of a data-driven method for predicting the evolution characteristics of the plastic zone in tunnel excavation, according to the present invention.

[0042] Figure 2 This is a schematic diagram of a finite element model of tunnel excavation according to an embodiment of the present invention.

[0043] Figure 3 The following is an example of an equivalent plastic strain distribution cloud map of the surrounding rock at the observation profile in an embodiment of the present invention; wherein, (a) corresponds to D=-2m; (b) corresponds to D=0m; (c) corresponds to D=2m; and (d) corresponds to D=4m.

[0044] Figure 4 This is an example diagram of parameter fitting for the characterization curve of the plastic zone evolution characteristics of tunnel excavation in an embodiment of the present invention.

[0045] Figure 5 This is a graph showing the predicted results of the evolution characteristic curve of the plastic zone during tunnel excavation in an embodiment of the present invention. Detailed Implementation

[0046] The present invention will now be described in further detail with reference to the accompanying drawings and specific preferred embodiments.

[0047] like Figure 1 As shown, the data-driven method for predicting the evolution characteristics of the plastic zone in tunnel excavation includes the following steps.

[0048] Step 1: Based on the main influencing factors of the evolution characteristics of the plastic zone in tunnel excavation, determine several input variables for predicting the evolution characteristics of the plastic zone in tunnel excavation and the value range of each input variable.

[0049] The evolution characteristics of the plastic zone in tunnel excavation refer to the evolution of the distribution characteristics of the plastic zone in the surrounding rock at the predicted profile during tunnel excavation, as the distance D between the predicted profile and the excavation face changes. Specifically, the distribution characteristics of the plastic zone in the surrounding rock are the radial average distribution depth T of the plastic zone in the surrounding rock. p .

[0050] Based on the main influencing factors of the evolution characteristics of the plastic zone in tunnel excavation, the specific process for determining the input variables and their value ranges for predicting the evolution characteristics of the plastic zone in tunnel excavation is as follows: Based on actual engineering data and prediction requirements, determine N, which affects the evolution characteristics of the plastic zone in tunnel excavation. p One of the main factors is used as N to predict the evolution characteristics of the plastic zone during tunnel excavation. p Input variable P i (i=1, 2, …, N) p ), and determine the reasonable range of values ​​for each input variable [a i ,b i (i=1, 2, …, N) p ).

[0051] In this embodiment, a water conveyance tunnel for a reservoir is about 4.2 km long with a large variation in burial depth. The burial depth of the main tunnel section ranges from 255 to 465 m. The rock strata along the tunnel are steeply dipping, and the lithology is mainly dolomite and limestone. The initial in-situ stress is mainly self-weight stress, and the tectonic stress is not significant. Different tunnel sections have different excavation plastic zone evolution characteristics. The tunnel excavation diameter is 7.2 m, with full-face excavation and an excavation cycle advance of 2 m.

[0052] The main influencing factors of the evolution characteristics of the plastic zone during tunnel excavation include lithology, burial depth, tunnel diameter, initial geostress field, and excavation scheme. In this embodiment, due to the large variations in burial depth and lithology of the water conveyance tunnel, in order to meet the need for rapid prediction of the evolution characteristics of the plastic zone during tunnel excavation under different burial depths and lithological conditions, the surrounding rock deformation modulus E, Poisson's ratio μ, internal friction angle φ, cohesion c, and tunnel burial depth H are used as the five main factors affecting the evolution characteristics of the plastic zone during tunnel excavation. The surrounding rock deformation modulus E, Poisson's ratio μ, internal friction angle φ, cohesion c, and tunnel burial depth H are respectively used as the five input variables P1, P2, P3, P4, and P5 for predicting the evolution characteristics of the plastic zone during excavation of different tunnel sections. The reasonable value ranges of each input variable determined based on the topographic and geological data of the tunnel are shown in Table 1.

[0053] Table 1 Input variables and their value ranges for predicting the evolution characteristics of the plastic zone in tunnel excavation.

[0054]

[0055] Step 2, Parametric characterization of the evolution characteristic curve T of the plastic zone during tunnel excavation. p ~D is preferably represented by a three-parameter exponential function, and its specific expression is as follows:

[0056] (1)

[0057] In the formula, T p To predict the radial average distribution depth of the plastic zone of tunnel excavation in the surrounding rock at the profile.

[0058] D is the distance between the predicted profile and the excavation face.

[0059] The maximum radial average depth of the excavated plastic zone in the surrounding rock is denoted as , and is an unknown parameter to be predicted.

[0060] α1 is T when D=0 p Values ​​and The ratio of to represents the unknown parameter to be predicted.

[0061] α2 is T p Increase the rate of change of the rate of change, which is an unknown parameter to be predicted.

[0062] Step 3: Establish a data-driven model for predicting the parameters of the evolution characteristic curve of the plastic zone in tunnel excavation.

[0063] Step 3-1: Select the basic model

[0064] The input to the basic model is N, which is determined in step 1. pThe input variables are: P1, P2, P3, P4 and P5 corresponding to the surrounding rock deformation modulus E, Poisson's ratio μ, internal friction angle φ, cohesion c and tunnel burial depth H.

[0065] The output of the basic model is the unknown parameters from the evolution characteristic curve of the plastic zone during tunnel excavation in step 2. In this embodiment, the preferred method is: The three unknown parameters are α1 and α2.

[0066] Furthermore, the aforementioned basic model is preferably a machine learning model, such as the Kriging machine learning model, with the specific expression as follows:

[0067]

[0068] in:

[0069]

[0070] X=[ … ]

[0071] In the formula, The predicted output of the basic model, 1≤m≤N p In this embodiment, N p =3, therefore, m=1,2,3.

[0072] r is an N-order correlation vector, composed of the Gaussian correlation function values ​​between the predicted input variable and the model input variable in the training samples; where N is the total number of samples in the training set in the subsequent sample library.

[0073] K is an N-order correlation matrix, composed of Gaussian correlation function values ​​between the model input variables in the training samples.

[0074] U is an N-order identity matrix; This is an intermediate quantity.

[0075] X is the set of predicted output values ​​for N samples in the training set. In this embodiment, N=30 is preferred.

[0076] Alternatively, the base model can be other machine learning models in existing technologies, such as neural network learning models.

[0077] In this embodiment, the training set S, also known as the sample library, preferably includes 30 training samples. The j-th sample is denoted as S. j (j=1, 2, …, N); the training set S includes the input sample library and its corresponding output sample library.

[0078] Step 3-2: Construct the input sample library: Sample within the sampling range of each input variable to obtain N=30 input samples, and denoted as N=30. (i=1, 2, …, N) p (j=1, 2, …, N), where j is the sample number; i is the input variable number.

[0079] The model input space of the parameter prediction model for the evolution characteristic curve of the plastic zone in tunnel excavation is composed of P i (i=1, 2,…, N) p N composed of ) p dimensional bounded space, P i ∈[a i , b i ].

[0080] In this embodiment, the model input space of the tunnel excavation plastic zone evolution characteristic curve parameter prediction model is composed of P i A 5-dimensional bounded space consisting of (i=1, 2, …, 5), P1∈[5,10], P2∈[0.24,0.28], P3∈[35,45], P4∈[0.5,1], P5∈[255,465]; The Sobol sequence sampling method based on uniform distribution is used to sample the model input space of the tunnel excavation plastic zone evolution characteristic curve parameter prediction model, obtaining 30 training samples of input data. (i=1,2, …, 5; j=1, 2, …, 30), as shown in Table 2 below.

[0081] Table 2 Input data for each sample in the training sample set S

[0082]

[0083] Step 3-3: For each input sample, construct a... Figure 2 The finite element model of the tunnel excavation shown uses the predicted profile as the central cross section of the model. The axial length is no less than 15 times the tunnel diameter, the center of the cross section is located at the center of the tunnel, and the vertical distance between the center of the cross section and the surrounding boundaries is no less than 5.5 times the tunnel diameter. This corresponds to the j-th sample S. j Input data The finite element model of the tunnel excavation is denoted as M. j .

[0084] In this embodiment, as Figure 2 As shown, the finite element model of tunnel excavation uses the predicted profile as the middle cross section of the model, with an axial length of 120m (16.67 times the tunnel diameter). The center of the cross section is located at the center of the tunnel, and the vertical distance between the center of the cross section and the surrounding boundaries is 40m (5.56 times the tunnel diameter).

[0085] Steps 3-4: For each tunnel excavation finite element model, analyze its T value under different D values. p Values, and all resulting T values p The values ​​are fitted to the corresponding D values ​​to obtain all the unknown parameter values ​​of the corresponding input samples; thus, all the unknown parameter values ​​corresponding to the N input samples constitute the output sample library.

[0086] I. Three-dimensional finite element analysis of tunnel excavation process

[0087] The three-dimensional finite element analysis results of the excavation process refer to the variation of the distribution range of equivalent plastic strain in the surrounding rock at the predicted profile obtained by the three-dimensional finite element analysis of the tunnel excavation process based on the established tunnel excavation finite element model, as the distance D between the predicted profile and the excavation face changes.

[0088] In this invention, the selection method for different D values ​​is as follows: the maximum value of D is not less than 4 times the tunnel diameter, the minimum value of D is the opposite of its maximum value, and the difference between adjacent D values ​​is taken as the excavation cycle advance.

[0089] In this embodiment, the preferred method for selecting different D values ​​is to take values ​​from -30m to 30m, with a 2m increment, for a total of 30 values.

[0090] In this example, for each tunnel excavation finite element model, T is shown under different D values. p The method for calculating the value includes the following steps.

[0091] Step 3-4A: Calculate the area A of the tunnel excavation plastic zone at the predicted profile for each D value. The preferred calculation method is as follows: For any value of D, based on the results of the three-dimensional finite element analysis of the excavation process, obtain... Figure 3 The equivalent plastic strain distribution cloud map of the surrounding rock at the predicted profile shown is used to calculate the sum of the areas of regions with equivalent plastic strain greater than 0 in the equivalent plastic strain distribution cloud map. This sum is the required value A, as shown in Table 3 below.

[0092] Step 3-4B: Calculate T for each D value p The value is calculated using the following formula:

[0093]

[0094] In the formula, R is the tunnel diameter.

[0095] In this embodiment, T corresponding to 30 D values ​​is calculated. p The values ​​are shown in Table 3 below.

[0096] Table 3. A and T under different values ​​of D p The calculation results (taking the tunnel excavation finite element model M1 as an example)

[0097]

[0098] Table 4 Output data of each sample in the training sample set S

[0099]

[0100] II. T p Fitting values ​​to D-value data

[0101] The finite element model for each tunnel excavation is denoted as M. j The corresponding 30 D values ​​and 30 T values p The value was fitted using formula (1), and the fitting process is as follows: Figure 4 As shown in Table 4 above, the characterization parameters of the 30 sets of tunnel excavation plastic zone evolution curves were obtained. Fitted values ​​of α1 and α2 , and Let j = 1, 2, …, 30, and let = , = , = The training set S containing 30 samples is required to complete the prediction model of the evolution characteristic curve of the plastic zone in tunnel excavation. j Construction of (j=1, 2, …,30).

[0102] Step 3-5: Using the input sample library constructed in Step 3-2 and the output sample library obtained in Step 3-4, the basic model in Step 3-1 is trained to obtain the parameter prediction model for the evolution characteristic curve of the plastic zone in tunnel excavation.

[0103] Step 4: Based on the tunnel excavation plastic zone evolution characteristic curve parameter prediction model established in Step 3, predict the tunnel excavation plastic zone evolution characteristic curve parameters under given input variables. The specific process is as follows: using the given input variables... (i=1, 2, …, N) p Using the established tunnel excavation plastic zone evolution characteristic curve parameters as model input, the corresponding model output is predicted by the model prediction model. (m=1, 2, 3), and P i * Predicted values ​​of the corresponding parameters of the evolution characteristic curve of the plastic zone in tunnel excavation , and They are respectively equal to , , .

[0104] In this embodiment, three typical tunnel sections (section I, section II, and section III) with different tunnel depths and surrounding rock mechanical parameters are selected along the tunnel route. The depth and surrounding rock mechanical parameters of each typical tunnel section are given input variables. (i=1,2, …, 5), as shown in Table 5.

[0105] Table 5. Model inputs for predicting the evolution characteristics of the plastic zone in three typical tunnel sections.

[0106]

[0107] The input variables in Table 5 Using the established tunnel excavation plastic zone evolution characteristic curve parameters as input, the corresponding model output is predicted. (m=1, 2, 3), the predicted values ​​are... , , That is, respectively equal to the predicted values ​​of the parameters of the plastic zone evolution characteristic curves of each typical tunnel section. , and As shown in Table 6.

[0108] Table 6. Predicted values ​​of parameters for the evolution characteristics curves of the plastic zone in three typical tunnel sections.

[0109]

[0110] Step 5: Substitute the parameters of the tunnel excavation plastic zone evolution characteristic curve obtained in Step 4 into Step 2 to obtain the tunnel excavation plastic zone evolution characteristic curve. The specific process is as follows: Let the parameters in equation (1) be... α1 and α2 are the predicted values ​​of the parameters of the evolution characteristic curve of the plastic zone in tunnel excavation, respectively. , and Plotting the function curve yields the predicted evolution characteristic curve of the plastic zone in tunnel excavation, as shown below. Figure 5 As shown.

[0111] The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention, various equivalent transformations can be made to the technical solutions of the present invention, and these equivalent transformations all fall within the protection scope of the present invention.

Claims

1. A data-driven method for predicting the evolution characteristics of the plastic zone in tunnel excavation, characterized by: The method comprises the following steps: Step 1, based on the main influencing factors of the evolution characteristics of the plastic zone of tunnel excavation, determine several input variables for predicting the evolution characteristics of the plastic zone of tunnel excavation and the value range of each input variable; Step 2, parameterize the evolution characteristics curve of the plastic zone of tunnel excavation; Step 3, establish a data-driven evolution characteristics curve parameter prediction model of the plastic zone of tunnel excavation; Step 4, based on the evolution characteristics curve parameter prediction model of the plastic zone of tunnel excavation established in step 3, predict the evolution characteristics curve parameters of the plastic zone of tunnel excavation under the condition of given input variables; Step 5, substitute the evolution characteristics curve parameters of the plastic zone of tunnel excavation obtained in step 4 into step 2 to obtain the evolution characteristics curve of the plastic zone of tunnel excavation.

2. The data-driven based tunnel excavation plastic zone evolution characteristics prediction method according to claim 1, characterized in that: In step 1, the main influencing factors of the evolution characteristics of the plastic zone of tunnel excavation include lithology, burial depth, tunnel diameter, initial ground stress field and excavation scheme.

3. The data-driven based tunnel excavation plastic zone evolution characteristics prediction method according to claim 1, characterized in that: In step 1, the evolution characteristics of the plastic zone of the tunnel excavation refer to the evolution process of the distribution characteristics of the plastic zone of the tunnel excavation at the prediction profile in the surrounding rock varying with the distance D between the prediction profile and the excavation face during the tunnel excavation process, and the distribution characteristics of the plastic zone of the tunnel excavation in the surrounding rock specifically refer to the radial average distribution depth T of the plastic zone of the tunnel excavation in the surrounding rock p .

4. The data-driven based tunnel excavation plastic zone evolution characteristics prediction method according to claim 3, characterized in that: In step 2, the evolution characteristic curve T of the plastic zone of the hole p The three-parameter exponential function is used to characterize the expression, which is specifically expressed as: ； In the formula, is the maximum value of the average radial distribution depth of the excavated plastic zone in the surrounding rock, which is an unknown parameter to be predicted; T when a1 is D = 0 p the value of the ratio of the value of a1 to a2 is the unknown parameter to be predicted; a2 is T p The rate of change of the increase is the unknown parameter to be predicted.

5. The data-driven based tunnel excavation plastic zone evolution characteristics prediction method according to claim 4, characterized in that: In step 3, the method for establishing the evolution characteristics curve parameter prediction model of the plastic zone of tunnel excavation comprises the following steps: Step 3-1, selecting a base model: the input of the base model is N input variables determined in step 1, and the output of the base model is the unknown parameter in the plastic zone evolution characteristic curve of the tunnel excavation in step 2. p Step 3-1, selecting a base model: the input of the base model is N input variables determined in step 1, and the output of the base model is the unknown parameter in the plastic zone evolution characteristic curve of the tunnel excavation in step 2. Step 3-2, construct an input sample library: sample in the sampling range of each input variable to obtain N input samples; Step 3-3, for each input sample, construct a tunnel excavation finite element model; Step 3-4, for each tunnel excavation finite element model, analyze its T p value under different D values, and fit all the obtained T p values with the corresponding D values to obtain all unknown parameter values of the corresponding input samples; thus, all unknown parameter values corresponding to N input samples constitute an output sample library; Step 3-5, use the input sample library constructed in step 3-2 and the output sample library obtained in step 3-4 to train the base model in step 3-1, thereby obtaining the evolution characteristics curve parameter prediction model of the plastic zone of tunnel excavation.

6. The data-driven based tunnel excavation plastic zone evolution characteristics prediction method according to claim 5, characterized in that: In step 3-1, the base model is a machine learning model.

7. The data-driven based tunnel excavation plastic zone evolution characteristics prediction method according to claim 5, characterized in that: In step 3-3, the tunnel excavation finite element model takes the predicted profile as the middle cross section, the axial length is not less than 15 times the tunnel diameter, the center of the cross section is located in the center of the tunnel, and the vertical distance between the center of the cross section and the surrounding boundary is not less than 5.5 times the tunnel diameter.

8. The data-driven based tunnel excavation plastic zone evolution characteristics prediction method according to claim 5, characterized in that: In step 3-4, the selection method of different D values is: the maximum value of D is not less than 4 times the tunnel diameter, the minimum value of D is the opposite of the maximum value, and the difference between adjacent D values is the excavation cycle footage.

9. The data-driven based tunnel excavation plastic zone evolution characteristics prediction method according to claim 5, characterized in that: In step 3-4, each tunnel excavation finite element model, T p The method for calculating the value of D includes the following steps: Step 3-4A, calculate the tunnel excavation plastic zone area A at the predicted profile under each D value; Step 3 - 4B, Calculate T for each D value p value, specifically calculated as: ； In the formula, R is the tunnel diameter.

10. The data-driven based tunnel excavation plastic zone evolution characteristics prediction method according to claim 9, characterized in that: In step 3-4A, the calculation method of the tunnel excavation plastic zone area A value at the predicted profile under each D value is: for any value of D, based on the three-dimensional finite element analysis results of the excavation process, the equivalent plastic strain distribution cloud diagram of the surrounding rock at the predicted profile is obtained, and the sum of the areas of the regions with equivalent plastic strain greater than 0 in the equivalent plastic strain distribution cloud diagram is calculated, which is the value of A.

Images