A collaborative design method and system for tunnel support systems based on artificial intelligence

Abstract

The present invention provides a collaborative design method and system for a tunnel support system based on artificial intelligence, relating to the technical field of tunnel support, including: obtaining geological exploration data of a target tunnel and a historical tunnel support system design scheme; performing feature extraction according to the historical tunnel support system design scheme to obtain a historical data vector set; performing machine learning modeling based on the historical data vector set and performing prediction processing based on the geological exploration data of the target tunnel to obtain a support design requirement report; performing topological design of a support structure according to the support design requirement report to obtain a topological model of the support structure; and performing optimization processing according to the topological model to obtain a tunnel support system design scheme. The present invention optimizes the topological relationship of the support structure through a force-directed algorithm, realizes the coordinated layout between support units, effectively reduces construction risks and costs, and provides a more intelligent and efficient solution for the design of the tunnel support system.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel support, and in particular, to a collaborative design method and system for a tunnel support system based on artificial intelligence. Background Art

[0002] At present, tunnel support design plays a crucial role in tunnel engineering, especially under complex geological environments and changing construction conditions. Existing tunnel support design methods usually rely on standardized design formulas and empirical formulas. Although these methods can solve common design problems to a certain extent, they still face many challenges. First, it is often difficult for existing technologies to comprehensively consider the complexity of geological conditions, especially the changes in dynamic factors such as soil type, groundwater level, and stress state, which directly affect the stability and safety of the support structure. Second, the synergistic effects and interdependent relationships between support units are often ignored, resulting in a less refined design of the support structure and the inability to fully utilize the functions of each support unit. In addition, existing design methods rely on manual experience for adjustment and cannot efficiently design complex environments in a short time. They often require multiple iterations and corrections, resulting in a long design cycle and low efficiency. These problems not only limit the adaptability of existing technologies but also increase the risks and costs during tunnel construction.

[0003] Based on the above disadvantages of the existing technology, there is an urgent need for a collaborative design method and system for a tunnel support system based on artificial intelligence. Summary of the Invention

[0004] The purpose of the present invention is to provide a collaborative design method and system for a tunnel support system based on artificial intelligence to improve the above problems. To achieve the above purpose, the technical solutions adopted by the present invention are as follows:

[0005] In a first aspect, the present application provides a collaborative design method for a tunnel support system based on artificial intelligence, including:

[0006] Obtaining geological exploration data of the target tunnel and the design scheme of the historical tunnel support system;

[0007] Performing feature extraction according to the design scheme of the historical tunnel support system to obtain a historical data vector set;

[0008] Performing machine learning modeling according to the historical data vector set and performing prediction processing based on the geological exploration data of the target tunnel to obtain a support design requirement report;

[0009] Performing topological design of the support structure according to the support design requirement report. By creating a topological framework of the support structure, clarifying the mutual relationships and spatial layouts between support units, a topological model of the support structure is obtained;

[0010] Optimize according to the topological model to obtain the design scheme of the tunnel support system.

[0011] In a second aspect, the present application also provides an artificial intelligence-based collaborative design system for a tunnel support system, including:

[0012] An acquisition module for acquiring geological exploration data of a target tunnel and a historical design scheme of a tunnel support system;

[0013] An extraction module for extracting features according to the historical design scheme of the tunnel support system to obtain a historical data vector set;

[0014] A prediction module for performing machine learning modeling according to the historical data vector set and performing prediction processing based on the geological exploration data of the target tunnel to obtain a support design requirement report;

[0015] A construction module for performing topological design of the support structure according to the support design requirement report, and by creating a topological framework of the support structure, clarifying the mutual relationship and spatial layout between support units to obtain a topological model of the support structure;

[0016] An optimization module for performing optimization processing according to the topological model to obtain the design scheme of the tunnel support system.

[0017] The beneficial effects of the present invention are as follows:

[0018] By deeply analyzing the historical tunnel support design scheme and the geological exploration data of the target tunnel, and using algorithms such as machine learning models and graph theory, the present invention accurately quantifies the dependence relationship and collaborative effect between support units, overcoming the problem that the existing design methods cannot effectively handle complex geological conditions and variable construction environments; by introducing advanced technologies such as graph convolutional networks and multi-objective optimization, it can optimize the spatial layout of support units, and while ensuring safety, stability and construction efficiency, reduce manual intervention and correction in the design process, greatly improving the design efficiency and accuracy; in addition, the present invention optimizes the topological relationship of the support structure through a force-directed algorithm, realizes the coordinated layout between support units, effectively reduces construction risks and costs, and provides a more intelligent and efficient solution for the design of the tunnel support system. Description of the Drawings

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present invention, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0020] Figure 1 It is a schematic flow chart of a collaborative design method for a tunnel support system based on artificial intelligence described in an embodiment of the present invention;

[0021] Figure 2 It is a schematic structural diagram of a collaborative design system for a tunnel support system based on artificial intelligence described in an embodiment of the present invention;

[0022] Figure 3 It is a schematic structural diagram of a collaborative design device for a tunnel support system based on artificial intelligence described in an embodiment of the present invention.

[0023] Reference numerals in the figure: 800, a collaborative design device for a tunnel support system based on artificial intelligence; 801, a processor; 802, a memory; 803, a multimedia component; 804, an I / O interface; 805, a communication component; 901, an acquisition module; 902, an extraction module; 903, a prediction module; 904, a construction module; 905, an optimization module. Detailed implementation manners

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a variety of different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but is merely representative of selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the protection scope of the present invention.

[0025] It should be noted that like reference numerals and letters denote like items in the following figures. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. At the same time, in the description of the present invention, the terms "first", "second", etc. are only used for descriptive distinction and cannot be construed as indicating or implying relative importance. Embodiment 1:

[0026] This embodiment provides a collaborative design method for a tunnel support system based on artificial intelligence.

[0027] See Figure 1 , which shows that this method includes steps S100 to S500.

[0028] Step S100, obtain the geological exploration data of the target tunnel and the historical tunnel support system design scheme;

[0029] It is understandable that geological exploration data includes information such as soil type, groundwater level, soil layer bearing capacity, rock formation structure, groundwater permeability, seismic activity, etc. These data can comprehensively reflect the geological characteristics of the tunnel construction area. Specifically, the soil type and soil layer characteristics directly affect the selection of the support design scheme, the groundwater level and permeability are related to the tunnel waterproof design, the rock formation structure and the development of fractures determine the stability requirements of the support structure, and the seismic activity and soil vibration characteristics require the design to consider the seismic performance. The accuracy of the geological exploration data determines the reliability of the support structure design. Therefore, the comprehensiveness and accuracy of the data are ensured through modern exploration techniques such as drilling, ground penetrating radar, and remote sensing. In addition, the historical tunnel support system design scheme provides a reference for the design, which includes the types of support structures used in past tunnel projects, construction techniques, material selection, and the challenges encountered and solutions.

[0030] Step S200: Extract features from the historical tunnel support system design scheme to obtain a historical data vector set;

[0031] It should be noted that the historical tunnel support system design scheme contains multiple key factors, such as the type of support structure, construction method, material use, support unit configuration, geological condition adaptation, etc. These information directly affect the design effect and construction efficiency of the support structure. When extracting features, first, it is necessary to identify the specific features in these schemes, such as the selection of the support structure (such as shotcrete, steel support, etc.), the strength of the material, the construction technology used, the construction sequence, etc. By converting these features into numerical data, it provides input for the subsequent machine learning model. Second, in order to handle the diversity between different design schemes, the design scheme needs to be standardized or normalized during the feature extraction process to make it suitable for machine learning algorithms. In addition, there may be successful and failed cases under different design schemes in the historical data. When extracting these features, it is also necessary to consider their impact on the tunnel support design to ensure that this information can reflect the support effect under different geological conditions. Through the feature extraction of the historical tunnel support design scheme, a historical data vector set is finally constructed. Each data vector represents the design scheme of a historical case and contains various features related to the support structure effect.

[0032] Step S300: Conduct machine learning modeling based on the historical data vector set and perform prediction processing based on the geological exploration data of the target tunnel to obtain a support design requirement report;

[0033] It is understandable that by performing machine learning modeling on the historical data vector set, the model can identify the potential relationships between the support structure configuration and design requirements under different geological conditions based on the experience of past tunnel support design schemes. By training the model, a system with intelligent learning ability can be obtained, which can automatically identify the most suitable support scheme according to the specific geological characteristics. The geological exploration data of the target tunnel will be used as the input of the model and combined with the design cases in the historical data vector set to further improve the prediction accuracy. In practical applications, these geological data not only affect the type selection of the support structure, but also affect many aspects such as the size, strength requirements and construction technology of the structure. By making predictions based on the geological data of the target tunnel, the machine learning model can intelligently adjust the prediction results to ensure that the design requirements are consistent with the specific construction environment.

[0034] Step S400: Conduct topological design of the support structure according to the support design requirement report. By creating a topological framework of the support structure, clarify the mutual relationship and spatial layout between support units to obtain a topological model of the support structure.

[0035] It should be noted that through topological design in this step, complex support design requirements can be transformed into clear structural layouts, effectively managing and optimizing the mutual relationship between support units. Through an accurate topological framework and optimized spatial layout, the support structure can not only meet geological and design requirements, but also improve construction efficiency, reduce possible conflicts and errors during construction, and enhance the overall stability and safety of the support structure.

[0036] Step S500: Perform optimization processing according to the topological model to obtain a tunnel support system design scheme.

[0037] It is understandable that the process of optimization processing involves matching the functional requirements and engineering constraints between support units. For example, the configuration of support units not only needs to consider geological conditions such as soil bearing capacity and groundwater level, but also needs to comprehensively consider factors such as construction period, cost, and material selection. Through multi-objective optimization algorithms (such as particle swarm optimization, genetic algorithms, etc.), a balance can be achieved between these objectives to ensure the rationality of the support unit configuration, avoid over-design or under-design, and thus maximize the performance and benefits of the support system.

[0038] Furthermore, step S200 includes steps S210 to S230.

[0039] Step S210: Conduct modeling of the dependency relationship between structures according to the historical tunnel support system design scheme. By quantifying the spatial and functional dependencies of different support units and calculating the weight relationship between support units, a weighted graph is obtained.

[0040] Specifically, first, the modeling of the inter-structure dependency relationship is achieved by analyzing historical tunnel support system design plans to identify the interactions and dependencies between support units. The functional roles of support units in the support structure are usually interrelated, and the performance of a certain support unit may be affected by other units. For example, shotcrete support needs to work in coordination with steel supports to improve the structural stability, or the spatial position of a support unit may be restricted by the layout of adjacent units. Therefore, when modeling, it is necessary to describe in detail information such as the functional requirements, construction sequence, and geological environment adaptability of each support unit. Next, to quantify the spatial and functional dependencies of different support units, these relationships can be transformed into numerical values through mathematical methods. For example, spatial dependency can be measured by the distance and relative position between support units, while functional dependency may involve parameters such as the bearing capacity and support strength of support units in the geological environment. Through the quantification of these parameters, the relationships between support units can be accurately described. Common quantification methods include representing the degree of coordination between different support units as a numerical value and reflecting the differences in relative importance between units through weighting. Calculating the weight relationship between support units further transforms these quantified dependency relationships into a weighted graph. In the weighted graph, support units are regarded as nodes of the graph, and the edges between nodes represent the dependency relationships between support units. The weight of an edge represents the degree of mutual influence or coordination between support units. The calculation method of the weight can be based on factors such as the distance between support units, functional dependency, and construction difficulty. For example, the weight between two support units may be inversely proportional to their physical distance and directly proportional to their close functional connection. Through this process, the obtained weighted graph can accurately reflect the dependency relationships and interactions between support units.

[0041] Step S220: Process the weighted graph based on the graph convolutional network. By performing node embedding on each support unit and its relationships, learn the collaborative characteristics of support units under different geological conditions to obtain support unit embedding features;

[0042] It can be understood that the graph convolutional network can effectively capture the dependencies between nodes by performing convolutional operations on the local neighborhoods of nodes in the graph. Different from traditional neural networks, the graph convolutional network can directly process graph-structured data. By aggregating the information of neighboring nodes, it learns the high-order features of nodes, so that the representation of nodes in the whole graph can reflect the relationship with surrounding nodes. In the process of node embedding, the graph convolutional network updates the embedding vector of each node through iterative calculation. Each layer of convolutional operation passes the information of adjacent nodes to the current node, thus gradually learning the high-order features of nodes. In the context of support design, node embedding is regarded as the "representation" of support units in the geological environment. It not only contains the feature information of a single support unit, but also can reflect the cooperative effect and mutual dependence between support units by aggregating the information of neighboring nodes. The node embedding features obtained through the training of the graph convolutional network can deeply reveal the dynamic cooperative characteristics between support units. For example, under different soil types, groundwater levels, rock formations and other conditions, how the interaction between support units affects the stability and safety of the overall design. This step utilizes the node embedding ability of the graph convolutional network to transform the complex dependencies between support units into feature vectors that can be processed by machine learning models. These embedding features not only accurately capture the local attributes of each support unit, but also comprehensively consider the mutual influence and cooperative effect between support units, enabling the model to better adapt to the support design requirements under different geological conditions.

[0043] Step S230: According to the support unit embedding features, generate a historical data vector set by fusing historical support design and corresponding geological environment data.

[0044] It can be understood that this fusion process forms a more comprehensive feature vector by connecting or weighted combining the numerical features of historical design schemes and the numerical features of geological data. This process can ensure that both the historical information of the design scheme and the actual situation of the corresponding geological environment are considered in the same vector set, thus more truly reflecting the relationship between support design requirements and geological conditions.

[0045] Furthermore, step S300 includes steps S310 to S330.

[0046] Step S310: According to the historical data vector set, identify the support design patterns under different geological conditions by clustering the historical design schemes, and obtain the support design requirement pattern group.

[0047] In this step, by performing clustering analysis on the historical data vector set, historical cases with similar design requirements and geological conditions can be grouped together to form a support design pattern group. Each group represents the common characteristics of the schemes and methods adopted in support design under a specific geological environment.

[0048] Step S320: According to the support design requirement pattern groups, combine the geological exploration data of the target tunnel to predict the support design requirements and obtain the prediction results.

[0049] Specifically, considering that tunnel support design is not just a simple match based on historical design data, but also needs to consider the complexity and diversity of geological features, an algorithm of weighted similarity matching is introduced. This algorithm first calculates the similarity between the geological exploration data of the target tunnel and each group in the historical data vector set, and assigns different weights according to the similarity, so that the groups more similar to the geological conditions of the target tunnel account for a greater proportion in the prediction. To further improve the prediction accuracy, a multi-level feature extraction model based on deep learning is adopted, such as a hybrid architecture combining convolutional neural network and long short-term memory network. The local features of geological data are extracted through the convolutional layer, and the long short-term memory network layer is used to capture the temporal dynamic information of geological changes. The model can not only learn the static geological-design relationship in historical data, but also learn the dynamic trend of geological condition changes, so as to better adapt to the changes of complex geological conditions in tunnel design. Through this improvement, the prediction method combines the advantages of weighted similarity matching and deep learning model, enabling the model to understand the complex relationship between the geological data of the target tunnel and historical design cases in a higher dimension. When generating the prediction of support design requirements, the geological exploration data of the target tunnel, historical design pattern groups and the similarity weights obtained by the algorithm are comprehensively considered to intelligently predict the most suitable support design scheme for the target tunnel, including technical parameters such as support structure type, material selection, and strength requirements.

[0050] Step S330: Perform fuzzy inference processing according to the prediction results, and obtain the support design requirement report by adaptively adjusting the weights of fuzzy rules.

[0051] It can be understood that the core of this process is to finely adjust the support design requirements through a fuzzy inference system, so that the prediction results can better meet the actual tunnel engineering requirements. First of all, fuzzy inference is a method of dealing with uncertainty and fuzziness through fuzzy logic, which is especially suitable for situations where there are uncertain or fuzzy relationships between geological conditions and support design requirements. In the prediction of support design requirements, the relationship between geological data and design requirements is often not clear, but there is a certain degree of fuzziness. For example, the relationship between soil type and support structure type may vary due to geological changes.

[0052] Therefore, in this step, preliminary fuzzy rules are first generated according to the prediction results, and these rules describe the relationships between the support design requirements under different geological conditions. For example, "if the soil type is soft soil and the groundwater level is high, then shotcrete support is used". These rules are generated based on experience or historical data in the initial stage, but their weights may need to be adjusted according to specific circumstances in actual applications. To ensure the accuracy and adaptability of the rules, a method of adaptively adjusting the weights of the fuzzy rules is adopted. This process learns and adjusts through the feedback of the actual geological conditions and design requirements, dynamically changing the influence of each fuzzy rule so as to more accurately reflect the design requirements of the target tunnel. Preferably, the adaptive adjustment can be achieved through a reinforcement learning algorithm. During each inference process, the weights of the fuzzy rules are adjusted according to the feedback results of the design requirements (such as the stability and economy of the support structure), enabling the system to gradually optimize the inference process and finally obtain more reasonable support design requirements. In this way, the fuzzy inference system can adapt to the changes in different tunnel designs in real time, improving the accuracy and operability of the prediction results.

[0053] Further, step S400 includes steps S410 to S430.

[0054] Step S410: According to the support design requirement report, construct a preliminary relationship graph between support units through a graph theory model guided by preset constraint conditions;

[0055] It should be noted that different from the previous weighted graph, the relationship graph here focuses more on the functional dependencies based on design requirements and the constraints in engineering practice, rather than directly quantifying the intensity of the mutual influence between support units. First of all, the support design requirement report provides the design requirements and required functions of each support unit, such as information on the type, configuration of the support unit, and its adaptability to geological conditions. This information provides the basis for establishing the relationship graph. In the graph theory model, nodes represent each support unit, and edges represent the relationships between support units. However, the edges here do not focus on the quantification of the dependence intensity, but are defined according to specific engineering constraints. For example, the spatial layout of some support units must meet specific spacing requirements, or need to be installed in a specific order to ensure the stability of the support. In this case, the connection of the edges and the relative positions between support units will be set according to these actual constraints. The preset constraints play a guiding role in constructing the preliminary relationship graph. These conditions include the construction sequence, the physical isolation requirements between support units, and the relative position relationship, etc. Through these constraints, it can be ensured that the layout between support units conforms to the limitations in actual construction, and situations that cannot be actually implemented in the design scheme can be avoided. For example, if some support units need to be close to the underground water source for watertightness control, then the spatial dependence between these units needs to be represented by connections in the relationship graph. In other cases, some support units may need to be far away from unstable geological areas, and such constraints also need to be reflected in the relationship graph.

[0056] Step S420: Quantitatively evaluate the priority and interaction of support units according to the preliminary relationship graph, calculate the importance of each support unit in the overall support structure by establishing a weight relationship matrix, and obtain the weight relationship between support units;

[0057] It can be understood that by quantifying the priority and interaction of support units, the role and status of each support unit in the overall design can be clarified. The establishment of the weight relationship matrix makes the optimization of support design more scientific and accurate, not only ensuring the synergy between support units, but also improving the efficiency and feasibility of the design.

[0058] Step S430: According to the weight relationship, construct a spatial distribution model of support units, and determine the positions and synergy of support units based on the mutual relationship and weight distribution to obtain a topological model.

[0059] It should be noted that by constructing a spatial distribution model of the support units, designers can determine the optimal position of each unit in the support structure according to the functional dependencies and interactions between the support units. Support units with high weight values need to be arranged in positions with strong mutual dependencies to enhance the stability of the structure. To achieve this goal, common optimization algorithms such as the force-directed algorithm or the particle swarm optimization algorithm can simulate the mutual attraction and repulsion between the support units, adjust their positions, and ensure that all parts of the structure can work in coordination. In this process, the synergistic effect not only includes the physical proximity of the support units but also involves the collaborative work requirements during the construction process. Finally, the optimized layout generates a topological model that details the positions and mutual dependencies of the support units. This model provides clear guidance for actual construction, ensuring the overall stability, construction efficiency, and safety of the support structure, reducing potential construction conflicts, and improving the accuracy and flexibility of the design.

[0060] Further, step S420 includes steps S421 to S423.

[0061] Step S421: According to the preliminary relationship diagram, use the adjacency matrix to quantify the dependencies and interactions between the support units and establish a functional dependency matrix;

[0062] Specifically, in graph theory, the adjacency matrix is a square matrix, and the elements in the matrix represent the relationship strength between the support units. The numerical values of the elements in the adjacency matrix represent the degree of interaction between the support units. The larger the value, the stronger the dependence or interaction between the support units. This matrix provides quantitative and operable data support for each pair of support units, enabling designers to more clearly understand the mutual relationships between the support units. Based on the adjacency matrix, further construct a functional dependency matrix to transform the functional requirements and interaction relationships of the support units into numerical features. The functional dependency matrix reflects the synergistic effect and functional dependencies between the support units. For example, some support units have strong functional dependencies and may need to be arranged at the same position or in a specific construction stage. The functional dependency matrix clearly represents these relationships, supporting the optimized design and subsequent decision-making of the support structure. Therefore, define the functional dependency matrix , where:

[0063] ;

[0064] In the formula, represents the functional dependency ratio of support unit on support unit ; represents the dependence strength of support unit on support unit ; and Indicates the serial number of the support unit; Is a summation variable; Represents the total number of support units.

[0065] Step S422: According to the functional dependency matrix, quantify the priority among the support units through the analytic hierarchy process, set the relative importance and priority of each support unit based on expert knowledge, and calculate the weight of each support unit in combination with the dependency matrix to obtain the importance score;

[0066] It can be understood that the functional dependency matrix has quantified the mutual relationship and functional dependency among the support units. The next step is to rank the importance of the support units according to these relationships. The analytic hierarchy process is a classic multi-criteria decision-making method. By constructing a judgment matrix, complex decision-making problems are transformed into simple ratio comparisons. Here, the analytic hierarchy process is used to quantify the priority among the support units. Through expert evaluation of the functional requirements, spatial layout, construction sequence, etc. of the support units, the relative importance of each support unit in the overall design is determined. Expert knowledge is usually based on experience and historical data and is used to evaluate the criticality of each support unit. For example, under certain complex geological conditions, some support units may be crucial for stability, so their weights are relatively large. Then, the dependency matrix and expert knowledge are combined through the analytic hierarchy process to calculate the weight of each support unit. The judgment matrix in the analytic hierarchy process calculates the weights of each support unit by pairwise comparison of the relative importance among the support units and obtains a comprehensive importance score. This score reflects the importance and priority of the support unit in the overall support structure. Through this method, the relative position and role of each support unit in the design can be quantified, ensuring that when optimizing the layout, those support units with higher importance are considered first.

[0067] Step S423: Calculate the weight relationship of each support unit in the overall support structure according to the importance score to obtain the weight relationship matrix.

[0068] It should be noted that the weight relationship matrix not only considers the relative importance of a single support unit but also comprehensively considers the mutual dependency relationship among the support units. Through this matrix, the weight distribution and interaction among the support units can be clearly understood, and based on this, the structure can be optimized and adjusted to ensure that the functional and spatial requirements of each unit can be met.

[0069] Furthermore, step S430 includes steps S431 to S433.

[0070] Step S431: Divide the spatial distribution of the support units according to the weight relationship to obtain the spatial distribution model;

[0071] It should be noted that the weight relationship matrix provides a quantified weight value for each pair of support units. These weight values not only characterize the relative importance between units but also reflect their interdependent relationships. When performing spatial distribution segmentation, it is necessary to determine the relative positions of support units in space based on these weight values. For support units with larger weight values, they may need to be arranged relatively close to each other to reduce functional conflicts and increase the synergistic effect of support. Conversely, units with smaller weights can be relatively dispersed or arranged in non-critical positions.

[0072] The construction of the spatial distribution model can be achieved through various optimization algorithms, such as the optimization method based on the force-directed algorithm. In this method, the relationship between support units is regarded as a "force field". Stronger attractive forces are generated between units with larger weights, while weaker attractive forces or repulsive forces are generated between units with smaller weights. By simulating the action of this force field, the positions of support units can be adjusted so that each unit automatically adjusts to the optimal position according to its relative importance and the dependence relationship with other units. This optimization not only considers the spatial distance between support units but also takes into account the contribution of each unit to the entire structure, ensuring the stability of the support structure and the smooth progress of construction.

[0073] Step S432: According to the spatial distribution model, use the force-directed algorithm to optimize the spatial layout of support units, and reduce the conflicts in the layout by simulating the mutual repulsive and attractive forces between support units to obtain the optimized spatial layout result;

[0074] The formula for the repulsive force is:

[0075] ;

[0076] The formula for the attractive force is:

[0077] ;

[0078] Where, represents the repulsive force between support unit and support unit ; represents the constant of the repulsive force; represents the distance between support unit and support unit ; represents the attractive force between support unit and support unit ; represents the constant of the attractive force.

[0079] Step S433: Optimize according to the optimized spatial layout result, and adjust the positions based on the mutual forces between support units to obtain the topological model.

[0080] It can be understood that the basic idea of the force-directed algorithm is to adjust the positions of nodes (support units) by simulating the repulsive and attractive forces between them. The repulsive force is usually stronger when the distance between support units is closer, aiming to prevent support units from getting too close. The attractive force is greater when the dependence relationship between support units is stronger, aiming to ensure that the relative positions of support units in space can reflect their functional cooperation requirements. In practical applications, the position update of support units is completed through an iterative process. In each iteration, nodes are adjusted according to the interaction forces until a balanced state is reached, that is, the interaction forces between all support units are balanced.

[0081] In the specific optimization process, the intensities of the repulsive and attractive forces are affected by the weight relationship matrix. Stronger attractive forces are generated between support units with larger weights, prompting them to approach each other, while units with smaller weights may be dispersed. Through this optimization process, the layout of support units can effectively reduce spatial conflicts, avoid unreasonable overlaps or overly compact layouts between support units, and at the same time ensure that the functional dependence relationships between support units are satisfied.

[0082] Furthermore, step S500 includes steps S510 to S530.

[0083] Step S510: Match the functional requirements of support units with engineering constraints according to the topological model, and obtain the support unit configuration result by analyzing the degree of fit between the functional requirements of each support unit in construction and the geological conditions.

[0084] It is understandable that the topological model has been constructed through the previous steps, providing the dependency relationships, spatial layouts, and functional synergies among the support units. On this basis, the functional requirements of the support units refer to the actual roles of each unit in the support structure, such as bearing capacity, stability, and support function. These functional requirements are usually affected by geological conditions, construction environments, and the design objectives of the support structure. For example, in areas with low soil bearing capacity, some support units may require higher strength or denser configurations, while in geological areas with harder rocks, different types of support units may be needed. Through the analysis of the fit between the functional requirements and geological conditions, it is possible to identify which support units can best adapt to specific geological conditions and ensure that they can meet the stability and safety requirements during construction. Secondly, engineering limitations include factors such as construction techniques, construction time, and budget. These limitations usually inevitably affect the configuration of the support units in actual design. For example, some support units may not be widely used in certain areas due to high construction difficulty, high material costs, etc., and need to be adjusted according to the actual construction situation. In this process, the configuration of the support units is adjusted according to geological conditions, construction requirements, and resource availability to ensure that the design scheme can meet both functional requirements and be efficiently implemented in actual construction.

[0085] Step S520: Conduct finite element analysis based on the support unit configuration results. By simulating the mechanical properties of the support units, verify the adaptability of the support structure in the geological environment to obtain the strength verification results.

[0086] Specifically, the support unit configuration results have been obtained through the previous step. These configuration results provide detailed guidance for the layout, material selection, and mutual relationships among the support units of the support structure. Based on these configuration results, the next step is to transform the design of the support units into a finite element model. In this model, the support units are regarded as a finite number of small units (or elements), and these elements are interconnected through nodes and boundary conditions to simulate the mechanical behavior of the entire support structure. Through the finite element method, each unit can be analyzed in detail, considering its deformation, stress, and strain under different load conditions.

[0087] When simulating mechanical properties, in addition to considering the impact of geological conditions (such as soil type, groundwater level, rock formation characteristics, etc.) on the support unit, it is also necessary to simulate the effects of external loads, such as soil pressure, seismic force, temporary loads during construction, etc. These factors will affect the overall stability of the support structure. Therefore, finite element analysis can provide the response of each support unit under specific working conditions and identify potential weak links or dangerous areas. Through finite element analysis, the final strength verification results will indicate the adaptability and safety of the support structure in the geological environment. These results will include key data such as strength distribution, stress concentration areas, and deformation of each support unit when subjected to mechanical loads.

[0088] Step S530: Optimize the support design scheme according to the strength verification result, and obtain the tunnel support system design scheme by optimizing and adjusting the configuration and mutual relationship of the support units.

[0089] It should be noted that the first step in the optimization process is to adjust the configuration of the support units to ensure that the interdependent support units are reasonably arranged in space. This not only involves adjusting the position of the support units, but also includes the improvement of the stress concentration areas found in the strength verification. For example, by increasing the number of support units, adjusting their arrangement, or adjusting the distance between units, the stress concentration in the local area can be reduced, thereby improving the stability of the overall structure.

[0090] Next, the relationship between the support units is optimized, taking into account the synergy between the support units. For example, if some support units are highly dependent on other units in terms of function, these units need to be configured more closely or coordinated in other ways. During the optimization process, designers may also need to adjust the size, material or structure type of the support units to meet the requirements of the strength verification. For example, some support units may need to be strengthened or use more suitable materials to ensure their stability and safety under specific geological conditions.

[0091] Finally, after optimization and adjustment, the tunnel support system design will be a verified and optimized support structure that can operate stably in the geological environment and meet the technical requirements of the construction process. This design not only enhances the safety and stability of the support structure, but also improves the operability and economy of the construction, ensuring the practical feasibility of the design. Embodiment 2:

[0092] like Figure 2 As shown, this embodiment provides a tunnel support system collaborative design system based on artificial intelligence, and the system includes:

[0093] An acquisition module 901 is used to acquire geological survey data of a target tunnel and a historical tunnel support system design scheme;

[0094] An extraction module 902, configured to perform feature extraction based on a historical tunnel support system design scheme to obtain a historical data vector set;

[0095] A prediction module 903, configured to perform machine learning modeling based on the historical data vector set, and perform prediction processing based on the geological exploration data of the target tunnel to obtain a support design requirement report;

[0096] A construction module 904, configured to perform topological design of the support structure according to the support design requirement report, and clarify the mutual relationship and spatial layout between support units by creating a topological framework of the support structure to obtain a topological model of the support structure;

[0097] An optimization module 905, configured to perform optimization processing according to the topological model to obtain a tunnel support system design scheme.

[0098] Further, in a specific implementation manner disclosed in the present invention, the extraction module 902 includes:

[0099] A first extraction unit, configured to perform structure - to - structure dependency relationship modeling based on the historical tunnel support system design scheme, obtain a weighted graph by quantifying the spatial and functional dependencies of different support units and calculating the weight relationship between support units;

[0100] A second extraction unit, configured to process the weighted graph based on a graph convolutional network, learn the collaborative characteristics of support units under different geological conditions by performing node embedding on each support unit and its relationship, and obtain support unit embedding features;

[0101] A third extraction unit, configured to generate a historical data vector set according to the support unit embedding features by fusing historical support designs and corresponding geological environment data.

[0102] Further, in a specific implementation manner disclosed in the present invention, the prediction module 903 includes:

[0103] A first prediction unit, configured to perform clustering processing on the historical design scheme according to the historical data vector set, identify support design patterns under different geological conditions, and obtain a support design requirement pattern group;

[0104] A second prediction unit, configured to perform support design requirement prediction according to the support design requirement pattern group in combination with the geological exploration data of the target tunnel to obtain a prediction result;

[0105] A third prediction unit, configured to perform fuzzy inference processing according to the prediction result, and obtain a support design requirement report by adaptively adjusting the weights of fuzzy rules. Example 3:

[0106] Corresponding to the above method embodiment, this embodiment also provides a collaborative design device for a tunnel support system based on artificial intelligence. A collaborative design device for a tunnel support system based on artificial intelligence described below can be correspondingly referred to the method for collaborative design of a tunnel support system based on artificial intelligence described above.

[0107] Figure 3 is a block diagram of a collaborative design device 800 for a tunnel support system based on artificial intelligence shown according to an exemplary embodiment. As Figure 3 shown, the collaborative design device 800 for a tunnel support system based on artificial intelligence may include: a processor 801, a memory 802. The collaborative design device 800 for a tunnel support system based on artificial intelligence may further include one or more of a multimedia component 803, an I / O interface 804, and a communication component 805.

[0108] Among them, the processor 801 is used to control the overall operation of the collaborative design device 800 for a tunnel support system based on artificial intelligence to complete all or part of the steps in the above-mentioned collaborative design method for a tunnel support system based on artificial intelligence. The memory 802 is used to store various types of data to support the operation of the collaborative design device 800 for a tunnel support system based on artificial intelligence. These data may include, for example, instructions for any application or method operating on the collaborative design device 800 for a tunnel support system based on artificial intelligence, as well as application-related data, such as contact data, sent and received messages, pictures, audio, video, and so on. The memory 802 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic memory, flash memory, a magnetic disk, or an optical disc. The multimedia component 803 may include a screen and an audio component. Among them, the screen may be a touch screen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signal may be further stored in the memory 802 or sent through the communication component 805. The audio component also includes at least one speaker for outputting audio signals. The I / O interface 804 provides an interface between the processor 801 and other interface modules, and the above-mentioned other interface modules may be a keyboard, a mouse, buttons, etc. These buttons may be virtual buttons or physical buttons. The communication component 805 is used for the collaborative design device 800 for a tunnel support system based on artificial intelligence to communicate with other devices in a wired or wireless manner. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, or 4G, or a combination of one or more of them. Therefore, the corresponding communication component 805 may include: a Wi-Fi module, a Bluetooth module, and an NFC module.

[0109] In an exemplary embodiment, an artificial intelligence-based collaborative design device 800 for a tunnel support system may be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components, and is used to execute the above-mentioned artificial intelligence-based collaborative design method for a tunnel support system.

[0110] In another exemplary embodiment, a computer-readable storage medium including program instructions is further provided. When the program instructions are executed by a processor, the steps of the above-mentioned artificial intelligence-based collaborative design method for a tunnel support system are implemented. For example, the computer-readable storage medium may be the above-mentioned memory 802 including program instructions, and the above program instructions may be executed by the processor 801 of an artificial intelligence-based collaborative design device 800 for a tunnel support system to complete the above-mentioned artificial intelligence-based collaborative design method for a tunnel support system.

[0111] As described above, the above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered within the protection scope of the present invention.

Claims

1. A collaborative design method for a tunnel support system based on artificial intelligence, characterized in that, Including: Obtain the geological exploration data of the target tunnel and the historical tunnel support system design scheme; Perform feature extraction according to the historical tunnel support system design scheme to obtain a historical data vector set; Perform machine learning modeling based on the historical data vector set, and perform prediction processing based on the geological exploration data of the target tunnel to obtain a support design requirement report; Perform topological design of the support structure according to the support design requirement report. By creating a topological framework of the support structure, clarify the mutual relationship and spatial layout between support units to obtain a topological model of the support structure; Perform optimization processing according to the topological model to obtain a tunnel support system design scheme; Among them, performing feature extraction according to the historical tunnel support system design scheme to obtain a historical data vector set includes: Perform modeling of the interdependence relationship between structures according to the historical tunnel support system design scheme. By quantifying the spatial and functional dependencies of different support units and calculating the weight relationship between support units, obtain a weighted graph; Process the weighted graph based on a graph convolutional network. By performing node embedding on each support unit and its relationship, learn the collaborative characteristics of support units under different geological conditions to obtain support unit embedding features; According to the support unit embedding features, generate a historical data vector set by fusing historical support designs and corresponding geological environment data.

2. The collaborative design method of a tunnel support system based on artificial intelligence according to claim 1, characterized in that, Perform machine learning modeling based on the historical data vector set, and perform prediction processing based on the geological exploration data of the target tunnel to obtain a support design requirement report, including: According to the historical data vector set, perform clustering processing on historical design schemes to identify support design patterns under different geological conditions to obtain a support design requirement pattern group; According to the support design requirement pattern group, combine the geological exploration data of the target tunnel to perform support design requirement prediction to obtain a prediction result; Perform fuzzy inference processing according to the prediction result. By adaptively adjusting the weights of fuzzy rules, obtain a support design requirement report.

3. A collaborative design method for a tunnel support system based on artificial intelligence according to claim 1, characterized in that, Perform topological design of the support structure according to the support design requirement report, including: According to the support design requirement report, construct a preliminary relationship graph between support units through a graph theory model guided by preset constraint conditions; Quantitatively evaluate the priority and interaction of support units according to the preliminary relationship graph. By establishing a weight relationship matrix, calculate the importance of each support unit in the overall support structure to obtain the weight relationship between support units; According to the weight relationship, construct a spatial distribution model of support units, and determine the position and collaborative effect of support units based on the mutual relationship and weight distribution to obtain a topological model.

4. A collaborative design method for a tunnel support system based on artificial intelligence according to claim 3, characterized in that, Quantitatively evaluate the priority and interaction of support units according to the preliminary relationship graph, including: According to the preliminary relationship graph, use an adjacency matrix to quantify the dependence relationship and interaction between support units, and establish a functional dependence matrix; According to the functional dependency matrix, the priorities among the support units are quantified by the analytic hierarchy process. Based on expert knowledge, the relative importance and priorities of each support unit are set, and the weights of each support unit are calculated in combination with the dependency matrix to obtain the importance scores. Based on the importance scores, the weight relationship of each support unit in the overall support structure is calculated to obtain the weight relationship matrix.

5. The collaborative design method of a tunnel support system based on artificial intelligence according to claim 4, characterized in that According to the weight relationship, by constructing a spatial distribution model of the support units and determining the positions and synergistic effects of the support units based on the mutual relationships and weight distributions, a topological model is obtained, including: The spatial distribution of the support units is segmented according to the weight relationship to obtain the spatial distribution model. According to the spatial distribution model, the force-directed algorithm is used to optimize the spatial layout of the support units. By simulating the mutual repulsive and attractive forces between the support units, the conflicts in the layout are reduced to obtain the optimized spatial layout result. Based on the optimized spatial layout result, the positions are adjusted based on the mutual forces between the support units to obtain the topological model.

6. The collaborative design method of a tunnel support system based on artificial intelligence according to claim 1, characterized in that Optimization processing is performed according to the topological model, including: The functional requirements of the support units are matched with the engineering constraints according to the topological model. By analyzing the fit between the functional requirements of each support unit during construction and the geological conditions, the support unit configuration result is obtained. Finite element analysis is performed according to the support unit configuration result. By simulating the mechanical properties of the support units, the adaptability of the support structure in the geological environment is verified to obtain the strength verification result. The support design scheme is optimized according to the strength verification result. By optimizing and adjusting the configuration and mutual relationships of the support units, the tunnel support system design scheme is obtained.

7. An artificial intelligence-based collaborative design system for tunnel support systems, characterized in that, Including: An acquisition module for acquiring the geological exploration data of the target tunnel and the historical tunnel support system design scheme. An extraction module for extracting feature vectors from the historical tunnel support system design scheme to obtain a historical data vector set. A prediction module for performing machine learning modeling based on the historical data vector set and performing prediction processing based on the geological exploration data of the target tunnel to obtain a support design requirement report. A construction module for performing topological design of the support structure according to the support design requirement report. By creating a topological framework of the support structure, the mutual relationships and spatial layout among the support units are clarified to obtain the topological model of the support structure. An optimization module for performing optimization processing according to the topological model to obtain the tunnel support system design scheme. Among them, the extraction module includes: A first extraction unit for modeling the structural inter-dependency relationship according to the historical tunnel support system design scheme. By quantifying the spatial and functional dependencies of different support units and calculating the weight relationship between the support units, a weighted graph is obtained. A second extraction unit for processing the weighted graph based on a graph convolutional network. By performing node embedding on each support unit and its relationship, the collaborative characteristics of the support units under different geological conditions are learned to obtain the support unit embedding features. A third extraction unit, configured to generate a historical data vector set by fusing historical support designs and corresponding geological environment data according to the support unit embedding features.

8. The collaborative design system for a tunnel support system based on artificial intelligence according to claim 7, characterized in that, The prediction module includes: A first prediction unit, configured to identify support design pattern groups for different geological conditions by performing clustering processing on the historical design schemes according to the historical data vector set, so as to obtain a support design requirement pattern group; A second prediction unit, configured to predict support design requirements by combining the geological exploration data of the target tunnel according to the support design requirement pattern group, so as to obtain a prediction result; A third prediction unit, configured to perform fuzzy inference processing according to the prediction result, and obtain a support design requirement report by adaptively adjusting the weights of fuzzy rules.

Images