Shield tunneling deformation optimization method and device based on intelligent algorithm

Abstract

The application belongs to the technical field of shield construction control, and particularly discloses a shield underpass existing tunnel deformation optimization method and equipment based on an intelligent algorithm. The method comprises the following steps: establishing an index system and constructing a sample data set; establishing a prediction model of tunnel lining top transverse and vertical deformation and tunnel lining top transverse and vertical deformation rate based on BO-CNN-three layer LSTM; taking a nonlinear mapping relation function obtained by the prediction model as a target optimization fitness function, constructing a multi-objective optimization model, and proposing a construction parameter setting suggestion range under the condition of minimum deformation. The application can realize multi-objective optimization of tunnel lining top transverse and vertical deformation and tunnel lining top transverse and vertical deformation rate, so that the optimization result can be used as an effective tool for shield approach tunnel construction parameter regulation and control and existing tunnel deformation regulation and control.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of shield construction control, and more particularly to a shield under-passing existing tunnel deformation optimization method and equipment based on an intelligent algorithm. BACKGROUND

[0002] The multi-objective optimization control of existing tunnel deformation caused by shield close under-passing construction mainly includes multi-objective optimization of existing tunnel deformation and optimization of construction parameters. There are generally several types of under-passing existing tunnels, including over-passing, under-passing and side-passing, among which under-passing is the most common. During the under-passing process of the shield, soil disturbance will be caused, the displacement field of the stratum will be transmitted to the existing tunnel, the original soil balance will be broken, and the existing tunnel structure and stratum will undergo vertical displacement, horizontal displacement and other deformations. When the longitudinal and horizontal displacements reach the limit deformation degree, the use function and structural safety of the existing tunnel structure will be affected. Many researchers have conducted extensive research on the response of the existing tunnel in under-passing engineering and have proposed a series of control measures, which greatly reduce the risk of the existing tunnel. At present, the research on the deformation of the existing tunnel caused by shield under-passing mainly focuses on theoretical formulas, model tests, field monitoring, numerical simulation, data-driven machine learning (ML) and other methods. The defects of the existing research on the deformation control of under-passing existing tunnels mainly include: (1) the traditional theoretical research, model test and numerical simulation consider limited factors, and inevitably involve simplification, often ignoring some influencing factors, resulting in inaccurate results. At the same time, a large amount of calculation work is time-consuming; (2) the field monitoring method cannot predict the actual situation in advance and requires processing a large amount of data; (3) the existing ML methods mainly focus on the prediction of the deformation and stress of the existing tunnel, and the research on the multi-objective optimization control of the under-passing existing tunnel deformation is extremely rare. In addition, in the construction of under-passing shield tunnels, the horizontal distance (HDFCAUT) between the front end of the tunneling machine and the center axis of the intersection point of the new and old tunnels and the health status of the existing tunnel have important influences on the deformation of the existing tunnel and the safe tunneling of the new tunnel. The existing research rarely considers the control of the deformation of the existing tunnel and the optimization of the construction parameters of the new tunnel under the influence of the health degree of the existing tunnel, different tunneling intervals and different time periods. That is, how to consider the regulation and control of the shield parameters before, during and after the shield under-passing to control the deformation of the existing tunnel is a problem worth continuous attention.

[0003] To overcome the limitations of traditional methods, ML has been widely used in shield-related parameters and deformation prediction in recent years, such as NGBoost, CNN, LSTM, etc. Wu, X., et al. used BO-NGBoost to predict the deformation of existing tunnels. Ye, X.-W., et al. used LSTM model to establish an intelligent prediction relationship between new tunnel excavation parameters and existing tunnel deformation. Although ML algorithms can obtain better evaluation results than traditional prediction methods, they have some shortcomings, NGBoost cannot handle complex problems that need to consider the influence of space and time; CNN has a convolutional weight-sharing structure that reduces the complexity of the neural network model while significantly reducing the number of neural network parameters, making it have strong feature self-learning ability and prevent overfitting. However, CNN ignores the long-term dependence between sequence data on the time level. LSTM can be used to solve time series problems, but it is not good at processing data with spatial features, and single-layer LSTM will have the problem of gradient disappearance or explosion as the output sequence grows, which needs to be supplemented by other measures, such as using multi-layer LSTM that can solve complex time series problems. SUMMARY

[0004] In view of the above defects or improvement needs of the prior art, the present application provides a shield close underpass construction caused by existing tunnel deformation multi-objective optimization control method and equipment based on BO-CNN-three layer LSTM-NSGA-III, which solves the problems of shield close underpass construction caused by existing tunnel deformation multi-objective optimization control and construction parameter optimization. The BO-CNN-three layer LSTM model is used to simulate and train the monitored sample data to provide a high-precision prediction function for NSGA-III optimization. Then, NSGA-III is used for multi-objective optimization to obtain the optimal shield segment parameter value, achieving the purpose of shield close underpass construction caused by existing tunnel deformation multi-objective optimization control.

[0005] To achieve the above object, according to one aspect of the present application, a shield tunneling under existing tunnel deformation optimization method and equipment based on Bayesian (BO) optimization combined with CNN-three layer LSTM and elitist non-dominated sorting genetic algorithm (NSGA)-III are provided, which mainly include the following steps: Step one, determining different approaching sections of shield tunneling near a tunnel, sensitive factors causing tunnel deformation, establishing an index system, collecting actual monitoring data in the shield tunneling process, and constructing a sample data set; Step two, based on the sample data set, establishing a prediction model of tunnel lining top transverse and vertical deformation and tunnel lining top transverse and vertical deformation rate based on BO-CNN-three layer LSTM and testing; Step three, taking the nonlinear mapping relationship function obtained by the prediction model as the objective optimization fitness function, constructing a multi-objective optimization model based on BO-CNN-three layer LSTM-NSGA-III, and proposing a recommended range of construction parameter settings under the condition of minimum deformation. The present application combines BO-CNN-three layer LSTM and NSGA-III, and through high-precision prediction of tunnel lining top transverse and vertical deformation and tunnel lining top transverse and vertical deformation rate, it can realize multi-objective optimization of tunnel lining top transverse and vertical deformation and tunnel lining top transverse and vertical deformation rate, so that the optimization result is more in line with the actual engineering demand, and it can be used as an effective tool for shield tunneling near a tunnel construction parameter regulation and existing tunnel deformation regulation.

[0006] Specifically, according to one aspect of the present application, a shield tunneling under existing tunnel deformation optimization method based on intelligent algorithm is provided, which includes the following steps:

[0007] Step one, determining different approaching sections of shield tunneling near a tunnel, sensitive factors causing tunnel deformation, establishing an index system, collecting actual monitoring data in the shield tunneling process, and constructing a sample data set;

[0008] Step two, establishing a prediction model of tunnel lining top transverse and vertical deformation and tunnel lining top transverse and vertical deformation rate based on BO-CNN-three layer LSTM, and training the prediction model using the sample data set to obtain a nonlinear mapping relationship function of tunnel lining top transverse and vertical deformation and tunnel lining top transverse and vertical deformation;

[0009] Step three, taking the nonlinear mapping relationship function as the objective optimization fitness function, constructing a multi-objective optimization model based on BO-CNN-three layer LSTM-NSGA-III, and proposing an optimal shield segment parameter setting recommendation value.

[0010] As a further preferred, in step one, the sensitive factors include the soil bin pressure, the cutter head rotating speed, the cutter head torque, the advancing speed, the total thrust, the auger speed, the auger torque, the grouting pressure, the horizontal distance from the working face to the intersection axis of the new and old tunnels.

[0011] The actual monitoring data include the transverse and vertical deformation of the top of the existing tunnel lining, the transverse and vertical deformation rate of the top of the tunnel lining.

[0012] As a further preferred, in step two, the grid search method is used to optimize the key parameters of the BO-CNN-three layer LSTM, and then the 5-fold cross-validation method is used to select the optimal parameter combination, and the BO-CNN-three layer LSTM prediction model is established based on the optimal parameter combination.

[0013] Among them, the BO is used to optimize the four hyperparameters of learning_rate, minibatch_frac, col_sample and n_estimators of the model.

[0014] As a further preferred, step two further comprises:

[0015] The root mean square error, the mean absolute error and the mean absolute percentage error are used to evaluate the reliability of the prediction model.

[0016] As a further preferred, step three comprises the following steps:

[0017] (31) Determine the fitness function, and take the nonlinear mapping relationship function as the target optimization fitness function:

[0018] min1=min(BO-CNN-three layer LSTM regression (X i ))

[0019] min2=min(BO-CNN-three layer LSTM regression (X i ))

[0020] min3=min(BO-CNN-three layer LSTM regression (X i ))

[0021] ming4=min(BO-CNN-three layer LSTM regression (X i ))

[0022] In the formula, X iThe parameters represent the soil chamber pressure, cutterhead rotation speed, cutterhead torque, propulsion speed, total thrust, screw conveyor speed, screw conveyor torque, grouting pressure, and horizontal distance from the working face to the intersection axis of the old and new tunnels. ming1 is the objective function for the horizontal displacement of the arch crown, ming2 is the objective function for the vertical displacement of the arch crown, ming3 is the objective function for the deformation rate of the vertical displacement of the arch crown, and ming4 is the objective function for the deformation rate of the horizontal displacement of the arch crown.

[0023] (32) Determine the constraint range and set constraints on the values ​​of the target decision parameters based on engineering requirements and existing project data:

[0024]

[0025] in, and These represent the lower and upper limits of the input parameter values, respectively.

[0026] (33) The NSGA-Ⅲ algorithm is used to realize the multi-objective optimization and control of existing tunnel deformation caused by shield tunneling near-underpass construction and the optimization of construction parameters.

[0027] As a further preferred approach, the NSGA-III algorithm is used to address the multi-objective optimization and control of existing tunnel deformation and the optimization of construction parameters caused by shield tunneling near-passage construction. The specific steps include:

[0028] (331): Set the external solution set This is used to store the algorithm and all currently obtained Pareto nondominated solutions, and to calculate each weight vector λ. j Generate a distance matrix d using Euclidean distances to other vectors. Select the T weight vectors with the smallest Euclidean distances. For each i = 1, ..., N, let B(i) = {i1, i2, ..., i...} T}, λ 1 ,…,λ T For λ i The most recent T weight vectors are used to randomly initialize the population {x}. 1 ,x 2 ,…,x N}, x j For the solution to the j-th subproblem, for each individual x in the population j , j = 1, 2, ..., N, calculate its objective function value FV j =F(x) j The ideal point z is initialized based on the MOP calculation. * =(z1,z2,…,z) m ),satisfy j∈{1,2,…,N P};

[0029] (332) Randomly select two individuals x k and x l from the set B(i), use a specific recombination operator to make the parents x k and x l cross and mutate to generate a new child x new , use a repair or heuristic method to repair x new into x re , evaluate the new child x re , calculate the F(x re ) function value, and update the ideal point z * according to the following rules:

[0030]

[0031] Update the neighborhood solution, if g tch (x re | λ l , z) ≤ g tch (x l | λ l , z) is satisfied, execute the replacement of the individual in the neighborhood, let x l = x re , FV l = F(x re ). Update the EP, judge the dominance relationship of all individuals with x re , remove all dominated vectors of x re from the EP, and move all non-dominated vectors of x re into the EP;

[0032] (333) Repeat (332) until the number of iterations is reached, and output the vertical deformation and the like as the Pareto solution set.

[0033] As a further preferred, a fuzzy decision method is used to select an optimal solution from the Pareto solution set, specifically as follows:

[0034] According to the distance of each scheme to the positive ideal solution and the negative ideal solution, the score of each scheme is calculated according to the following formula:

[0035]

[0036] In the formula, S i is the score corresponding to the i-th Pareto solution, f i - is the distance of the i-th Pareto solution to the negative ideal solution, f i + is the distance of the i-th Pareto solution to the positive ideal solution, f i - and fi + The calculation formula is as follows:

[0037]

[0038] In the formula, m is the number of optimization targets, x ij is the i-th Pareto solution under the j-th target, and is the input value corresponding to the negative ideal solution and the positive ideal solution under the j-th target, and the calculation formula is as follows:

[0039]

[0040] The improvement percentage of each Pareto solution is calculated, wherein the improvement of each target is:

[0041]

[0042] The overall improvement of each scheme is:

[0043]

[0044] In the formula, Imp ij is the improvement of the j-th target of each i-th Pareto solution, Org j,ave is the original average value of the j-th target, and Opt ij is the optimization value of the j-th target of each i-th Pareto solution, and Impoverall i is the overall improvement of each i-th Pareto solution.

[0045] According to another aspect of the present application, there is also provided a shield tunneling under existing tunnel deformation optimization system based on intelligent algorithm, comprising: a first master control module for determining different approaching sections of shield tunneling near a tunnel, sensitive factors causing tunnel deformation, establishing an index system, collecting actual monitoring data in the shield tunneling process, and constructing a sample data set; a second master control module for establishing a prediction model of tunnel lining top transverse and vertical deformation and tunnel lining top transverse and vertical deformation rate based on BO-CNN-three layer LSTM, and training the prediction model using the sample data set to obtain a nonlinear mapping relationship function of tunnel lining top transverse and vertical deformation and tunnel lining top transverse and vertical deformation rate; and a third master control module for taking the nonlinear mapping relationship function as a target optimization fitness function, constructing a multi-objective optimization model based on BO-CNN-three layer LSTM-NSGA-III, and proposing an optimal shield segment parameter setting suggestion value.

[0046] According to another aspect of the present application, there is also provided an electronic device comprising:

[0047] at least one processor, at least one memory, and a communication interface; wherein,

[0048] the processor, the memory, and the communication interface communicate with each other;

[0049] the memory stores program instructions executable by the processor, and the processor invokes the program instructions to perform the intelligent algorithm-based deformation optimization method for shield underpassing existing tunnel according to any embodiment or combination of multiple embodiments.

[0050] According to another aspect of the present application, there is also provided a non-transitory computer readable storage medium storing computer instructions, which cause the computer to perform the intelligent algorithm-based deformation optimization method for shield underpassing existing tunnel according to any embodiment or combination of multiple embodiments.

[0051] Overall, compared with the prior art, the above technical solutions conceived by the present application mainly have the following technical advantages:

[0052] 1. The present application can combine decision trees and non-parametric statistics based on discrete or continuous data sets to form regression or classification functions by using BO-CNN-three layer LSTM, which can well avoid overfitting, has strong generalization, realizes high-precision prediction of existing tunnel deformation, and realizes multi-objective optimization control of existing tunnel deformation caused by shield close underpassing construction and optimization of construction parameters.

[0053] 2. The present application uses NSGA-III with fast non-dominated sorting operator, individual crowding distance operator and elite strategy selection operator to establish a multi-objective optimization model, which can quickly find the optimal scheme and perform selective sorting, greatly improving the efficiency and accuracy of optimization.

[0054] 3. The present application replaces the traditional mathematical model with the BO-CNN-three layer LSTM prediction model, which is used as the fitness function of target optimization, and is combined with NSGA-III, which well reflects the complex nonlinear mapping relationship between shield construction parameters and existing tunnel deformation, and provides a basis for realizing multi-objective optimization control of existing tunnel deformation caused by shield close underpassing construction and accurate optimization of construction parameters.

[0055] 4. The present application simultaneously uses the horizontal and vertical deformation of the top of the existing tunnel lining, the horizontal and vertical deformation rate of the top of the tunnel lining as the optimization target for multi-objective optimization, which can comprehensively consider the contradiction relationship between multiple objectives, so that the optimized scheme is more in line with the actual engineering requirements. BRIEF DESCRIPTION OF DRAWINGS

[0056] Figure 1 Flowchart of the method of the present application.

[0057] Figure 2 Shield tunnel green intelligent design method based on BO-CNN-three layer LSTM-NSGA-III provided by the embodiment of the present application and equipment structure schematic diagram.

[0058] Figure 3 Physical structure schematic diagram of the electronic equipment provided by the embodiment of the present application.

[0059] Figure 4 BO-CNN-three layer LSTM prediction tunnel lining vault vertical displacement hyperparameter optimization result graph in the embodiment of the present application.

[0060] Figure 5 BO-CNN-three layer LSTM prediction tunnel lining vault horizontal displacement hyperparameter optimization result graph in the embodiment of the present application.

[0061] Figure 6 BO-CNN-three layer LSTM prediction tunnel lining vault vertical displacement deformation rate hyperparameter optimization result graph in the embodiment of the present application.

[0062] Figure 7 BO-CNN-three layer LSTM prediction tunnel lining vault horizontal displacement deformation rate hyperparameter optimization result graph in the embodiment of the present application.

[0063] Figure 8 BO-CNN-three layer LSTM prediction tunnel vault vertical displacement training curve in the embodiment of the present application.

[0064] Figure 9 BO-CNN-three layer LSTM prediction tunnel vault horizontal displacement training curve in the embodiment of the present application.

[0065] Figure 10 BO-CNN-three layer LSTM prediction tunnel vault vertical displacement deformation rate training curve in the embodiment of the present application.

[0066] Figure 11 BO-CNN-three layer LSTM prediction tunnel vault horizontal displacement deformation rate training curve in the embodiment of the present application.

[0067] Figure 12is a BO-CNN-three layer LSTM predicted tunnel vault vertical displacement prediction result graph in the embodiment of the application;

[0068] Figure 13 is a BO-CNN-three layer LSTM predicted tunnel vault horizontal displacement prediction result graph in the embodiment of the application;

[0069] Figure 14 is a BO-CNN-three layer LSTM predicted tunnel vault vertical displacement prediction result graph in the embodiment of the application;

[0070] Figure 15 is a BO-CNN-three layer LSTM predicted tunnel vault vertical displacement prediction result graph in the embodiment of the application;

[0071] Figure 16 is a Pareto solution of four targets of each section in the embodiment of the application. DETAILED DESCRIPTION

[0072] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application. In addition, the technical features involved in each embodiment of the application described below can be combined with each other as long as they do not conflict with each other.

[0073] In order to achieve the above-mentioned object, the present application provides a multi-objective optimization control method for deformation of an existing tunnel caused by shield close underpass construction based on BO-CNN-three layer LSTM-NSGA-III, which is used to solve the problems of multi-objective optimization control and construction parameter optimization of deformation of an existing tunnel caused by shield close underpass construction. The BO-CNN-three layer LSTM model is used to simulate and train the monitored sample data to provide a high-precision prediction function for NSGA-III optimization. Then, NSGA-III is used for multi-objective optimization to obtain the optimal shield segment parameter value, thereby achieving the purpose of multi-objective optimization control of deformation of an existing tunnel caused by shield close underpass construction. The method comprises the following steps:

[0074] (1) Construction of influencing factor index system

[0075] Through a large number of engineering practices and literature research, the existing tunnel deformation evaluation index system includes two parts of influencing factors and evaluation indexes. The sensitive factors mainly involve three categories of segment parameters, shield construction parameters and geological parameters, including soil chamber pressure, cutter head speed, cutter head torque, pushing speed, total thrust, screw machine speed, screw machine torque, grouting pressure, and horizontal distance from working face to new and old tunnel intersection axis. The existing tunnel deformation and cost evaluation indexes are the horizontal and vertical deformations of the tunnel lining top of the existing tunnel, and the horizontal and vertical deformation rates of the tunnel lining top.

[0076] (2) BO-CNN-three layer LSTM prediction model establishment and evaluation

[0077] S1: Data processing and sample division

[0078] By collecting actual monitoring data, a total of 221 typical sample data were obtained. In order to better train and verify the model, 80% of the sample data were randomly selected as the training set, and the remaining 20% were used to test the model. The sample data were normalized between [-1, 1] to eliminate the influence of different sample feature sizes on prediction accuracy and efficiency. The calculation formula of the normalization processing is as follows:

[0079]

[0080] where x nor represents the normalized standard value, and are 1 and 0 by default, respectively, representing the maximum and minimum values after normalization, x represents the sample value, x max and x min are the maximum and minimum values of the sample value, respectively.

[0081] S2: Model parameter optimization

[0082] In order to prevent the CNN-three layer LSTM model from under-learning and ensure good generalization ability of the model, the hyperparameters of the model need to be optimized to obtain better prediction results. The optimal parameter combination with the highest prediction accuracy is selected. BO is used to optimize the main hyperparameters of the CNN-three layer LSTM model (see Table 1 for meanings). Based on the Python platform, the training model is established, the MSE is used as the loss function, the number of training iterations is set to 200, and the hyperparameter combination with the best training performance is found.

[0083] Table 1 BO-CNN-three layer LSTM model hyperparameter meanings and ranges

[0084]

[0085] In this paper, BO is used for hyperparameter selection, and the MSE of K-fold cross-validation is used as the selection basis to find the parameters corresponding to the model with the highest accuracy, so as to determine the final preferred parameters.

[0086] S3: Prediction model establishment and evaluation

[0087] The sample training set is used to train the BO-CNN-three layer LSTM prediction model of both the tunnel lining top transverse and vertical deformation and the tunnel lining top transverse and vertical deformation rate, and then the test set is used to test the generalization performance and prediction accuracy of the BO-CNN-three layer LSTM prediction model of both the tunnel lining top transverse and vertical deformation and the tunnel lining top transverse and vertical deformation rate.

[0088] In order to effectively evaluate the prediction performance of the prediction model, the root mean square error (RMSE), the mean absolute error (MAE), and the mean absolute percentage error (MAPE) are used to evaluate the reliability of the model, and the calculation formulas of the three performance indicators are as follows:

[0089]

[0090] In the formula, n is the number of data samples, and are the detection value and the prediction value, respectively.

[0091] (3) NSGA-Ⅲ multi-objective optimization

[0092] S1: Determine the objective function and constraint condition

[0093] Using BO-CNN-three layer LSTM existing tunnel arch top settlement, arch bottom horizontal displacement, arch bottom vertical displacement and settlement / d regression prediction algorithm to replace the traditional mathematical function as the objective function in the multi-objective genetic algorithm can well solve the problem that the specific function expression cannot be given due to the complex nonlinear relationship between the input variables and the output targets. The objective function of the BO-CNN-three layer LSTM existing tunnel lining top transverse and vertical deformation and the tunnel lining top transverse and vertical deformation rate can be represented as:

[0094]

[0095] S2: Determine the constraint condition

[0096] In order to make the generated shield construction parameter combination more reasonable and feasible, it is also necessary to set a limit range for each decision variable according to the actual situation of the project and relevant specifications to form the constraint condition of the variable, and the general form of the constraint condition is as follows:

[0097]

[0098] in, and These represent the lower and upper limits of the target parameter values, respectively.

[0099] S3:NSGA-III Multi-Objective Optimization

[0100] The NSGA-III algorithm is used to solve the multi-objective optimization and control of existing tunnel deformation and construction parameter optimization problems caused by shield tunneling near-passage construction. The specific implementation steps are as follows:

[0101] Step 1: Initialization. Setting up the external solution set (elite population). This is used to store the algorithm and all currently calculated Pareto non-dominated solutions. Calculate each weight vector λ. j Generate a distance matrix d using Euclidean distances to other vectors, and select the T weight vectors with the smallest Euclidean distances. For each i = 1, ..., N, let B(i) = {i1, i2, ..., ...} i T},λ 1 ,…,λ T For λ i The most recent T weight vectors. Randomly initialize the population {x}. 1 ,x 2 ,…,x N}, x j This is the solution to the j-th subproblem. For each individual x in the population... j , j = 1, 2, ..., N, calculate its objective function value FV j =F(x) j The ideal point z is initialized based on the MOP calculation. * =(z1,z2,…,z) m ),satisfy j∈{1,2,…,N P}

[0102] Step 2: Individual Update. Randomly select two individuals x from set B(i). k and x l Using a specific recombination operator, the parent individual x k and x l Gene crossover and mutation are performed to generate new offspring individuals x. new Use repair or heuristic methods to fix x new Repair to x re For the new sub-individual x re Perform function evaluation and calculate F(x) re The function value is updated according to the following rules for the ideal point z. * :

[0103]

[0104] Update the neighborhood solution. If g is satisfied tch (x re λ l ,z)≤g tch (x l |λ l If x ∈ (x, z), then the replacement of individuals in the neighborhood is performed, let x ∈ (x, z). l =x re FV l =F(x) re Update EP and determine the relationship between all individuals and x. re The dominance relationship, removing all those by x from EP re Dominant vector, move in all vectors not controlled by x re Dominant vector.

[0105] Step 3: Repeat step 2 until the required number of iterations is reached, and output the vertical deformation, etc., as the Pareto solution set.

[0106] (4) Obtaining the optimal solution using the TOPSIS method

[0107] The Pareto optimal solution obtained using the NSGA-III algorithm is a set of solutions that meets the requirements of multi-objective optimization, corresponding to multiple objective parameter decision schemes. However, in actual tunnel engineering, only one decision scheme is needed to guide the multi-objective optimization and control of existing tunnel deformation caused by shield tunneling near-underpass construction, as well as the optimization of construction parameters. Therefore, a method is required to select an optimal solution from the Pareto solution set. Fuzzy decision-making is used to select the optimal decision scheme, avoiding the influence of subjective human preferences on the results during the selection process, thus obtaining a more scientific and objective result. The implementation process is as follows:

[0108] The Pareto optimal solution obtained using the NSGA-III algorithm is a set of solutions that satisfy the Pareto optimal state decision variables. However, in actual tunnel construction, often only one decision scheme is needed to guide the tunnel boring machine's excavation. Therefore, the TOPSIS method is used to select the optimal solution from the Pareto solution set. The core of this method is to score each scheme based on its distance from the positive and negative ideal solutions. The scoring formulas for each scheme are shown below:

[0109]

[0110] In the formula, S i f is the score corresponding to the i-th Pareto solution. i - f is the distance from the i-th Pareto solution to the negative ideal solution. i +Distance of the i-th Pareto solution to the positive ideal solution. i - And f i + The calculation formula is shown in (8) and (9):

[0111]

[0112] In the formula, m is the number of optimization targets, x ij is the i-th Pareto solution under the j-th target, and are the input values corresponding to the negative ideal solution and the positive ideal solution under the j-th target, and the calculation formula is shown in (10) and (11):

[0113]

[0114] In order to better present the improvement degree of the optimization scheme, the improvement percentage of each Pareto solution is calculated. The improvement of each target is calculated by formula (12), and the overall improvement of each scheme is calculated by formula (13):

[0115]

[0116] In the formula, Imp ij is the improvement of the j-th target of each i-th Pareto solution, Org j,ave is the original average value of the j-th target, Opt ij is the optimization value of the j-th target of each i-th Pareto solution, Impoverall i is the overall improvement of each i-th Pareto solution. For the existing tunnel deformation selected in the present study, the greater the deformation, the more dangerous it is. Therefore, a minimum type index can be used to evaluate the target in TOPSIS.

[0117] (5) BO-CNN-three layer LSTM-NSGA-Ⅲ multi-objective optimization result evaluation

[0118] To verify the effectiveness of the BO-CNN-three layer LSTM-NSGA-Ⅲ optimization model, the average values of the monitored shield parameters are compared with the optimized average values, the optimization ratios of each parameter are calculated, and the feasibility of the provided method is proved. The comparison between the optimized data and the construction monitoring data proves that the method is effective and feasible for multi-objective optimization in actual engineering.

[0119] More specifically, in one embodiment of the present application,

[0120] The BO-CNN-three layer LSTM-NSGA-III-based method and equipment for multi-objective optimization control of existing tunnel deformation caused by shield close underpass construction and optimization of construction parameters mainly include the following steps:

[0121] (1) Construction of influencing factor index system and collection of sample data

[0122] Through a large number of engineering practices and literature researches, nine influencing factors including soil bin pressure, cutterhead speed, cutterhead torque, advancing speed, total thrust, screw machine speed, screw machine torque, grouting pressure and HDFCAUT are selected as input variables. The horizontal and vertical deformations of the top of the existing tunnel lining and the horizontal and vertical deformation rates of the top of the tunnel lining are selected as output variables. 1237 groups of field monitoring data are selected as sample data sets, part of the data is shown in Table 2, 990 groups are randomly selected as training sets, and the remaining 247 groups are selected as test sets.

[0123] Table 2 Input and output parameter data information

[0124]

[0125]

[0126] (2) BO-CNN-three layer LSTM parameter optimization

[0127] In order to make the BO-CNN-three layer LSTM regression model obtain better prediction effect, the BO is adopted to optimize the four hyperparameters of learning_rate, minibatch_frac, col_sample and n_estimators of the model, and the model accuracy is verified by combining the 5-fold cross-validation method. Before optimizing the parameters, the initial ranges of the four parameters are given as shown in Table 3, and the hyperparameter search results of the horizontal and vertical deformations of the top of the existing tunnel lining and the horizontal and vertical deformation rates of the top of the tunnel lining are shown in Figure 4 、 Figure 5 、 Figure 6 、 Figure 7

[0128] Table 3 BO-CNN-three layer LSTM model hyperparameter meaning and range

[0129]

[0130] Figure 4 、 Figure 5 、 Figure 6 、 Figure 7 ​​It can be seen that the combination of the tunnel lining top transverse and vertical deformation, the tunnel lining top transverse and vertical deformation rate prediction model error reaches the minimum value as shown in Table 4 below.

[0131] Table 4 BO-CNN-three layer LSTM hyperparameter range setting and optimization results

[0132]

[0133]

[0134] (3) BO-CNN-three layer LSTM prediction model establishment and evaluation

[0135] Based on the BO-CNN-three layer LSTM parameter optimization results, the training set is used for learning simulation, and the BO-CNN-three layer LSTM tunnel tunnel lining top transverse and vertical deformation, tunnel lining top transverse and vertical deformation rate prediction model training is established respectively, as shown in Figures 8-11 , the test set is used to test the prediction model of the training set, and the prediction results are as shown in Figures 12-15 .

[0136] From Figure 8 and Figure 12 , it can be seen that the BO-CNN-three layer LSTM model can better realize the prediction of the vertical displacement of the tunnel vault. Figure 8 The BO-CNN-three layer LSTM prediction results of the tunnel bed settlement training set. The results show that the BO-CNN-three layer LSTM model fully learns the relationship between the input variables and output indicators of the training sample set, and the results obtained by predicting the bed settlement through the training set are basically consistent with the actual observation value, the mean square error is 0.011, and the goodness of fit is 0.970. Figure 12 For test set prediction results comparison, it can be seen that the predicted value of the vault settlement in the test set is very consistent with the actual observation value, the mean square error is 0.015, the goodness of fit is 0.966, and the results obtained by the training set and the test set show that the BO-CNN-three layer LSTM prediction model has high precision and good generalization performance, and the fitted nonlinear prediction function has high accuracy.

[0137] Similarly, from Figure 9 and Figure 13 , Figure 10 and Figure 14 , Figure 11 and Figure 15It can be seen that the BO-CNN-three layer LSTM has good accuracy in predicting the horizontal displacement, vertical displacement and deformation rate of the tunnel vault.

[0138] (4) Establish the objective function

[0139] The BO-CNN-three layer LSTM regression function is introduced as the objective function of NSGA-III.

[0140] The BO-CNN-three layer LSTM-based vault vertical displacement objective function ming1:

[0141] ming1(BO-CNN-three layer LSTM(χ1,χ2,χ3,χ4,χ5,χ6,χ7,χ8,χ9))

[0142] The BO-CNN-three layer LSTM-based vault horizontal displacement objective function ming2:

[0143] ming2(BO-CNN-three layer LSTM(χ1,χ2,χ3,χ4,χ5,χ6,χ7,χ8,χ9))

[0144] The BO-CNN-three layer LSTM-based vault vertical displacement deformation rate objective function ming3:

[0145] ming3(BO-CNN-three layer LSTM(χ1,χ2,χ3,χ4,χ5,χ6,χ7,χ8,χ9))

[0146] The BO-CNN-three layer LSTM-based vault horizontal displacement deformation rate objective function ming4:

[0147] ming4(BO-CNN-three layer LSTM(χ1,χ2,χ3,χ4,χ5,χ6,χ7,χ8,χ9))

[0148] Where χ1, χ2, χ3, χ4, χ5, χ6, χ7, χ8, χ9 represent the soil pressure, cutterhead speed, cutterhead torque, advancing speed, total thrust, screw machine speed, screw machine torque, grouting pressure, HDFCAUT (horizontal distance from working face to new and old tunnel intersection axis).

[0149] (5) Determine the constraint range

[0150] The constraint range of each influencing factor parameter is set according to the engineering actual situation. When the NSGA-III algorithm is used to find the optimal solution of the shield construction parameters, the decision range of the initial population must be set first to ensure that the initial population has practical significance. In order to avoid the optimized shield machine parameter results being too different from the actual engineering parameters of the shield machine, causing a large parameter adjustment span and potential safety hazards, the range of the initial decision variable is set according to the fluctuation range of each input parameter in the selected sample as the main reference, as shown in Table 5. In addition, the range of the input parameter HDFCAUT (x9) is [-2D, 2D].

[0151] Table 5 Reasonable range of shield parameters

[0152]

[0153]

[0154] (6) Multi-objective optimization and decision based on NSGA-III

[0155] Before performing the NSGA-III multi-objective optimization, the number of genetic algorithm objectives, population size, crossover and mutation operator values, and optimization stopping criteria are first determined. Considering that a proper population size and iteration number can promote the convergence of multi-objective optimization, the crossover operator of the NSGA-III algorithm is taken as 0.7, the mutation operator is taken as 0.01, the population size is taken as 200, and the maximum evolution and stopping generations are taken as 100. After the parameter setting is completed, the NSGA-III algorithm is run, and 692 pairs of Pareto front solutions are obtained. The Pareto front solution graph is shown in Figure 16 .

[0156] Further, based on the content of the above-mentioned device embodiment, as an optional embodiment, the embodiment of the present application provides a shield tunneling deformation optimization system based on intelligent algorithm, which is used for realizing and executing the method of any embodiment or combination of multiple embodiments, comprising: a first master control module, used for determining different approaching sections of shield tunneling near a tunnel, sensitive factors causing tunnel deformation, establishing an index system, collecting actual monitoring data in the shield tunneling process, and constructing a sample data set; a second master control module, used for establishing a prediction model of tunnel lining top horizontal and vertical deformation and tunnel lining top horizontal and vertical deformation rate based on BO-CNN-three layer LSTM, and training the prediction model by using the sample data set to obtain a nonlinear mapping relationship function of tunnel lining top horizontal and vertical deformation and tunnel lining top horizontal and vertical deformation rate; and a third master control module, used for taking the nonlinear mapping relationship function as a target optimization fitness function, constructing a multi-objective optimization model based on BO-CNN-three layer LSTM-NSGA-III, and proposing an optimal shield segment parameter setting suggestion value.

[0157] The method of the embodiment of the present application is realized by relying on an electronic device, and therefore it is necessary to introduce the related electronic device. For this purpose, the embodiment of the present application provides an electronic device, as shown in the accompanying drawings, which comprises at least one processor, a communications interface, at least one memory and a communications bus, wherein the at least one processor, the communications interface and the at least one memory complete mutual communication through the communications bus. The at least one processor can call logical instructions in the at least one memory to execute all or part of the steps of the method provided by the foregoing various method embodiments. Figure 3

[0158] ​In addition, the logic instructions in the at least one memory described above can be implemented in the form of a software function unit and sold or used as an independent product, and can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various method embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various program code storage media.

[0159] The device embodiments described above are only schematic, wherein the units described as separate components can or can not be physically separated, and the components displayed as units can or can not be physical units, i.e., can be located in one place, or can be distributed on multiple network units. Part or all of the modules can be selected to achieve the purpose of the present embodiment according to actual needs. Those skilled in the art can understand and implement without creative labor.

[0160] From the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be realized by means of software and the necessary general hardware platform, and of course can also be realized by hardware. Based on such understanding, the above technical solutions essentially or the part that contributes to the prior art can be embodied in the form of a software product. The computer software product can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0161] The computer program product of the present application can be a computer program product comprising a computer-readable medium bearing computer program code embodied therein for use with a computer. In operation, the computer program code causes a computer to perform the steps of the present application. The computer program code can be written in any form suitable for use with the computer for example, in the form of source code, object code, interpretative code, etc. The computer program code can be supplied to a user on a computer-readable medium such as a floppy disk, CD ROM, or over a computer network such as the Internet. The computer-readable medium can be magnetic, optical, or other computer storage medium. The computer-readable medium can be a distributed computer-readable medium that is remotely accessible via the Internet or other computer network.

[0162] Those skilled in the art will readily understand that the above description is only one of the many preferred embodiments of the present application and that many modifications, e.g., adaptations, changes, substitutions, improvements, etc., can be made by those skilled in the art in the art without departing from the spirit and scope of the present application. Accordingly, the present application is not intended to be limited to the above described preferred embodiments but rather is to be accorded with the widest scope consistent with the principles and features disclosed herein.

Claims

1. A shield tunneling under existing tunnel deformation optimization method based on intelligent algorithm, characterized in that, The method comprises the following steps: Step one, determine the sensitive factors causing tunnel deformation in different approaching sections in the shield approaching tunnel construction stage, establish an index system, collect actual monitoring data in the shield tunneling process, and construct a sample data set; Step two, establish a prediction model of tunnel lining top horizontal and vertical deformation and tunnel lining top horizontal and vertical deformation rate based on BO-CNN-three layer LSTM, and train the prediction model using the sample data set to obtain a nonlinear mapping relationship function of tunnel lining top horizontal and vertical deformation and tunnel lining top horizontal and vertical deformation rate; Step three, use the nonlinear mapping relationship function as the objective optimization fitness function, construct a multi-objective optimization model based on BO-CNN-three layer LSTM-NSGA-III, and propose an optimal shield segment parameter setting suggestion value; Step three comprises the following steps: (31) Determine the fitness function, and use the nonlinear mapping relationship function as the objective optimization fitness function: ming1 = min(BO - CNN - three layer LSTM regression (X i )) ming2 = min(BO - CNN - three layer LSTM regression (X i )) ming3 = min(BO - CNN - three layer LSTM regression (X i )) ming4 = min(BO - CNN - three layer LSTM regression (X i )) wherein X i The soil pressure, the cutterhead rotating speed, the cutterhead torque, the advancing speed, the total thrust, the screw machine speed, the screw machine torque, the grouting pressure, the horizontal distance from the working face to the intersection axis of the new and old tunnels, ming1 is the vault horizontal displacement target function, ming2 is the vault vertical displacement target function, ming3 is the vault vertical displacement deformation rate target function, and ming4 is the vault horizontal displacement deformation rate target function. (32) Determine the constraint range, and set constraint conditions for the value of the target decision parameter according to the engineering requirements and existing project data: wherein and respectively represent the lower and upper limits of the input parameter value. (33) Use the NSGA-III algorithm to realize multi-objective optimization control of existing tunnel deformation caused by shield approaching underpass construction and construction parameter optimization problem.

2. The intelligent algorithm-based deformation optimization method for shield tunneling under existing tunnels according to claim 1, characterized in that, In step two, the grid search method is used to optimize the key parameters of BO-CNN-three layer LSTM, and then the 5-fold cross-validation method is used to select the optimal parameter combination, and the BO-CNN-three layer LSTM prediction model is established based on the optimal parameter combination; Wherein, the BO is used to optimize the four hyperparameters of learning_rate, minibatch_frac, col_sample and n_estimators. Wherein, learning_rate is the learning rate, minibatch_frac is the hyperparameter optimization gradient descent process, col_sample is the sampling rate, and n_estimators is the number of decision trees.

3. The intelligent algorithm-based shield tunneling deformation optimization method according to claim 1, characterized in that, Step two further comprises: The reliability of the prediction model is evaluated by using three performance indicators of root mean square error, mean absolute error and mean absolute percentage error.

4. The intelligent algorithm-based shield tunneling deformation optimization method according to claim 3, characterized in that, The NSGA-III algorithm is used to realize the multi-objective optimization control of existing tunnel deformation caused by shield approaching underpass construction and construction parameter optimization problem, and the vertical deformation is output as a Pareto solution set.

5. A shield tunneling under existing tunnel deformation optimization system based on intelligent algorithm, characterized in that, It comprises: The first main control module is used to determine the sensitive factors causing tunnel deformation in different approaching sections in the shield approaching tunnel construction stage, establish an index system, collect actual monitoring data in the shield tunneling process, and construct a sample data set; The second master module is configured to establish a prediction model of the lateral and vertical deformation of the tunnel lining top and the lateral and vertical deformation rate of the tunnel lining top based on BO-CNN-three layer LSTM, and train the prediction model by using the sample data set to obtain a nonlinear mapping relationship function of the lateral and vertical deformation of the tunnel lining top and the lateral and vertical deformation rate of the tunnel lining top. The third master module is configured to use the nonlinear mapping relationship function as a target optimization fitness function, and construct a multi-objective optimization model based on BO-CNN-three layer LSTM-NSGA-III to propose an optimal shield segment parameter setting suggestion value. The third master module is configured to perform the following steps: (31) determining a fitness function, and using the nonlinear mapping relationship function as a target optimization fitness function; ming1 = min(BO - CNN - three layer LSTM regression (X i )) ming2 = min(BO - CNN - three layer LSTM regression (X i )) ming3 = min(BO - CNN - three layer LSTM regression (X i )) ming4 = min(BO - CNN - three layer LSTM regression (X i )) wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X i wherein X (32) determining a constraint range, and setting a constraint condition for the value of the target decision parameter according to the engineering requirement and the existing data of the project; wherein and respectively represent the lower and upper limits of the input parameter value. (33) using the NSGA-III algorithm to realize multi-objective optimization control of the deformation of the existing tunnel caused by the shield close approach underpass construction and optimization of the construction parameter.

6. An electronic device, comprising: The system comprises: at least one processor, at least one memory and a communication interface; wherein the processor, the memory and the communication interface communicate with each other; the memory stores program instructions executable by the processor, and the processor invokes the program instructions to execute the method according to any one of claims 1 to 4.

7. A non-transitory computer-readable storage medium, comprising: The non-transitory computer readable storage medium stores computer instructions, and the computer instructions enable the computer to execute the method according to any one of claims 1 to 4. The non-transitory computer readable storage medium stores computer instructions, and the computer instructions enable the computer to execute the method according to any one of claims 1 to 4.

Images