Method for determining construction design parameters of isolation piles between newly-built foundation pit and existing tunnel

Abstract

The present application relates to the field of building construction, specifically to a method for determining the construction design parameters of isolation piles between a newly built foundation pit and an existing tunnel. Through three-dimensional numerical simulation and DE-GWO-SVM nonlinear mapping regression model, the horizontal displacement and other performance indicators of the tunnel can be accurately predicted. This prediction capability enables real-time adjustment of the isolation pile design parameters during construction, effectively controlling the deformation of the tunnel, avoiding structural damage, ensuring the safety of construction, and reducing the risk during the implementation process. By introducing a multi-objective optimization framework based on DE-GWO-SVM-NSGA-II, the design parameters of the isolation pile are further optimized under the premise of meeting the safety requirements of the tunnel. The optimization objectives include not only the minimization of tunnel deformation but also the minimization of construction cost of the isolation pile. Ultimately, the construction parameters of the isolation pile determined by the TOPSIS method can significantly reduce the project cost while ensuring safety, achieving the purpose of improving economic efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of building construction, in particular to a method for determining construction design parameters of isolation piles between a newly-built foundation pit and an existing tunnel. BACKGROUND

[0002] With the accelerated urbanization process, high-rise buildings and large infrastructure construction projects are increasing. As a key part of such projects, the importance of deep foundation pit engineering is self-evident. In particular, in the urban area adjacent to the existing subway tunnel and densely populated, the newly-built foundation pit is extremely challenging. Improper construction handling may cause irreversible damage to the tunnel, and even induce safety accidents such as collapse. How to effectively protect the subway tunnel, which is the artery of urban transportation, is crucial to the operation of the city.

[0003] The isolation pile technology is a common practice to deal with such working conditions, which can effectively block the influence of foundation pit excavation on the tunnel, and ensure construction efficiency and economy. However, in current engineering practice, the construction parameters of isolation piles are often determined by experience or overly conservative design. This may lead to inaccurate parameter selection, increasing the risk of engineering safety and economic cost. In particular, in complex geological conditions and dense urban environment, the traditional design method is difficult to cope with the complex interaction and uncertainty between underground structures. Therefore, it is particularly urgent to develop a more scientific and systematic method to determine these construction parameters. SUMMARY

[0004] To solve the above problems, the embodiment of the present application provides a method for determining construction design parameters of isolation piles between a newly-built foundation pit and an existing tunnel.

[0005] The method for determining construction design parameters of isolation piles between a newly-built foundation pit and an existing tunnel provided by the embodiment of the present application comprises the following steps.

[0006] S1, determining and acquiring geological data and engineering information parameters related to the newly-built foundation pit and the existing tunnel, and acquiring physical and mechanical parameters of the soil in the site range of the newly-built foundation pit and the existing tunnel;

[0007] S2, using the acquired parameters in combination with the actual engineering of the newly-built foundation pit and the existing tunnel, determining the construction design parameters affecting the bearing mechanism of the isolation pile, setting up orthogonal design schemes of different isolation pile construction design parameters through orthogonal test; based on three-dimensional finite difference calculation, constructing a numerical analysis model of the newly-built foundation pit excavation under each design scheme according to the orthogonal design scheme, and carrying out simulation calculation of the whole process of foundation pit construction to obtain the performance evaluation index of the existing tunnel under different design scheme working conditions;

[0008] S3, taking the isolation pile construction design parameters as input and the existing tunnel performance evaluation indexes corresponding to the isolation pile construction design parameters as output, a sample database is built; DE-GWO-SVM is used to learn and train the sample database, and a DE-GWO-SVM nonlinear mapping regression function is constructed;

[0009] S4, the DE-GWO-SVM nonlinear mapping regression function and the isolation pile cost function are combined to construct a DE-GWO-SVM-NSGA-II multi-objective optimization framework with the minimum deformation and the minimum cost of the existing tunnel as the target, so as to optimize the isolation pile construction design parameters between the new foundation pit and the existing tunnel.

[0010] Optionally,

[0011] The geological data and engineering information parameters include: the length, width and depth of the proposed foundation pit, the radius, segment thickness, segment specific gravity, segment elastic modulus and segment Poisson's ratio of the existing tunnel, and the vertical distance between the new foundation pit edge and the existing tunnel;

[0012] The physical and mechanical parameters include: soil type and thickness, elastic modulus, Poisson's ratio, bulk density, cohesion and friction angle.

[0013] Optionally,

[0014] The isolation pile construction design parameters include: the distance between the isolation pile and the existing tunnel, the pile top burial depth, the pile length, the pile diameter and the pile spacing.

[0015] Optionally, the specific process of constructing the DE-GWO-SVM nonlinear mapping regression function in S3 includes:

[0016] 1) Preprocessing sample data in the sample database

[0017] In order to improve the convergence speed and accuracy of the model and eliminate the influence of sample data size, the sample data is normalized before model training, and the sample data is mapped to the interval [0, 1];

[0018] 2) Selecting a kernel function

[0019] In order to improve the accuracy of the model, a radial basis kernel function is used, and the expression is:

[0020]

[0021] In the formula, σ is the kernel width parameter, and ||x-x i || represents the distance between data x and the center point x i ;

[0022] 3) Determining the optimal parameters of SVM

[0023] In order to make the SVM model have higher prediction accuracy, the differential evolution algorithm and grey wolf algorithm DE-GWO are combined with the pretreated sample data to optimize the hyperparameters of the SVM model, and the optimal penalty parameter and kernel function parameter are determined;

[0024] 4) Establishing DE-GWO-SVM nonlinear mapping regression function

[0025] Based on the optimal penalty parameter, kernel function parameter and sample data determined by DE-GWO, the SVM model is trained to establish the DE-GWO-SVM nonlinear mapping regression function.

[0026] Optionally, in the process of establishing the DE-GWO-SVM nonlinear mapping regression function, the mean absolute error MAE, the mean square error MSE and the mean absolute percentage error MAPE are used as the evaluation criteria of the model, and the formulas are as follows:

[0027]

[0028]

[0029]

[0030] Wherein, N is the number of sample data, y i,pre is the predicted value of the model, y i,act is the corresponding actual value.

[0031] The smaller the numerical MAE, MSE and MAPE in the training process, the better the prediction fitting degree of the nonlinear mapping regression function.

[0032] Optionally, the process of constructing the DE-GWO-SVM-NSGA-II multi-objective optimization framework in S4 specifically includes:

[0033] 1) Establishing objective function

[0034] Define the objective function f i (X), in order to meet the requirements of tunnel deformation and the best cost of isolation pile, define two objective functions f1(X) and f2(X):

[0035] f i (X)=(f1(X),f2(X))

[0036] The DE-GWO-SVM nonlinear mapping regression function replaces the traditional mathematical function and is used as the first objective function f1(X) of NSGA-II, representing the minimization of tunnel deformation, and is expressed as:

[0037] f1(X)=min[DE-GWO-SVM(X1,X2,X3,X4,X5)]

[0038] wherein, f1(X) is the minimum value of the existing tunnel horizontal position, DE-GWO-SVM(X1, X2, X3, X4, X5) is a nonlinear mapping regression function, X1 is the pile length, X2 is the pile diameter, X3 is the pile spacing, X4 is the pile top buried depth, and X5 is the distance between the pile tunnels;

[0039] The completed isolation pile is regarded as a wall, and the total cost of the unit length wall material is taken as the second objective function f2(X), representing cost minimization:

[0040]

[0041] wherein, f2(X) represents the total cost, C con is the unit volume concrete cost of the isolation pile, C reb is the unit mass of the steel reinforcement cost of the isolation pile, V con is the concrete volume of a single isolation pile, M lr and M sr are the longitudinal reinforcement mass and the stirrup mass of a single isolation pile, respectively;

[0042] The total volume of a single isolation pile is approximated as the concrete volume V con , and thus:

[0043]

[0044] 2) Determine the constraint range of the isolation pile construction design parameters

[0045] To prevent the optimized structure of the isolation pile construction design parameters from being too different from the actual engineering parameters, the constraint range of the isolation pile construction design parameters is set according to relevant construction specifications and engineering practices:

[0046]

[0047] wherein, X i represents the i-th isolation pile construction design parameter, and are the upper limit and the lower limit of the i-th isolation pile construction design parameter, respectively;

[0048] Specifically, the constraint range of the objective function is as follows:

[0049]

[0050] wherein, s.t. represents the constraint condition.

[0051] Optionally, the process of optimizing the isolation pile construction design parameters between the newly built foundation pit and the existing tunnel in S4 includes:

[0052] The NSGA-II algorithm is used for multi-objective optimization of the isolation pile construction design parameters to obtain a Pareto optimal solution set, which specifically includes:

[0053] Step 1, determine the basic parameters

[0054] Determine the basic parameters that affect the efficiency and solution quality of the NSGA-II algorithm, including population size, crossover rate, mutation rate, and iteration number;

[0055] Step 2, population initialization

[0056] Randomly generate an initial population, where each individual in the initial population represents a possible isolation pile construction design parameter;

[0057] Step 3, fitness evaluation and non-dominated sorting

[0058] Calculate the objective function value for each individual in the population, and perform non-dominated sorting based on the objective function value to distinguish different optimization levels, obtaining a Pareto frontier layer of multiple isolation pile construction design parameters. Each individual in each layer represents an optimal solution that is not dominated by other individuals in all objective functions;

[0059] Step 4, crowding distance calculation

[0060] For each individual in the Pareto frontier layer, calculate the crowding distance of the individual in the objective function space to ensure the diversity of the solution set. The crowding distance d i The calculation formula is as follows:

[0061]

[0062] Where d i represents the crowding distance of the i-th individual, represents the value of the i-th individual in the j-th objective function;

[0063] Step 5, genetic operation

[0064] Perform selection strategy based on crowding distance, and generate a new generation of population through crossover and mutation operations;

[0065] Step 6, iteration update

[0066] Repeat steps 1-5 until the iteration number is met. At the end of the algorithm, a Pareto optimal solution set is formed, which contains a series of design parameter combinations that achieve the best balance in the objective functions. Based on the Pareto optimal solution set, determine the final isolation pile construction design parameters.

[0067] Optionally, the process of determining the final isolation pile construction design parameters based on the Pareto optimal solution set includes:

[0068] The TOPSIS method is used to calculate the optimal solution for the Pareto optimal solution set in order to determine the final design parameters for the isolation pile construction. Specifically, this includes:

[0069] 1) Determine the ideal solution and the inferior solution

[0070] The ideal solution X+ is the solution that optimizes each objective function value. For the objectives of minimizing existing tunnel deformation and minimizing isolation pile cost, it is the solution with the minimum deformation and the lowest cost. The inferior solution X-, on the other hand, selects the solution with the maximum deformation and the highest cost.

[0071] 2) Calculate the distance between the ideal solution and the inferior solution.

[0072] For each Pareto solution in the Pareto optimal solution set, calculate its Euclidean distance to the ideal and inferior solutions, using the following formulas:

[0073]

[0074]

[0075] Where, d + (X k The distance from the k-th Pareto solution to the ideal solution; the smaller the distance, the closer the Pareto solution is to the optimal state across all objective functions; d-(X k The distance from the k-th Pareto solution to the worst-case solution; a larger distance indicates that the Pareto solution is further from the worst-case state across all objective functions; X k f represents the k-th Pareto solution; i (X k ) represents the value of the k-th Pareto solution on the i-th objective function; m is the number of objective functions;

[0076] 3) Calculate the relative closeness

[0077] Calculate the relative proximity CC of each Pareto solution. k :

[0078]

[0079] Among them, CC k This represents the relative proximity of the k-th Pareto solution. The closer the value is to 1, the closer the Pareto solution is to the ideal solution.

[0080] 4) Select the optimal solution

[0081] Based on the relative proximity values, the solutions are sorted from largest to smallest, and the Pareto solution with the highest relative proximity is selected as the final design parameter for the isolation pile construction.

[0082] Compared with the prior art, the application has the beneficial effects that:

[0083] (1) The application accurately determines the construction design parameters of the isolation pile by combining field engineering practice, using orthogonal test design and three-dimensional finite difference calculation. This method not only considers key parameters such as the distance between the pile and the tunnel, the pile top burial depth, the pile length, the pile diameter, and the pile spacing, but also simulates the whole process of foundation pit construction through a numerical analysis model. Compared with the traditional design method relying on experience, the application can significantly improve the scientificity and accuracy of parameter selection and ensure the rationality of the design.

[0084] (2) The application provides a more secure construction parameter optimization method for the possible impact on the existing tunnel during the foundation pit excavation process. Through three-dimensional numerical simulation and DE-GWO-SVM nonlinear mapping regression model, the horizontal displacement and other performance indicators of the tunnel can be accurately predicted. This prediction capability enables real-time adjustment of the isolation pile design parameters during construction, effectively controlling the deformation of the tunnel, avoiding structural damage, ensuring the safety of construction, and reducing the risk during the implementation of the project.

[0085] (3) The application further optimizes the design parameters of the isolation pile under the premise of meeting the safety requirements of the tunnel by introducing a multi-objective optimization framework based on DE-GWO-SVM-NSGA-II. The optimization objectives include not only the minimization of tunnel deformation but also the minimization of construction cost of the isolation pile. The isolation pile construction parameters determined by the TOPSIS method can significantly reduce the project cost while ensuring safety, achieving the purpose of improving economic efficiency.

[0086] (4) The method of the application is particularly suitable for complex urban underground space development needs. Through the optimization of isolation pile construction parameters under various geological conditions, the application demonstrates its adaptability in different geological environments. Whether there are differences in parameters such as soil thickness, elastic modulus, and Poisson's ratio, or the distance and depth between the foundation pit and the tunnel are different, the application can provide optimal construction design parameters to ensure the smooth implementation of the project under complex geological conditions and improve the fine management level of urban infrastructure construction. BRIEF DESCRIPTION OF DRAWINGS

[0087] The accompanying drawings, which are included to provide a further understanding of the application and constitute a part of this application, illustrate certain non-limiting embodiments of the application. In the drawings:

[0088] Figure 1 A method flowchart for determining the construction design parameters of the isolation pile between the newly built foundation pit and the existing tunnel is provided for the embodiments of the application.

[0089] Figure 2A three-dimensional model diagram for determining construction design parameters of isolation piles between a newly-built foundation pit and an existing tunnel is provided for an embodiment of the present application.

[0090] Figure 3 A location schematic diagram of isolation piles, an existing tunnel and a newly-built foundation pit is provided for an embodiment of the present application.

[0091] Figure 4 A principle block diagram of a method for determining construction design parameters of isolation piles between a newly-built foundation pit and an existing tunnel is provided for an embodiment of the present application.

[0092] In the figure, 1 represents an existing tunnel; 2 represents a newly-built foundation pit; 3 represents isolation piles; 4 represents a field soil body; and 5 represents isolation piles. DETAILED DESCRIPTION

[0093] To make the objectives, technical solutions and advantages of the present application clearer, further detailed description will be made to the present application in combination with embodiments and drawings. Herein, the illustrative embodiments of the present application and their descriptions are used to explain the present application but not to limit the present application.

[0094] Referring to Figure 1 The method for determining construction design parameters of isolation piles between a newly-built foundation pit and an existing tunnel provided by an embodiment of the present application comprises the following steps.

[0095] S1, determining and acquiring geological data and engineering information parameters related to the newly-built foundation pit and the existing tunnel, and acquiring physical and mechanical parameters of the soil body within the field range of the newly-built foundation pit and the existing tunnel.

[0096] In the implementation, the project background of the newly-built foundation pit adjacent to the existing tunnel is determined, the geological conditions of the construction site are surveyed, the geological data and engineering information parameters related to the newly-built foundation pit and the existing tunnel are acquired, the indoor mechanical test of each type of soil body in the field of the newly-built foundation pit and the existing tunnel is carried out, and the physical and mechanical parameters of the soil body within the field range of the newly-built foundation pit and the existing tunnel are acquired. Specifically,

[0097] The geological data and engineering information parameters include the length, width and depth of the proposed foundation pit, the radius, segment thickness, segment specific gravity, segment elastic modulus and segment Poisson's ratio of the existing tunnel, and the vertical distance between the edge of the newly-built foundation pit and the existing tunnel.

[0098] The physical and mechanical parameters include the soil layer type and thickness, elastic modulus, Poisson's ratio, unit weight, cohesion and friction angle.

[0099] S2, combining the acquired parameters with the actual engineering of the newly-built foundation pit and the existing tunnel, the construction design parameters affecting the bearing mechanism of the isolation pile are determined, and the orthogonal design scheme of different construction design parameters of the isolation pile is set through orthogonal test; based on three-dimensional finite difference calculation, the numerical analysis model of the newly-built foundation pit excavation under each design scheme is constructed according to the orthogonal design scheme, and the simulation calculation of the whole process of the foundation pit construction is carried out to obtain the performance evaluation index of the existing tunnel under different design schemes.

[0100] In implementation, the three-dimensional finite difference calculation program refers to the FLAC3D program, and the FISH language built in the program is used for modeling of the isolation pile, the existing tunnel and the newly-built foundation pit and construction simulation of each process.

[0101] The construction design parameters of the isolation pile include the distance between the isolation pile and the existing tunnel, the pile top burial depth, the pile length, the pile diameter and the pile spacing.

[0102] S3, taking the construction design parameters of the isolation pile as input and the performance evaluation index of the existing tunnel corresponding to the construction design parameters of the isolation pile as output, a sample database is built; the sample database is trained and learned by using the support vector machine DE-GWO-SVM, and the DE-GWO-SVM nonlinear mapping regression function is constructed.

[0103] In implementation, the support vector machine nonlinear mapping model is the support vector machine DE-GWO-SVM optimized by using the differential evolution-gray wolf algorithm. 80% of the sample database is used as the training data set, and the remaining 20% of the data is used as the test data set.

[0104] Specifically, the process of constructing the DE-GWO-SVM nonlinear mapping regression function includes:

[0105] 1) Preprocessing sample data in the sample database

[0106] In order to improve the convergence speed and accuracy of the model and eliminate the influence of the sample data size, the sample data is normalized before the model training, and the sample data is mapped to the interval [0, 1];

[0107] 2) Selecting a kernel function

[0108] In order to improve the accuracy of the model, the radial basis kernel function is used, and the expression is:

[0109]

[0110] In the formula, σ is the kernel width parameter, and ||x-x i || represents the distance between data x and the center point x i ;

[0111] 3) Determining the optimal parameters of SVM

[0112] In order to make the SVM model have higher prediction accuracy, combined with the pretreatment sample data, the differential evolution algorithm and grey wolf algorithm DE-GWO are used to optimize the hyperparameters of the SVM model, and the best penalty parameter and kernel function parameter are determined;

[0113] 4) Establishing DE-GWO-SVM nonlinear mapping regression function

[0114] Based on the best penalty parameter, kernel function parameter and sample data determined by DE-GWO, the SVM model is trained to establish the DE-GWO-SVM nonlinear mapping regression function. The training sample data is used to help simulate the establishment of the mathematical model, and the test sample data is used to test the prediction effect of the model.

[0115] In the implementation, in the process of establishing the DE-GWO-SVM nonlinear mapping regression function, the mean absolute error MAE, the mean square error MSE and the mean absolute percentage error MAPE are used as the evaluation criteria of the model, and the formulas are as follows:

[0116]

[0117]

[0118]

[0119] Wherein, N is the number of sample data, y i,pre is the predicted value of the model, y i,act is the corresponding actual value;

[0120] The smaller the numerical MAE, MSE and MAPE in the training process, the better the prediction fitting degree of the nonlinear mapping regression function.

[0121] S4, combine the DE-GWO-SVM nonlinear mapping regression function and the isolation pile cost cost function to construct a DE-GWO-SVM-NSGA-II multi-objective optimization framework with the minimum tunnel deformation and the minimum cost as the target, so as to optimize the construction design parameters of the isolation pile between the newly built foundation pit and the existing tunnel.

[0122] In the implementation, the process of constructing the DE-GWO-SVM-NSGA-II multi-objective optimization framework specifically includes:

[0123] 1) Establishing objective function

[0124] Define the objective function f i (X), in order to meet the tunnel deformation requirements and the best isolation pile cost, define two objective functions f1(X) and f2(X):

[0125] fi (X) = (f1(X), f2(X))

[0126] DE-GWO-SVM nonlinear mapping regression function replaces the traditional mathematical function as the first objective function f1(X) of NSGA-II, representing the minimum tunnel deformation, expressed as:

[0127] f1(X) = min[DE-GWO-SVM(X1, X2, X3, X4, X5)]

[0128] Where f1(X) is the minimum value of the existing tunnel horizontal position, DE-GWO-SVM(X1, X2, X3, X4, X5) is a nonlinear mapping regression function, X1 is the length of the isolation pile, X2 is the pile diameter, X3 is the pile spacing, X4 is the pile top buried depth, and X5 is the distance between the pile tunnels;

[0129] The completed isolation pile is regarded as a wall, and the total cost of unit length of wall material is taken as the second objective function f2(X), representing the minimum cost:

[0130]

[0131] Where f2(X) represents the total cost, C con is the unit volume cost of concrete of the isolation pile, C reb is the unit mass cost of steel reinforcement of the isolation pile, V con is the concrete volume of a single isolation pile, M lr and M sr are the longitudinal reinforcement mass and stirrup mass of a single isolation pile, respectively;

[0132] The total volume of a single isolation pile is approximately equal to the concrete volume V con , then:

[0133]

[0134] 2) Determine the constraint range of the isolation pile construction design parameters

[0135] To prevent the optimized structure of the isolation pile construction design parameters from deviating too much from the actual engineering parameters, the constraint range of the isolation pile construction design parameters is set according to relevant construction specifications and engineering practices:

[0136]

[0137] Where X i represents the i-th isolation pile construction design parameter, and are the upper limit and lower limit of the i-th isolation pile construction design parameter, respectively;

[0138] Specifically, the constraint range of the objective function is as follows:

[0139]

[0140] Where s.t. represents a constraint condition.

[0141] In implementation, the process of optimizing the construction design parameters of the isolation pile between the newly-built foundation pit and the existing tunnel includes:

[0142] The NSGA-II algorithm is used to perform multi-objective optimization on the construction design parameters of the isolation pile to obtain a Pareto optimal solution set, which specifically includes:

[0143] Step 1, determine the basic parameters

[0144] Determine the basic parameters that affect the running efficiency and solution quality of the NSGA-II algorithm, including population size, crossover rate, mutation rate, and iteration number;

[0145] Step 2, population initialization

[0146] Randomly generate an initial population, where each individual in the initial population represents a possible construction design parameter X of the isolation pile k ;

[0147] Step 3, fitness evaluation and non-dominated sorting

[0148] Calculate the objective function value for each individual in the population, and perform non-dominated sorting based on the objective function value to distinguish different optimization levels and obtain a Pareto frontier layer of multiple isolation pile construction design parameters. Each individual in each layer represents an optimal solution that is not dominated by other individuals in all objective functions;

[0149] Step 4, crowding degree calculation

[0150] For each individual in the Pareto frontier layer, calculate the crowding degree of the individual in the objective function space to ensure the diversity of the solution set. The crowding degree d i The calculation formula is as follows:

[0151]

[0152] Where d i represents the crowding degree of the i-th individual, represents the value of the i-th individual in the j-th objective function;

[0153] Step 5, genetic operation

[0154] Perform a selection strategy based on the crowding degree, and generate a new generation of population through crossover and mutation operations;

[0155] Step 6, iteration update

[0156] Repeat steps 1-5 until the number of iterations is met, and a Pareto optimal solution set is formed at the termination of the algorithm, which contains a series of design parameter combinations that achieve the best balance on the objective functions. The final isolation pile construction design parameters are determined based on the Pareto optimal solution set.

[0157] In implementation, the process of determining the final isolation pile construction design parameters based on the Pareto optimal solution set includes:

[0158] TOPSIS method is used to select the best solution from the Pareto optimal solution set to determine the final isolation pile construction design parameters, which specifically includes:

[0159] 1) Determine the ideal solution and the inferior solution

[0160] The ideal solution X+ is the solution that makes each objective function value optimal. For the existing tunnel deformation minimization and isolation pile cost minimization objectives, it is the solution with minimum deformation and lowest cost. The inferior solution X- is the opposite, which is the solution with maximum deformation and highest cost.

[0161] 2) Calculate the distance of the ideal solution and the inferior solution

[0162] For each Pareto solution in the Pareto optimal solution set, calculate its Euclidean distance to the ideal solution and the inferior solution, and the calculation formulas are as follows:

[0163]

[0164] Where d + (X k ) is the distance of the kth Pareto solution to the ideal solution, and the smaller the distance, the closer the Pareto solution is to the optimal state on all objective functions; d - (X k ) is the distance of the kth Pareto solution to the inferior solution, and the larger the distance, the farther the Pareto solution is from the worst state on all objective functions; X k represents the kth Pareto solution; f i (X k ) represents the value of the kth Pareto solution on the ith objective function; and m is the number of objective functions.

[0165] 3) Calculate the relative closeness

[0166] Calculate the relative closeness CC k of each Pareto solution:

[0167]

[0168] Where CC kThe relative closeness degree of the kth Pareto solution, the value is closer to 1, the closer the Pareto solution is to the ideal solution;

[0169] 4) Select the optimal solution

[0170] According to the value of the relative closeness degree, sort from large to small, select the Pareto solution with the highest relative closeness degree as the final isolation pile construction design parameter.

[0171] Embodiment:

[0172] (1) Through on-site investigation and research of the project, determine the geological profile data and engineering parameter information related to the new foundation pit and the existing tunnel.

[0173] (2) Through on-site sampling and laboratory testing, obtain the physical and mechanical parameters of various soil bodies in the construction site.

[0174] (3) Combined with the mechanical action mechanism of the isolation pile, take the pile length X1, pile diameter X2, pile spacing X3, pile top burial depth X4, and pile tunnel distance X5 as the isolation pile construction design parameters. According to the principle of orthogonal test, select the pile length X1, pile diameter X2, pile spacing X3, pile top burial depth X4, and pile tunnel distance X5 as the influencing factors. According to relevant specifications and engineering experience, determine the upper and lower limit ranges of the isolation pile design parameters. Divide the parameters into 5 levels according to the established range, as shown in Table 1. The numerical simulation scheme adopts the orthogonal table L25(5^6) of six factors and five levels, a total of 25 groups of numerical simulation schemes, and the isolation pile construction parameter orthogonal test table is shown in Table 2.

[0175] Table 1 Isolation pile design parameter factors and levels

[0176]

[0177]

[0178] Table 2 Orthogonal test scheme

[0179]

[0180] (4) Based on the above orthogonal test scheme, carry out numerical simulation analysis of the excavation of the new foundation pit, obtain the lateral deformation value of the existing tunnel under different test schemes, and the numerical analysis model is shown in Figure 2 . Take the isolation pile construction parameters as input variables and the displacement value of the existing tunnel as output variables to construct a sample database, and the total number of samples is 25.

[0181] The sample is normalized, and 25 data in the sample database are randomly divided into two groups, 80% as the training set and 20% as the test set, that is, 20 data are used for training the model, and 5 data are used for testing the prediction performance of the model. Select the pile length, pile diameter, pile spacing, pile top buried depth and pile tunnel distance as the input variables of the DE-GWO-SVM nonlinear prediction model, and the existing tunnel transverse displacement as the output variable. The DE-GWO algorithm is used to globally optimize the two key parameters in the SVM model, the penalty parameter C and the kernel function parameter g. To improve the accuracy of the model, the final result tolerance is set to 0.001. The existing tunnel deformation prediction model based on DE-GWO-SVM is constructed, and the algorithm specific process is as shown in Figure 4

[0182] According to the DE-GWO optimization result, the training set is used for nonlinear learning, and the test set is used for verification to establish the existing tunnel horizontal displacement prediction model. DE-GWO-SVM can well predict the displacement change of the existing tunnel when the new foundation pit is constructed. The evaluation index value of the DE-GWO-SVM model is calculated, among which the average absolute error MAE is 0.219, the mean square error MSE is 0.135, and the MAPE value is 1.047. It shows that the model has good prediction ability and good stability.

[0183] (5) Based on the DE-GWO-SVM model, NSGA-II is used to perform multi-objective optimization on the construction design parameters of the isolation pile. The optimization target is to minimize the horizontal displacement of the existing tunnel and the economic cost of the isolation pile, and the corresponding objective functions are

[0184] f1(X)=min[DE-GWO-SVM(X1,X2,X3,X4,X5)]

[0185]

[0186] (6) According to the relevant specifications and engineering practice, the constraint range of the construction parameters of the isolation pile is set as follows:

[0187]

[0188] (7) Multi-objective optimization based on NSGA-II. Based on the obtained multi-objective optimization function and the determined constraint range, the NSGA-II algorithm is used to perform multi-objective optimization on the construction parameters of the isolation pile. In this embodiment, the population size of the NSGA-II algorithm is set to 100, the iteration number is 200, the crossover distribution index is 20, the mutation probability is 0.1, the crossover probability is 0.9, and the mutation distribution index is 20. After determining the above algorithm parameters, the global optimization is performed according to the steps of the NSGA-II algorithm, and the Pareto optimal solution set is obtained. ​

[0189] (8) In order to select the optimal scheme from multiple decision schemes, the TOPSIS method is used to calculate the solution with the highest relative closeness in the Pareto optimal solution set as the optimal scheme of the construction parameters of the isolation pile.

[0190] The above merely describes preferred embodiments of the present application, but is not intended to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for determining the construction design parameters of isolation piles between a new foundation pit and an existing tunnel, characterized in that, include: S1. Determine and acquire geological data and engineering information parameters related to the new foundation pit and existing tunnel, and acquire the physical and mechanical parameters of the soil within the site of the new foundation pit and existing tunnel. S2. Using the obtained parameters and combining them with the actual site conditions of the new foundation pit and the existing tunnel, the construction design parameters that affect the bearing capacity mechanism of the isolation piles are identified, and orthogonal design schemes with different construction design parameters for the isolation piles are set through orthogonal experiments. Based on three-dimensional finite difference calculation and orthogonal design scheme, numerical analytical models of new foundation pit excavation under each design scheme are constructed, and simulation calculations of the entire foundation pit construction process are carried out to obtain the performance evaluation index of the existing tunnel under different design scheme working conditions. S3. Using the construction design parameters of the isolation piles as input and the existing tunnel performance evaluation index corresponding to the construction design parameters of the isolation piles as output, a sample database is built; the DE-GWO-SVM support vector machine is used to learn and train the sample database to construct the DE-GWO-SVM nonlinear mapping regression function. S4. Combine the DE-GWO-SVM nonlinear mapping regression function and the cost function of the isolation pile to construct a DE-GWO-SVM-NSGA-II multi-objective optimization framework with the objectives of minimizing the deformation and cost of the existing tunnel, thereby optimizing the construction design parameters of the isolation pile between the new foundation pit and the existing tunnel. The process of constructing the DE-GWO-SVM-NSGA-II multi-objective optimization framework in S4 specifically includes: 1) Establish the objective function Define the objective function To meet tunnel deformation requirements and optimize the cost of isolation piles, the following definition is provided. and Two objective functions: ； The DE-GWO-SVM nonlinear mapping regression function replaces the traditional mathematical function as the first objective function of NSGA-II. , representing the minimization of tunnel deformation, is expressed as: ； in, The minimum horizontal position of the existing tunnel, DE-GWO- X1 is a nonlinear mapping regression function, where X1 is the length of the isolation pile, X2 is the pile diameter, X3 is the pile spacing, X4 is the pile top embedment depth, and X5 is the distance between the pile and the tunnel. Treating the completed isolation piles as walls, the total cost of wall material per unit length is taken as the second objective function. This represents cost minimization: ； in, Indicates the total cost. The cost per unit volume of concrete for the isolation piles. The cost of steel reinforcement per unit weight for isolation piles. The concrete volume of a single isolation pile, and These refer to the quality of the longitudinal reinforcement and stirrups of a single isolation pile, respectively. The total volume of a single isolation pile is approximated by the volume of concrete. Then we have: ； 2) Determine the constraint range of the construction design parameters for the isolation piles. To prevent the optimized structural parameters of the isolation pile construction design from deviating too much from the actual engineering parameters, constraints on the isolation pile construction design parameters are set according to relevant construction specifications and actual engineering conditions: ； Among them, X i This represents the construction design parameters for the i-th isolation pile. and These are the upper and lower limits of the construction design parameters for the i-th isolation pile, respectively. Specifically, the constraints of the objective function are as follows: ； Where st represents the constraint condition.

2. The method for determining the construction design parameters of the isolation piles between the new foundation pit and the existing tunnel as described in claim 1, characterized in that, Geological data and engineering information parameters include: the length, width and depth of the proposed foundation pit, the radius, segment thickness, segment unit weight, segment elastic modulus and segment Poisson's ratio of the existing tunnel, and the vertical distance between the edge of the new foundation pit and the existing tunnel; The physical and mechanical parameters include: soil type and thickness, elastic modulus, Poisson's ratio, unit weight, cohesion, and friction angle.

3. The method for determining the construction design parameters of the isolation piles between the new foundation pit and the existing tunnel as described in claim 1, characterized in that, The design parameters for the construction of isolation piles include: the distance between the isolation pile and the existing tunnel, the pile top burial depth, pile length, pile diameter, and pile spacing.

4. The method for determining the construction design parameters of the isolation piles between the new foundation pit and the existing tunnel as described in claim 1, characterized in that, The specific process of constructing the DE-GWO-SVM nonlinear mapping regression function in S3 includes: 1) Preprocess sample data in the sample database To improve the convergence speed and accuracy of the model and eliminate the influence of sample data size, the sample data is normalized before model training, mapping the sample data to the interval [0, 1]. 2) Selecting the kernel function To improve the accuracy of the model, a radial basis function kernel is used, with the following expression: ； In the formula, For kernel width parameter, Representing data Distance from the center point The distance; 3) Determine the optimal parameters for SVM To improve the prediction accuracy of the SVM model, the differential evolution algorithm and the Grey Wolf algorithm (DE-GWO) are used to optimize the hyperparameters of the SVM model by combining preprocessed sample data, and the optimal penalty parameters and kernel function parameters are determined. 4) Establish the DE-GWO-SVM nonlinear mapping regression function The SVM model is trained based on the optimal penalty parameters, kernel function parameters, and sample data determined by DE-GWO, and a DE-GWO-SVM nonlinear mapping regression function is established.

5. The method for determining the construction design parameters of the isolation piles between the new foundation pit and the existing tunnel as described in claim 4, characterized in that, In establishing the DE-GWO-SVM nonlinear mapping regression function, the mean absolute error (MAE), mean squared error (MSE), and mean absolute percentage error (MAPE) are used as evaluation criteria for the model, as shown in the following formulas: ； ； ； Where N is the number of sample data. The predicted value of the model. This corresponds to the actual value; The smaller the numerical values ​​of MAE, MSE, and MAPE during the training process, the better the prediction fit of the nonlinear mapping regression function.

6. The method for determining the construction design parameters of the isolation piles between the new foundation pit and the existing tunnel as described in claim 1, characterized in that, The process of optimizing the construction design parameters of the isolation piles between the new foundation pit and the existing tunnel in S4 includes: The NSGA-II algorithm was used to perform multi-objective optimization of the construction design parameters of the isolation piles, resulting in the Pareto optimal solution set, which includes: Step 1, Determine basic parameters Determine the basic parameters that affect the running efficiency and solution quality of the NSGA-II algorithm, including population size, crossover rate, mutation rate, and number of iterations; Step 2, Population Initialization An initial population is randomly generated, where each individual in the initial population represents a possible design parameter for the construction of isolation piles; Step 3: Fitness Assessment and Non-Dominant Ranking The objective function value is calculated for each individual in the population, and non-dominated sorting is performed based on the objective function value to distinguish different optimization levels, resulting in Pareto front layers of multiple isolation pile construction design parameters. Individuals in each layer represent the optimal solution that is not dominated by other individuals among all objective functions. Step 4: Crowding Calculation For each individual in the Pareto front layer, calculate the crowding degree of that individual in the objective function space to ensure the diversity of the solution set. The calculation formula is as follows: ； in, This represents the crowding level of the i-th individual. This represents the value of the i-th individual on the j-th objective function; Step 5, Genetic manipulation Implement a crowding-based selection strategy and generate a new generation of population through crossover and mutation operations; Step 6: Iterative Update Repeat steps 1-5 until the required number of iterations is met. When the algorithm terminates, a Pareto optimal solution set is formed. The Pareto optimal solution set contains a series of design parameter combinations that achieve the best balance on the objective function. The final construction design parameters for the isolation piles are determined based on the Pareto optimal solution set.

7. The method for determining the construction design parameters of the isolation piles between the new foundation pit and the existing tunnel as described in claim 6, characterized in that, The process of determining the final construction design parameters for the isolation piles based on the Pareto optimal solution set includes: The TOPSIS method is used to calculate the optimal solution for the Pareto optimal solution set in order to determine the final design parameters for the isolation pile construction. Specifically, this includes: 1) Determine the ideal solution and the inferior solution The ideal solution X+ is the solution that optimizes each objective function value. For the objectives of minimizing existing tunnel deformation and minimizing isolation pile cost, it is the solution with the minimum deformation and the lowest cost. The inferior solution X-, on the other hand, selects the solution with the maximum deformation and the highest cost. 2) Calculate the distance between the ideal solution and the inferior solution. For each Pareto solution in the Pareto optimal solution set, calculate its Euclidean distance to the ideal and inferior solutions, using the following formulas: ； ； in, The smaller the distance from the k-th Pareto solution to the ideal solution, the closer the Pareto solution is to the optimal state across all objective functions. The distance from the k-th Pareto solution to the worst solution is the greater the distance, the further the Pareto solution is from the worst state across all objective functions. This represents the k-th Pareto solution; This represents the value of the k-th Pareto solution on the i-th objective function; m is the number of objective functions; 3) Calculate the relative closeness Calculate the relative proximity of each Pareto solution. : ； in, This represents the relative proximity of the k-th Pareto solution. The closer the value is to 1, the closer the Pareto solution is to the ideal solution. 4) Select the optimal solution Based on the relative proximity values, the solutions are sorted from largest to smallest, and the Pareto solution with the highest relative proximity is selected as the final design parameter for the isolation pile construction.

Images