Prediction method for adjacent tunnel deformation induced by foundation pit excavation and electronic equipment

Abstract

The invention relates to a method and electronic equipment for predicting adjacent tunnel deformation induced by foundation pit excavation, and the method comprises the following steps: obtaining input layer parameters of foundation pit excavation data, including foundation pit structure geometric parameters, foundation pit structure mechanical parameters and soil parameters; inputting the obtained foundation pit excavation data into the trained ANN prediction model to obtain the maximum horizontal displacement of the enclosure wall; the maximum horizontal displacement of the enclosure wall is calculated through a displacement control two-stage analysis method, and tunnel deformation induced by foundation pit excavation is obtained; the soil parameters of the input layer of the ANN prediction model are optimized by using an inverse analysis method. Compared with the prior art, the method has the advantages that intelligent prediction of adjacent tunnel deformation induced by foundation pit excavation is achieved, and the prediction precision is high.

Description

Technical Field

[0001] The invention relates to the technical field of tunnel engineering, and in particular to a method for predicting deformation of adjacent tunnels induced by foundation pit excavation. Background Art

[0002] With the rapid development of urban construction, a large number of deep foundation pit projects are excavated near existing subway tunnels or municipal pipelines. The stress release caused by excavation will inevitably lead to soil displacement, thereby inducing tunnel deformation and structural damage, such as lining cracks and dislocation. These problems will affect the normal use of subways and municipal pipelines, and even cause serious harm. Therefore, it is crucial for engineering technicians to accurately predict tunnel deformation caused by deep foundation pit excavation.

[0003] The randomness of soil distribution and the significant uncertainty of its characteristics, coupled with the complex interactions between soil, support structure, tunnel structure and construction personnel, make it quite difficult to predict tunnel deformation caused by foundation pit excavation. At present, tunnel deformation induced by foundation pit excavation is often predicted by numerical simulation. However, to construct a numerical model with reasonable results, strong comprehensive knowledge in numerical, mechanical and other aspects is required, especially in the determination of soil parameters, which requires a lot of experience, which often leads to large errors in calculation.

[0004] Machine learning methods can effectively learn the nonlinear relationship between various factors, thereby considering the impact of the construction process. Many scholars have used neural network methods to predict foundation pit deformation. However, this type of method still has problems such as limited universality and insufficient data sets, which restricts its application. In addition, in the current research on the application of neural networks in foundation pit engineering, the input parameters are usually limited and cannot fully reflect the deformation mechanism of the foundation pit. In summary, it is necessary to develop a method to predict the impact of foundation pit excavation on tunnel deformation under actual working conditions.

[0005] How to achieve intelligent prediction of deformation of adjacent tunnels induced by foundation pit excavation has become a technical problem that needs to be solved. Summary of the invention

[0006] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and provide a method for predicting deformation of adjacent tunnels induced by foundation pit excavation.

[0007] The purpose of the present invention can be achieved by the following technical solutions:

[0008] According to one aspect of the present invention, a method for predicting deformation of adjacent tunnels induced by foundation pit excavation is provided, comprising the following steps:

[0009] Step S101, obtaining input layer parameters of foundation pit excavation data, including foundation pit structure geometric parameters, foundation pit structure mechanical parameters and soil parameters;

[0010] Step S102, inputting the acquired foundation pit excavation data into the trained ANN prediction model to obtain the maximum horizontal displacement of the retaining wall;

[0011] Step S103, using a displacement control two-stage analysis method to calculate the maximum horizontal displacement of the retaining wall in step S102, and obtain the tunnel deformation induced by the foundation pit excavation;

[0012] The soil parameters of the input layer of the ANN prediction model are optimized using the back analysis method.

[0013] Preferably, the optimization process of the ANN prediction model includes:

[0014] Step S1, inputting the input layer parameters into the trained ANN prediction model, and outputting the maximum horizontal displacement of the retaining wall;

[0015] Step S2, using a displacement control two-stage analysis method to calculate the maximum horizontal displacement of the retaining wall to obtain an unoptimized tunnel deformation;

[0016] Step S3, optimizing the soil parameters of the input layer of the ANN prediction model by using the back analysis method;

[0017] Step S4, inputting the input layer parameters into the ANN prediction model of the optimized soil parameters, and directly outputting the optimized tunnel deformation after executing steps S2 to S3 again;

[0018] Step S5, compare the unoptimized and optimized tunnel deformations with the measured values ​​respectively. If the optimized tunnel deformation is closer to the measured value, it means that the soil parameters of the ANN prediction model optimized by the inverse analysis method are reasonable, that is, the ANN prediction model has been optimized; otherwise, return to step S3 to continue optimizing the soil parameters.

[0019] Preferably, the training process of the ANN prediction model includes:

[0020] Generate synthetic specimens through 2D finite element simulation, including selecting finite element analysis library, defining the problem, building the model, setting simulation parameters, running the simulation, post-processing and generating synthetic specimens;

[0021] The synthetic samples are divided into training set and validation set, and the parameters of the ANN prediction model are adjusted to minimize the error between the predicted value and the actual value;

[0022] The ANN prediction model is a model based on the ANN neural network;

[0023] The simulation parameter setting includes input layer parameters and output layer parameters, wherein the output layer parameters are the maximum horizontal deformation of the foundation pit retaining wall.

[0024] Preferably, the activation function of the ANN prediction model is ReLu, specifically:

[0025] g(x)=max(0,x)

[0026] Among them, x is the input value, including the geometric parameters of the foundation pit structure, the mechanical parameters of the foundation pit structure and the soil parameters; g(x) is the output value, that is, the maximum horizontal deformation of the foundation pit retaining wall.

[0027] Preferably, the displacement controlled two-stage analysis method comprises:

[0028] The free-field deformation of soil is calculated using the calculation formula for the three-dimensional deformation of adjacent strata induced by foundation pit excavation;

[0029] The Euler-Bernoulli beam theory is used to calculate tunnel deformation.

[0030] More preferably, the calculation formula for the three-dimensional deformation of the adjacent strata induced by the foundation pit excavation is specifically:

[0031]

[0032] Among them, u max is the maximum lateral deformation of the retaining wall; a x 、b x and c x are all coefficients; x, y, z are the three-dimensional space coordinate system values ​​established with a certain point on the soil surface as the origin, H is the excavation depth; D is the embedding depth of the retaining wall; R is the tunnel radius, and u(x, y, z) is the free field deformation of the soil.

[0033] More preferably, the calculation formula of the Euler-Bernoulli beam theory is specifically:

[0034]

[0035] Among them, u is the deformation of each section of the tunnel lining, E is the elastic modulus of the beam, x is the length of the beam in the axial direction, w is the distributed load on the beam, I is the inertia matrix, and w is the distributed load on the beam.

[0036] Preferably, the soil parameters of the input layer of the ANN prediction model are optimized using the inverse analysis method, specifically: using Matlab and Macro coupling programming, and calling the corresponding finite element program for iterative optimization to optimize the soil parameters of the input layer of the ANN prediction model, wherein each layer of optimized soil parameters only includes stiffness modulus and shear strain.

[0037] More preferably, the finite element program uses a weighted least squares error equation to evaluate the analysis results, and the governing equation is as follows:

[0038] S(b)=[yy′ (b)] T ω[yy ′ (b)]

[0039] Where b is the parameter vector to be optimized, y is the measured result vector, and y ′ (b) is the calculation result vector, ω is the weight matrix;

[0040] The finite element program uses two convergence criteria to determine whether the optimization results converge and when to stop the iteration process, including:

[0041] When the measured result vector y and the calculated result vector y ′ (b) If the maximum value of the parameter change is less than the value defined by the user, the iteration converges and the iteration process stops;

[0042] Or when the change in the objective equation S(b) is less than the value defined by the user, the iteration converges and the iteration process stops.

[0043] According to another aspect of the present invention, an electronic device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the method described above is implemented when the processor executes the program.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] 1) The present invention uses the back analysis method to optimize the soil parameters of the ANN neural network input layer parameters, and uses the advantages of the ANN neural network in identifying complex interaction patterns of multiple factors to predict the maximum horizontal displacement of the retaining wall of the foundation pit excavation support structure; the neural network method is combined with the theoretical formula of soil mechanics, and the displacement control two-stage analysis method is used to analyze the tunnel deformation, thereby realizing an intelligent prediction method for the deformation of adjacent tunnels induced by foundation pit excavation.

[0046] 2) The maximum horizontal displacement of the retaining wall and the tunnel deformation prediction obtained by comparing the back analysis method before and after optimization show that the ANN prediction model without optimizing the soil parameters overestimates the wall deformation to a certain extent. After optimization, the maximum horizontal displacement of other retaining walls and the calculated tunnel deformation are closer to the measured values, and the prediction accuracy is significantly improved compared with that before optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic flow chart of a method for predicting deformation of adjacent tunnels induced by foundation pit excavation in one embodiment of the present invention;

[0048] Figure 2 It is a schematic flow chart of a method for predicting deformation of adjacent tunnels induced by foundation pit excavation in another embodiment of the present invention;

[0049] Figure 3It is a spatial schematic diagram of the three-dimensional deformation of the stratum in the present invention;

[0050] Figure 4 This is a schematic diagram of the working condition in an embodiment of the present invention;

[0051] Figure 5 A schematic diagram showing a comparison between the finite element calculation results and the output results of the ANN prediction model in an embodiment of the present invention;

[0052] Figure 6 It is a schematic diagram comparing the measured deformation of the retaining wall under two working conditions in an embodiment of the present invention and the maximum deformation output by the ANN prediction model;

[0053] Figure 7 It is a schematic diagram comparing the horizontal deformation of the retaining wall before and after the reverse analysis optimization and the actual measurement in the embodiment of the present invention;

[0054] Figure 8 It is a schematic diagram comparing tunnel deformation before and after back-analysis optimization and actual measurement in an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0056] This embodiment provides a method for predicting deformation of adjacent tunnels induced by foundation pit excavation. Figure 1 As shown, the following steps are included:

[0057] Step S1, constructing an ANN (Artificial Neural Network, ANN) prediction model, specifically including:

[0058] Step S1-1, establish and train the ANN prediction model:

[0059] Generate synthetic samples through two-dimensional finite element simulation, including: select finite element analysis library, define problem, build model, set simulation parameters, run simulation, post-process and generate synthetic samples.

[0060] The simulation parameters include input layer parameters and output layer parameters. The input layer parameters (i.e., the input values ​​below) include the geometric parameters of the foundation pit structure, the mechanical parameters of the foundation pit structure, and the soil parameters. The output layer parameters (i.e., the output values ​​below) are the maximum horizontal deformation of the foundation pit retaining wall. The activation function of the ANN prediction model is the linear correction unit (ReLu), which is:

[0061] g(x)=max(0,x)

[0062] Among them, x is the input value, including the geometric parameters of the foundation pit structure, the mechanical parameters of the foundation pit structure and the soil parameters; g(x) is the output value, that is, the maximum horizontal deformation of the foundation pit retaining wall.

[0063] The ANN prediction model is trained using PyTorch, including: dividing the data set (i.e., the synthetic samples generated above) into a training set and a validation set, adjusting the parameters of the ANN prediction model to minimize the error between the predicted value and the actual value; and the Adam optimization algorithm is used as the training optimization method.

[0064] Step S1-2, verify the ANN prediction model:

[0065] 70% of the samples are randomly selected as the training set, and the remaining samples are the validation set. The convergence condition is that the error of the training set no longer decreases, while the error of the validation set increases.

[0066] The deformation of the retaining wall predicted by the ANN prediction model was compared with the results of finite element simulation, and the performance of the trained neural network prediction model was evaluated based on the mean square error (MSE) using a validation set.

[0067] The mean square error (MSE) formula is:

[0068]

[0069] Among them, y m is the true value, is the predicted value, and M is the sample size for calculation.

[0070] Step S1-3, forming an ANN prediction model for predicting the maximum horizontal deformation of the foundation pit retaining wall.

[0071] After training and verification, an ANN prediction model for predicting the maximum horizontal deformation of foundation pit retaining wall was obtained.

[0072] Step S2: inputting model values ​​(input layer parameters) into the ANN prediction model to obtain deformation results, specifically including:

[0073] Step S2-1, input the input layer parameters into the ANN prediction model and output the maximum horizontal displacement of the retaining wall.

[0074] In step S2-2, the ANN prediction model uses a displacement-controlled two-stage analysis method to calculate the maximum horizontal displacement of the retaining wall and obtains the tunnel deformation induced by foundation pit excavation (unoptimized tunnel deformation).

[0075] Two-stage analytical approach, including:

[0076] The free-field deformation of soil is calculated using the calculation formula for the three-dimensional deformation of adjacent strata induced by foundation pit excavation;

[0077] The Euler-Bernoulli beam theory is used to calculate tunnel deformation, and its basic assumptions are clarified: cross-section deformation assumption and rigid cross-section assumption. The finite element method is used to model and calculate tunnel deformation.

[0078] The calculation formula for the three-dimensional deformation of the adjacent strata induced by foundation pit excavation is as follows:

[0079]

[0080] Among them, u max is the maximum lateral deformation of the retaining wall; a x 、b x 、c x are all set coefficients; x, y, and z are the values ​​of the three-dimensional space coordinate system established with a certain point on the soil surface as the origin, such as Figure 3 As shown: H is the excavation depth; D is the embedding depth of the retaining wall; R is the tunnel radius. Figure 3 The base in is the same as u max Coincident horizontal planes.

[0081] The Euler-Bernoulli formula is shown below:

[0082]

[0083] Where: u is the deformation of each section of the tunnel lining, E is the elastic modulus of the beam, x is the length of the beam in the axial direction, w is the distributed load on the beam, I is the inertia matrix, and w is the distributed load on the beam.

[0084] Figure 4 The maximum horizontal displacement of the retaining wall predicted by the ANN neural network and the 46 results of the finite element simulation are compared. Figure 5 As shown in the figure, the dotted straight line indicates that the finite element calculation results coincide with the output results of the ANN prediction model, that is, there is no error; the discrete points close to the dotted line are the relationship between the synthetic samples generated by the two-dimensional finite element simulation (the same input results correspond to the actual output numerical simulation results) and the output results of the ANN prediction model; the closer to the dotted line, the higher the degree of consistency between the two results, and the smaller the error. The maximum horizontal displacement of the retaining wall predicted by the ANN prediction model is consistent with the numerical simulation results, with an average relative error of 12.2%.

[0085] Step S3, optimizing soil parameters using the back analysis method, specifically including:

[0086] The back analysis method is run. This process uses Matlab and Macro coupling programming to call the corresponding finite element program to optimize the soil parameters of the input layer of the ANN prediction model. The optimized soil parameters of each layer only include stiffness modulus and shear strain.

[0087] The finite element program uses a weighted least squares equation to evaluate the analysis results. The governing equation is as follows:

[0088] S(b)=[yy ′ (b)] T ω[yy ′ (b)]

[0089] Where b is the parameter vector to be optimized, y is the measured result vector, and y ′ (b) is the calculation result vector, and ω is the weight matrix.

[0090] The finite element program uses the following two convergence criteria to determine whether the optimization results have converged and when to stop the iteration process: ① During the iteration process, the maximum parameter change value (y and y ′ (b) If the change in the target equation S(b) is less than the user-defined value, the iteration converges and the iteration process stops. ② If the change in the target equation S(b) is less than the user-defined value, the iteration converges and the iteration process stops.

[0091] Step S4, re-predicting using the ANN neural network of the optimized soil parameters and the two-stage analysis method of displacement control to obtain the optimized tunnel deformation result.

[0092] Step S5, compare the unoptimized and optimized tunnel deformations with the measured values ​​respectively. If the optimized tunnel deformation is closer to the measured value, it means that the soil parameters of the ANN prediction model optimized by the back analysis method are reasonable, that is, the ANN prediction model has been trained and can be used for tunnel deformation prediction based on field data; otherwise, return to step S3 and continue to optimize the soil parameters.

[0093] The ANN prediction model was inverted using the measured data obtained at a depth of 13 m during the excavation process.

[0094] The back analysis results of stiffness modulus and shear strain are shown in Table 1, where:

[0095] is the reference secant modulus of the triaxial drained shear test;

[0096] is the reference tangent modulus in the consolidation test;

[0097] is the reference loading and unloading modulus of the triaxial drained shear test;

[0098] is the reference initial modulus for small strain stiffness test;

[0099] ε is the strain.

[0100] Table 1

[0101]

[0102] Excavation of foundation pit to 13m( Figure 6 1) and 16m( Figure 6 The measured deformation of the retaining wall (retaining wall) in the two working conditions (depth 2) and the maximum deformation of the ANN prediction model are as follows Figure 6 As shown in the figure, before optimization means that the anti-analysis optimization is not performed, and after optimization means that the anti-analysis optimization is performed.

[0103] Before the back-analysis optimization, the ANN prediction model overestimated the wall deformation to a certain extent. When the excavation depth was shallow, the maximum error was about 8mm. As the construction progressed, the maximum error between the measured value and the predicted value was reduced to 5mm. Finally, the prediction accuracy of the retaining wall deformation was significantly improved. Figure 7 .

[0104] Figure 8 is the measured value of tunnel deformation induced by foundation pit excavation ( Figure 8 The measured value) and predicted value ( Figure 8 From the results of the two-stage model (after optimization by the inverse analysis method) and the analysis results of directly inputting the measured retaining wall deformation into the two-stage model (before optimization by the inverse analysis method), it can be seen that the tunnel deformation after optimization by the inverse analysis method is closer to the measured value than that without optimization.

[0105] Example 2

[0106] This embodiment provides a prediction method for the deformation of adjacent tunnels induced by foundation pit excavation. Figure 2 , the method comprising:

[0107] Step S101, obtaining on-site foundation pit excavation data, including foundation pit structure geometric parameters, foundation pit structure mechanical parameters and soil parameters;

[0108] Step S102, inputting the acquired foundation pit excavation data into the trained ANN prediction model to obtain the predicted value of the maximum horizontal displacement of the retaining wall;

[0109] Step S103, using a displacement-controlled two-stage analysis method to calculate the maximum horizontal displacement of the retaining wall in step S102, and obtain the tunnel deformation induced by the foundation pit excavation.

[0110] Example 3

[0111] The electronic device of the present invention includes a central processing unit (CPU), which can perform various appropriate actions and processes according to computer program instructions stored in a read-only memory (ROM) or loaded from a storage unit into a random access memory (RAM). In the RAM, various programs and data required for device operation can also be stored. The CPU, ROM and RAM are connected to each other via a bus. An input / output (I / O) interface is also connected to the bus.

[0112] Multiple components in the device are connected to the I / O interface, including: input units, such as keyboards, mice, etc.; output units, such as various types of displays, speakers, etc.; storage units, such as disks, optical disks, etc.; and communication units, such as network cards, modems, wireless communication transceivers, etc. The communication unit allows the device to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunication networks.

[0113] The processing unit performs the various methods and processes described above, such as methods S101 to S105. For example, in some embodiments, methods S101 to S105 may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit. In some embodiments, part or all of the computer program may be loaded and / or installed on the device via a ROM and / or a communication unit. When the computer program is loaded into the RAM and executed by the CPU, one or more steps of methods S101 to S105 described above may be performed. Alternatively, in other embodiments, the CPU may be configured to execute methods S101 to S105 in any other appropriate manner (e.g., by means of firmware).

[0114] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0115] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0116] In the context of the present invention, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0117] The preferred specific embodiments of the present invention are described in detail above. It should be understood that a person skilled in the art can make many modifications and changes based on the concept of the present invention without creative work. Therefore, any technical solution that can be obtained by a person skilled in the art through logical analysis, reasoning or limited experiments based on the concept of the present invention on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A method for predicting deformation of adjacent tunnels induced by foundation pit excavation, characterized in that: The following steps are involved: Step S101, obtaining input layer parameters of foundation pit excavation data, including foundation pit structure geometric parameters, foundation pit structure mechanical parameters and soil parameters; Step S102, inputting the acquired foundation pit excavation data into the trained ANN prediction model to obtain the maximum horizontal displacement of the retaining wall; Step S103, using a displacement control two-stage analysis method to calculate the maximum horizontal displacement of the retaining wall in step S102, and obtain the tunnel deformation induced by the foundation pit excavation; The soil parameters of the input layer of the ANN prediction model are optimized using the back analysis method.

2. The method for predicting deformation of adjacent tunnels induced by foundation pit excavation according to claim 1, characterized in that: The optimization process of the ANN prediction model includes: Step S1, inputting the input layer parameters into the trained ANN prediction model, and outputting the maximum horizontal displacement of the retaining wall; Step S2, using a displacement control two-stage analysis method to calculate the maximum horizontal displacement of the retaining wall to obtain an unoptimized tunnel deformation; Step S3, optimizing the soil parameters of the input layer of the ANN prediction model by using the back analysis method; Step S4, inputting the input layer parameters into the ANN prediction model of the optimized soil parameters, and directly outputting the optimized tunnel deformation after executing steps S2 to S3 again; Step S5, compare the unoptimized and optimized tunnel deformations with the measured values ​​respectively. If the optimized tunnel deformation is closer to the measured value, it means that the soil parameters of the ANN prediction model optimized by the inverse analysis method are reasonable, that is, the ANN prediction model has been optimized; otherwise, return to step S3 to continue optimizing the soil parameters.

3. The method for predicting deformation of adjacent tunnels induced by foundation pit excavation according to claim 1, characterized in that: The training process of the ANN prediction model includes: Generate synthetic specimens through 2D finite element simulation, including selecting finite element analysis library, defining the problem, building the model, setting simulation parameters, running the simulation, post-processing and generating synthetic specimens; The synthetic samples are divided into training set and validation set, and the parameters of the ANN prediction model are adjusted to minimize the error between the predicted value and the actual value; The ANN prediction model is a model based on the ANN neural network; The simulation parameter setting includes input layer parameters and output layer parameters, wherein the output layer parameters are the maximum horizontal deformation of the foundation pit retaining wall.

4. The method for predicting deformation of adjacent tunnels induced by foundation pit excavation according to claim 1, characterized in that: The activation function of the ANN prediction model is ReLu, specifically: g(x)=max(0,x) Among them, x is the input value, including the geometric parameters of the foundation pit structure, the mechanical parameters of the foundation pit structure and the soil parameters; g(x) is the output value, that is, the maximum horizontal deformation of the foundation pit retaining wall.

5. The method for predicting deformation of adjacent tunnels induced by foundation pit excavation according to claim 1, characterized in that: The displacement control two-stage analysis method includes: The free-field deformation of soil is calculated using the calculation formula for the three-dimensional deformation of adjacent strata induced by foundation pit excavation; The Euler-Bernoulli beam theory is used to calculate tunnel deformation.

6. A method for predicting deformation of adjacent tunnels induced by foundation pit excavation according to claim 5, characterized in that: The calculation formula for the three-dimensional deformation of the adjacent strata induced by the foundation pit excavation is specifically: Among them, u max is the maximum lateral deformation of the retaining wall; a x 、b x and c x are all coefficients; x, y, z are the three-dimensional space coordinate system values ​​established with a certain point on the soil surface as the origin, H is the excavation depth; D is the embedding depth of the retaining wall; R is the tunnel radius, and u(x, y, z) is the free field deformation of the soil.

7. The method for predicting deformation of adjacent tunnels induced by foundation pit excavation according to claim 5, characterized in that: The calculation formula of the Euler-Bernoulli beam theory is specifically: Among them, u is the deformation of each section of the tunnel lining, E is the elastic modulus of the beam, x is the length of the beam in the axial direction, w is the distributed load on the beam, I is the inertia matrix, and w is the distributed load on the beam.

8. The method for predicting deformation of adjacent tunnels induced by foundation pit excavation according to claim 1, characterized in that: The soil parameters of the input layer of the ANN prediction model are optimized using the inverse analysis method, specifically: Matlab and Macro coupling programming is used, and the corresponding finite element program is called for iterative optimization to optimize the soil parameters of the input layer of the ANN prediction model, wherein each layer of optimized soil parameters only includes stiffness modulus and shear strain.

9. A method for predicting deformation of adjacent tunnels induced by foundation pit excavation according to claim 8, characterized in that: The finite element program uses a weighted least squares equation to evaluate the analysis results. The governing equation is as follows: S(b)=[y-y ′ (b)] T ω[y-y ′ (b)] Where b is the parameter vector to be optimized, y is the measured result vector, and y ′ (b) is the calculation result vector, ω is the weight matrix; The finite element program uses two convergence criteria to determine whether the optimization results converge and when to stop the iteration process, including: When the measured result vector y and the calculated result vector y ′ (b) If the maximum value of the parameter change is less than the value defined by the user, the iteration converges and the iteration process stops; Or when the change in the objective equation S(b) is less than the value defined by the user, the iteration converges and the iteration process stops.

10. An electronic device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the program, the method according to any one of claims 1 to 9 is implemented.

Images