Intelligent design method for optimal discrete angle of tunnel small clear distance section based on deep learning

Abstract

The embodiment of the invention discloses a tunnel small clear distance section optimal discrete angle intelligent design method based on deep learning. The method comprises the steps of obtaining an initial value of a shield tunnel small clear distance section discrete angle and a current value of a small clear distance section length; performing combined assignment on a plurality of construction parameters in the reinforcement scheme by using an AI model based on deep learning; tunnel-stratum three-dimensional finite element simulation analysis is conducted, and whether the safety requirement is met or not is determined according to the mechanical property and the deformation property; if yes, re-checking the project budget; if the safety requirement is not met or the engineering budget is not met, the current value of the discrete angle is increased according to the linear step length, and the value of the discrete angle is increased; and determining a first discrete angle with the shortest construction period and the cost less than the highest budget and a second discrete angle with the lowest cost and the construction period less than the planned construction period from the feasible scheme set to form an optimal interval of the discrete angles. According to the embodiment, the optimal discrete angle can be determined.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of tunnel construction technology, and in particular to a deep learning-based intelligent design method for optimal discrete angles in tunnel sections with small clearances. Background Art

[0002] A small clearance tunnel refers to a special tunnel structure specified in railway or highway construction to ensure a safe distance between trains or vehicles and the tunnel wall, while taking into account the mutual influence between two parallel tunnels. Its characteristic is that the clear distance between the two parallel tunnels is small, usually less than 1.5 times the width of the tunnel excavation section. This design form is particularly common when the geological conditions are complex, the terrain is narrow, or when the line layout space needs to be compressed. The design of a small clearance tunnel requires special measures to reduce the mutual influence between the two caverns to ensure the safety and stability of the tunnel structure. The "Highway Design Code" JTGD70 stipulates: A small clearance tunnel refers to a special tunnel layout form in which the thickness of the middle rock wall in the tunnel is less than the minimum clearance of the separate independent double tunnels.

[0003] The design of small-clearance tunnels must first determine a reasonable clearance, and secondly, different countermeasures should be adopted for different types of small clearances. Existing shield tunnel design of small-clearance sections is primarily based on manual experience, generally focusing on meeting on-site functional and safety requirements. Due to the numerous control factors involved, a linear dynamic equilibrium relationship with key construction control indicators such as budget and construction period has not been established, and there is no basis for determining the optimal breakaway angle. Summary of the Invention

[0004] An embodiment of the present invention provides an intelligent design method for the optimal discrete angle of a tunnel with a small clearance distance based on deep learning to solve the above technical problems.

[0005] In a first aspect, an embodiment of the present invention provides a method for intelligently designing the optimal breakaway angle for a small clearance section of a tunnel based on deep learning, comprising:

[0006] Obtaining an initial value of a discrete angle of a small clearance section of a shield tunnel, and using the initial value as a current value of the discrete angle;

[0007] Determining a current value of the length of the small clearance section according to a current value of the discrete angle and a set shield tunnel spacing;

[0008] Using stratum grouting and reinforcement pile installation as a small-spacing reinforcement method for shield tunnels, a deep learning-based AI model is used to combine and assign values ​​to multiple construction parameters in the reinforcement scheme corresponding to the current values ​​of the discrete angle and the small-clearance segment length, where the multiple construction parameters include pile diameter, pile spacing, pile length, stratum grouting amount, and construction step distance;

[0009] Based on the combined values, a three-dimensional finite element simulation analysis of the tunnel and the stratum is performed to determine whether the reinforcement scheme meets the safety requirements based on the mechanical properties and deformation properties;

[0010] If the security requirements are met, the project budget is reviewed; if the project budget is met, the reinforcement plan is stored in the feasible plan set;

[0011] If the safety requirements or the project budget are not met, the current value of the discrete angle is increased according to a linear step size, and the process of determining the current value of the small clearance section length according to the current value of the discrete angle and the set shield tunnel spacing is returned to, until the final value of the discrete angle reaches the set maximum threshold;

[0012] From the final set of feasible solutions, determine the first discrete angle with the shortest duration and cost less than the maximum budget, and the second discrete angle with the lowest cost and duration less than the planned duration;

[0013] An interval between the first discrete angle and the second discrete angle is determined as an optimal interval of discrete angles.

[0014] In a second aspect, an embodiment of the present invention provides an electronic device, comprising:

[0015] one or more processors;

[0016] a memory for storing one or more programs,

[0017] When the one or more programs are executed by the one or more processors, the one or more processors implement the intelligent design method for optimal discrete angle of small clearance section of tunnel based on deep learning as described in any embodiment.

[0018] In a third aspect, an embodiment of the present invention further provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the intelligent design method for optimal discrete angle of small clearance section of tunnel based on deep learning as described in any embodiment.

[0019] In summary, the present invention provides a deep learning-based intelligent design method for the optimal discrete angle of small-clearance sections in tunnels. Building on the empirical design of small-clearance sections in traditional shield tunnels, this method introduces AI intelligent value assignment and three-dimensional numerical simulation analysis technology. Under the conditions of ensuring tunnel functionality and safety, the AI ​​intelligent value assignment comprehensively covers cyclic calculations to determine the range of discrete angle values, helping designers find the discrete angle θe that best suits engineering requirements and satisfies the optimal balance between "small-clearance section length, construction period, and budget." Specifically, this embodiment can achieve the following beneficial effects:

[0020] 1. Breaking through the limitation of traditional design experience value calculation that can only obtain a limited number of point values, through AI loop calculation, a scientific, accurate, complete and comprehensive discrete angle value range of small clearance section can be obtained for designers' reference.

[0021] 2. Comprehensively consider the key control factors such as safety, construction period, and budget in the small clearance section of shield tunnel projects, establish a dynamic relationship between the design of the discrete angle and safety, construction period, and budget, and construct a discrete angle θe design method with the optimal balance relationship of "small clearance section length-construction period-budget" that meets safety requirements.

[0022] 3. By introducing AI autonomous learning and combination empowerment technology and combining it with three-dimensional numerical simulation analysis technology, the intelligent design method is more scientific, efficient, convenient, accurate, complete and covers the entire range than the traditional design method. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0024] Figure 1 This is a flow chart of a method for intelligently designing optimal discrete angles for small clearance sections in tunnels based on deep learning, provided by an embodiment of the present invention;

[0025] Figure 2 This is a schematic diagram of a planar design of a small clearance section of a shield tunnel provided by an embodiment of the present invention;

[0026] Figure 3 This is a schematic diagram of the structure of a CNN model provided by an embodiment of the present invention;

[0027] Figure 4 This is a flowchart of another method for intelligently designing optimal discrete angles for small clearance sections in tunnels based on deep learning, provided by an embodiment of the present invention;

[0028] Figure 5 A schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0029] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention are described clearly and completely below. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort are also within the scope of protection of the present invention.

[0030] In the description of the present invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings and are intended solely to facilitate and simplify the description of the present invention. They are not intended to indicate or imply that the devices or components referred to must have, be constructed, or operate in a specific orientation, and therefore should not be construed as limitations on the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0031] In the description of the present invention, it should also be noted that, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood broadly. For example, they may refer to fixed, detachable, or integral connections; mechanical or electrical connections; direct or indirect connections through an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on the specific circumstances.

[0032] As mentioned in the background, the existing design of shield tunnel sections with small clearances is mainly based on manual experience, and is generally based on meeting on-site functional and safety requirements as the main considerations. Due to the large number of control factors involved, it fails to establish a linear dynamic equilibrium relationship with key construction control indicators such as budget and construction period, and lacks a basis for judging the optimal separation angle. Specifically, the existing design methods have the following problems:

[0033] 1. The existing discrete angle design is mainly based on empirical assignment, which lacks a basis for judgment and cannot determine whether it is the optimal discrete angle.

[0034] 2. The existing discrete angle design lacks a linear dynamic correlation with key construction control indicators such as budget and construction period. That is, under the premise of meeting the functional requirements and safety, the discrete angle value range, the construction period control range, and the cost control range are unclear.

[0035] 3. Manual design uses empirical value assignment and trial calculation as the main method, and can only select certain value points for trial calculation based on experience.

[0036] Based on the above problems, Figure 1 This is a flow chart of a method for intelligently designing the optimal discrete angle of a small clearance section of a tunnel based on deep learning provided by an embodiment of the present invention. The method is suitable for execution by an electronic device. Figure 1 As shown, the method specifically includes:

[0037] S110 , obtaining an initial value of a discrete angle of a small clearance section of a shield tunnel, and using the initial value as a current value of the discrete angle.

[0038] According to the actual situation of the project, the schematic diagram of the plan design of the small clearance section of the shield tunnel is as follows: Figure 2 As shown in the figure, the discrete angle refers to the angle between the left / right line of the shield tunnel and the centerline of symmetry. Considering engineering geology, functional use, safety, and site conditions, this step preliminarily determines a relatively small discrete angle θ0 as the initial value based on conventional experience. This initial value is used as the current value of the discrete angle, and subsequent calculation cycles are initiated.

[0039] S120. Determine the current value of the small clearance section length according to the current value of the discrete angle and the set shield tunnel spacing.

[0040] According to the shield tunnel design specifications, the spacing less than 3-5 times the shield diameter is generally determined as the minimum clearance segment length M0. Figure 2 , set the shield tunnel spacing to 3-5 times the shield diameter. According to the geometric relationship between the shield tunnel spacing and the discrete angle, the length of the small clearance segment corresponding to the current discrete angle can be determined, and this length can be used as the current value of the small clearance segment length.

[0041] S130. Using stratum grouting and setting reinforcement piles as a small-spacing reinforcement method for shield tunnels, using an AI model based on deep learning, a combination of multiple construction parameters in the reinforcement scheme corresponding to the current values ​​of the discrete angle and the small-clearance segment length is assigned, wherein the multiple construction parameters include pile diameter, pile spacing, pile length, stratum grouting amount, and construction step distance.

[0042] Common methods for reinforcing shield tunnels with small spacing are stratum grouting and setting reinforcement piles. Therefore, this embodiment selects five key construction factors, namely pile diameter, pile length, pile spacing, stratum grouting volume, and left and right shield tunnel construction step (spacing), as control indicators of the engineering plan.

[0043] Among them, the pile diameter D is generally 0.5 to 2m;

[0044] The pile spacing S is generally 0 to 5m;

[0045] The pile length H is generally set to be 1 to 3 times the tunnel diameter below the outer bottom surface of the tunnel;

[0046] The general principle for determining the value of the formation grouting volume V is that the length is the length of the small clear distance section, the width is the clear distance of the tunnel, and the height is the pile length, V = length × width × height;

[0047] The construction step distance ΔL refers to the distance between the construction of two tunnels, which is generally 10 to 20 times the diameter of the tunnel.

[0048] In a specific implementation, the values ​​of the above five key construction parameters in similar engineering cases and specifications can be collected to form a database, and machine learning or artificial neural network algorithms can be used to perform big data learning and training on the AI ​​design model. After a certain amount of repeated training and learning of big data, the ability to independently combine and assign values ​​can be achieved.

[0049] Optionally, the design model uses CNN (Convolutional Neural Network) as the basic structure. Figure 3 The basic structure of a CNN network is shown as an example. A multi-layer perceptron can be added to the front end of the CNN network to expand the input parameters. The expanded features are input into the CNN network to further extract deep features. The CNN network includes multiple layers of convolution units, and each convolution unit includes a convolution layer and a pooling layer. The output of the previous convolution unit serves as the input of the next convolution unit. The expanded features are input into the CNN network, and the deep features are gradually extracted through multiple layers of convolution units, and the implicit constraints between the discrete angles and the lengths of the small clearance segments and the multiple construction parameters are learned layer by layer. The output of the last layer of convolution units is input into a fully connected layer, where the dimension is reduced to the dimension of the output vector, that is, a 5-dimensional vector composed of five construction parameters.

[0050] Optionally, the discrete angle, the length of the small clear distance segment and the value combinations of the multiple construction parameters in the same reinforcement scheme (or engineering scheme) in the database can be used as samples to construct a sample set; and the value ranges of the above five construction parameters are used as constraints, and the sample set is used to train the above design model, so that the input of the model is the value combination of the discrete angle and the length of the small clear distance segment, and the output is the optimal combination of each construction parameter that meets the value range.

[0051] Specifically, during the training process, valid samples that meet the value range can be screened from the sample set; the discrete angle and the length of the small clear distance segment in a valid sample are input into the multi-layer perceptron, and after being processed by the multi-layer perceptron and the CNN model, a 5-dimensional vector is output, corresponding to the values ​​of 5 construction parameters respectively; the difference between the output vector and the value vector of the five construction parameters in the valid sample is calculated as the loss function, and the model parameters are updated by minimizing the loss function.

[0052] After the model training is completed, the current values ​​of the discrete angle and the length of the small clear distance segment are input into the trained model. The output of the model is the combination of values ​​of multiple construction parameters in the reinforcement scheme corresponding to the current values, thereby realizing the AI ​​assignment of the design model.

[0053] S140. Perform a three-dimensional finite element simulation analysis of the tunnel-stratum based on the value combination, and determine whether it meets safety requirements based on mechanical properties and deformation properties.

[0054] This step uses AI-generated values ​​from the design model to conduct a three-dimensional finite element simulation analysis of the tunnel and strata. The mechanical and deformation properties of the simulation results determine whether the current engineering solution, including the breakaway angle, minimum clearance length, and various construction parameters, meets safety regulations and design requirements. The detailed finite element simulation analysis process is currently available and will not be detailed here.

[0055] S150. If the safety requirements are met, the project budget is reviewed; if the project budget is met, the current project solution is stored in the set of feasible solutions; if the project budget is not met, the current value of the discrete angle is increased according to the linear step size, and the process returns to S120 for the next round of calculation until the final value of the discrete angle reaches the set maximum threshold.

[0056] If the current project plan meets the safety requirements, continue to review the project budget of the current project plan.

[0057] If the project budget meets the requirements, the current project plan is considered a feasible plan.

[0058] If the project budget does not meet the requirements, the discrete angle assignment is returned to, the current value of the discrete angle is linearly increased, and the operation of S120 is returned based on the new current value to start the next round of loop calculation until the new current value exceeds the maximum value range of the discrete angle, and the loop is terminated.

[0059] S160: If the safety requirement is not met, increase the current value of the discrete angle according to a linear step size, and return to S120 to perform the next round of cyclic calculation until the final value of the discrete angle reaches the set maximum threshold.

[0060] If the current engineering solution does not meet the safety requirements, the discrete angle assignment is returned to the same way, the current value of the discrete angle is linearly increased, and the operation of S120 is returned to start the next round of loop calculation according to the new current value, until the new current value exceeds the maximum value range of the discrete angle. The whole loop process is shown in Figure 4 .

[0061] S170. From the final set of feasible solutions, determine the discrete angle with the shortest construction period and a cost less than the highest budget, and the discrete angle with the lowest budget and a construction period less than the planned construction period, and determine the interval between the two discrete angles as the optimal interval of the discrete angles.

[0062] The above loop of S120-S160 ultimately yields multiple feasible solutions, forming a set of feasible solutions. From this set, we can determine the discrete angle θ1 with the shortest duration and cost ≤ maximum budget, and the discrete angle θ2 with the lowest budget and duration ≤ planned duration. This allows us to determine the design range of the discrete angle θ, θ1 ≤ θ ≤ θ2, i.e., θ∈[θ1,θ2].

[0063] S180. Determine the optimal interval of the small clearance section length according to the optimal interval of the discrete angle and the set shield tunnel spacing.

[0064] Based on the design value range of the discrete angle θ, the design value range of the small clearance segment length M can be further determined to be M1≤M≤M2, that is, M∈[M1,M2].

[0065] The above two optimal intervals are provided to the designers, who can select the optimal balance angle θe of "small clearance section length-construction period-budget" within these optimal intervals based on the actual engineering conditions such as tunnel stratum geological conditions, usage functions, environmental climate and other human factors. Figure 4 Understand.

[0066] In summary, the present invention provides a deep learning-based intelligent design method for the optimal discrete angle of small-clearance sections in tunnels. Building on the empirical design of small-clearance sections in traditional shield tunnels, this method introduces AI intelligent value assignment and three-dimensional numerical simulation analysis technology. Under the conditions of ensuring tunnel functionality and safety, the AI ​​intelligent value assignment comprehensively covers cyclic calculations to determine the range of discrete angle values, helping designers find the discrete angle θe that best suits engineering requirements and satisfies the optimal balance between "small-clearance section length, construction period, and budget." Specifically, this embodiment can achieve the following beneficial effects:

[0067] 1. Breaking through the limitation of traditional design experience value calculation that can only obtain a limited number of point values, through AI loop calculation, a scientific, accurate, complete and comprehensive discrete angle value range of small clearance section can be obtained for designers' reference.

[0068] 2. Comprehensively consider the key control factors such as safety, construction period, and budget in the small clearance section of shield tunnel projects, establish a dynamic relationship between the design of the discrete angle and safety, construction period, and budget, and construct a discrete angle θe design method with the optimal balance relationship of "small clearance section length-construction period-budget" that meets safety requirements.

[0069] 3. By introducing AI autonomous learning and combination empowerment technology and combining it with three-dimensional numerical simulation analysis technology, the intelligent design method is more scientific, efficient, convenient, accurate, complete and covers the entire range than the traditional design method.

[0070] Figure 5 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention is shown in FIG. Figure 5 As shown, the device includes a processor 60, a memory 61, an input device 62 and an output device 63; the number of processors 60 in the device can be one or more. Figure 5In the embodiment, a processor 60 is used as an example; the processor 60, the memory 61, the input device 62 and the output device 63 in the device can be connected by a bus or other means. Figure 5 The bus connection is taken as an example.

[0071] Memory 61, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the deep learning-based intelligent design method for optimal standoff angles in tunnels with small clearance distances, as described in the embodiments of the present invention. Processor 60 executes the software programs, instructions, and modules stored in memory 61 to perform various functional applications and data processing, thereby implementing the aforementioned deep learning-based intelligent design method for optimal standoff angles in tunnels with small clearance distances.

[0072] The memory 61 may primarily include a program storage area and a data storage area. The program storage area may store an operating system and at least one application required for a function; the data storage area may store data generated based on the use of the terminal. Furthermore, the memory 61 may include high-speed random access memory and non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state memory device. In some instances, the memory 61 may further include memory remotely located relative to the processor 60, and these remote memories may be connected to the device via a network. Examples of such networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0073] The input device 62 may be used to receive input digital or character information and generate key signal input related to user settings and function control of the device. The output device 63 may include a display device such as a display screen.

[0074] An embodiment of the present invention also provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the program implements the intelligent design method for optimal discrete angle of small clearance section of tunnel based on deep learning of any embodiment.

[0075] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. Computer-readable media can be computer-readable signal media or computer-readable storage media. Computer-readable storage media can be, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or components, or any combination thereof. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, 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 thereof. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by an instruction execution system, device or device or used in combination with it.

[0076] A computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0077] Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0078] Computer program code for performing the operations of the present invention can be written in one or more programming languages, or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as C or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0079] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. A deep learning-based intelligent design method for optimal discrete angles in tunnels with small clearance distances, characterized in that: include: Obtaining an initial value of a discrete angle of a small clearance section of a shield tunnel, and using the initial value as a current value of the discrete angle; Determining a current value of the length of the small clearance section according to a current value of the discrete angle and a set shield tunnel spacing; Using stratum grouting and reinforcement pile installation as a small-spacing reinforcement method for shield tunnels, a deep learning-based AI model is used to combine and assign values ​​to multiple construction parameters in the reinforcement scheme corresponding to the current values ​​of the discrete angle and the small-clearance segment length, where the multiple construction parameters include pile diameter, pile spacing, pile length, stratum grouting amount, and construction step distance; Based on the combined values, a three-dimensional finite element simulation analysis of the tunnel and the stratum is performed to determine whether the reinforcement scheme meets the safety requirements based on the mechanical properties and deformation properties; If the security requirements are met, the project budget is reviewed; if the project budget is met, the reinforcement plan is stored in the feasible plan set; If the safety requirements or the project budget are not met, the current value of the discrete angle is increased according to a linear step size, and the process of determining the current value of the small clearance section length according to the current value of the discrete angle and the set shield tunnel spacing is returned to, until the final value of the discrete angle reaches the set maximum threshold; From the final set of feasible solutions, determine the first discrete angle with the shortest duration and cost less than the maximum budget, and the second discrete angle with the lowest cost and duration less than the planned duration; An interval between the first discrete angle and the second discrete angle is determined as an optimal interval of discrete angles.

2. The method according to claim 1, characterized in that Before using the deep learning-based AI model to combine and assign values ​​to a plurality of construction parameters in the reinforcement scheme corresponding to the current values ​​of the discrete angle and the small clearance segment length, the method further includes: Determine the value range of each construction parameter; Extracting values ​​of various construction parameters from engineering cases and / or specifications, and combining the discrete angle, the length of the small clearance section, and the values ​​of the plurality of construction parameters in the same reinforcement scheme as samples; With the value range as a constraint, the deep learning-based AI model is trained using a sample set, so that the input of the model is a combination of values ​​of discrete angles and small clearance segment lengths, and the output is the optimal combination of each construction parameter that satisfies the value range.

3. The method according to claim 2, characterized in that The AI ​​model based on deep learning is trained using a sample set with the value range as a constraint, so that the input of the model is a combination of values ​​of the discrete angle and the length of the small clearance segment, and the output is the optimal combination of each construction parameter that meets the value range, including: Screening valid samples that meet the value range from the sample set; The combination of the discrete angle and the length of the small clearance segment in each valid sample is input into a multi-layer CNN model based on deep learning, and the difference between the output of the model and the combination of the values ​​of each construction parameter in each valid sample is calculated; The model parameters are updated by constraining the minimization of the difference.

4. The method according to claim 3, characterized in that The multi-layer CNN model includes a multi-layer perceptron, a convolutional layer, a pooling layer and a fully connected layer, wherein: The multilayer perceptron is used to expand the input parameters to obtain expanded features with higher dimensions; The convolution layer and the pooling layer together constitute a convolution unit, and the convolution unit has multiple layers, which is used to learn the constraints between the discrete angle and the length of the small clearance segment and the multiple construction parameters layer by layer; The fully connected layer is used to reduce the dimension of the output of the last convolutional unit to a combination of values ​​of the multiple construction parameters.

5. The method according to claim 2, characterized in that The AI ​​model based on deep learning is used to combine and assign values ​​to multiple construction parameters in the reinforcement scheme corresponding to the current values ​​of the discrete angle and the length of the small clearance segment, including: The current values ​​of the discrete angle and the length of the small clear distance segment are input into the trained model, and the output of the model is used as the combined value assignment of multiple construction parameters in the reinforcement scheme corresponding to the current values.

6. The method according to claim 2, characterized in that The shield tunnel spacing is set to be 3-5 times the shield diameter; The range of pile diameter is 0.5 to 2 m; The range of pile spacing is 0 to 5m; The range of pile length is 1-3 times the tunnel diameter below the outer bottom surface of the tunnel; The value range of the formation grouting volume V satisfies: V = length × width × height, where length is the length of the small clear distance section, width is the clear distance of the tunnel, and height is the pile length; The construction step distance ranges from 10 to 20 times the tunnel diameter.

7. The method according to claim 1, characterized in that After determining the interval between the first discrete angle and the second discrete angle as the optimal interval of discrete angles, the method further includes: The optimal interval of the length of the small clearance section is determined according to the optimal interval of the discrete angle and the set shield tunnel spacing.

8. An electronic device, characterized in that: include: one or more processors; a memory for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the deep learning-based intelligent design method for optimal discrete angle of small clearance sections in tunnels as described in any one of claims 1-7.

9. A computer-readable storage medium, characterized in that A computer program is stored thereon, which, when executed by a processor, implements the intelligent design method for optimal discrete angle of a tunnel with small clearance section based on deep learning as described in any one of claims 1-7.

Images