Intelligent analysis and processing decision system and method for tunnel segment ovality

Abstract

The present invention relates to the technical field of ellipticity analysis and processing decision systems, in particular to a tunnel segment ellipticity intelligent analysis and processing decision system and a method thereof, comprising: a data acquisition module, for: acquiring data of each node in the whole life cycle of a tunnel, and constructing a multi-source heterogeneous data set of the tunnel; an intelligent analysis module, communicatively connected to the data acquisition module, for: receiving multi-source heterogeneous data sent by the data acquisition module; analyzing the segment ellipticity based on three-dimensional point cloud data; constructing an ellipticity change prediction model; a processing solution recommendation module, communicatively connected to the intelligent analysis module, for: recommending an ellipticity processing solution based on an expert knowledge graph; a human-computer interaction module, communicatively connected to the intelligent analysis module and the processing solution recommendation module, for: interactively displaying intelligent analysis results; supporting the input and optimization of analysis models, being able to improve the processing efficiency and quality of tunnel segment ellipticity problems, and being able to provide new ideas and tools for the overall management of tunnel engineering.

Description

Technical Field

[0001] The present invention relates to the technical field of ovality analysis and processing decision systems, in particular to a tunnel segment ovality intelligent analysis and processing decision system and method thereof. Background Art

[0002] With the acceleration of urbanization and the increasing demand for underground space development, tunnel engineering plays an increasingly important role in modern infrastructure construction. However, an important challenge faced during tunnel construction and operation is the ovality of the segments, which not only affects the structural safety of the tunnel, but may also cause a series of problems such as leakage and deformation, thus affecting the service life and operational efficiency of the tunnel.

[0003] Traditional tunnel segment ovality analysis and processing methods mainly rely on manual measurement and empirical judgment. This method has many limitations: first, manual measurement is often time-consuming and labor-intensive, making it difficult to achieve large-scale, high-frequency monitoring; second, measurement accuracy is greatly affected by human factors, making it difficult to ensure data consistency and reliability; third, judgments and decisions based on experience often lack systematicity and scientificity, making it difficult to cope with complex and changing engineering environments.

[0004] In recent years, with the development of information technology, some digital and automated segment ovality monitoring and analysis methods have begun to be applied in practice. For example, laser scanning technology is used to measure tunnel sections, or numerical simulation software is used to predict segment deformation. However, these methods still have some shortcomings: on the one hand, they often only focus on data collection or single-aspect analysis, lacking systematic consideration of the entire problem; on the other hand, these methods are usually static and isolated, and it is difficult to adapt to the dynamic changes of the entire life cycle of tunnel projects, and it is impossible to make full use of accumulated experience and knowledge.

[0005] In addition, the existing segment ovality processing decision mainly relies on the personal experience of engineers and lacks a standardized and intelligent decision support system. This not only easily leads to subjectivity and inconsistency in decision-making, but also makes it difficult to cope with the increasingly complex engineering environment and higher safety requirements.

[0006] In view of the above problems, there is an urgent need for a system that can comprehensively, intelligently and dynamically analyze the ovality of tunnel segments and provide decision support for processing. The system should be able to integrate multi-source heterogeneous data, realize automated ovality analysis and prediction, and make full use of expert knowledge to provide scientific and reasonable treatment solution recommendations. Summary of the invention

[0007] The tunnel segment ovality intelligent analysis and processing decision system provided by the present invention is designed to meet the above requirements. The system realizes comprehensive, intelligent and dynamic management of tunnel segment ovality problems by integrating functional modules such as data collection, intelligent analysis, processing solution recommendation and human-computer interaction.

[0008] The present invention proposes a tunnel segment ovality intelligent analysis and processing decision system, comprising:

[0009] Data acquisition module for:

[0010] Collect data from each node throughout the tunnel life cycle and build a multi-source heterogeneous data set for the tunnel;

[0011] The intelligent analysis module is connected to the data acquisition module for:

[0012] Receiving multi-source heterogeneous data sent by the data acquisition module;

[0013] Analyze the ovality of the pipe segments based on 3D point cloud data;

[0014] Construct a prediction model for ellipticity change;

[0015] The processing solution recommendation module is connected to the intelligent analysis module for:

[0016] Recommend ellipticity processing solutions based on expert knowledge graph;

[0017] The human-computer interaction module is in communication with the intelligent analysis module and the processing solution recommendation module, and is used to:

[0018] Interactively display the results of intelligent analysis;

[0019] Supports the input and optimization of analysis models.

[0020] Preferably, the data acquisition module comprises:

[0021] Tunnel design data acquisition unit, used to collect engineering quantity tables, building crossing tables, stratum lithology information, support parameters, surrounding rock grade, curve parameters and line parameters;

[0022] Monitoring data acquisition unit, used to collect ovality measurement data, deformation monitoring data and BIM model;

[0023] Construction process data collection unit, used to collect construction plans, segment assembly data and surrounding rock quality evaluation data;

[0024] The expert experience data collection unit is used to collect basic information of experts, expert knowledge graphs and ellipticity processing solution information.

[0025] Preferably, the intelligent analysis module comprises:

[0026] The first analysis unit is used to extract key segment data based on the BIM model and point cloud data, perform ellipticity analysis on the key segment data, and support 3D visualization;

[0027] The ellipticity change prediction unit is used to identify various factors that cause ellipticity changes based on the Bayesian network model, according to the segment ellipticity data and construction data, and obtain the ellipticity change prediction during the tunnel operation life based on the time series reasoning framework;

[0028] The decision-making unit is used to analyze the applicability of various ellipticity treatment solutions based on expert experience data and ellipticity change predictions, implement multi-objective comprehensive decision-making, and output recommended decision-making solutions.

[0029] Preferably, the processing solution recommendation module includes:

[0030] The second analysis unit is used to build a knowledge graph of multi-source heterogeneous ellipticity processing solutions based on temporal reasoning theory, supporting the definition and update operations of entities and relationships;

[0031] The multi-objective comprehensive decision-making unit is used to determine the final ellipticity processing solution based on the ellipticity dynamic change prediction results and expert knowledge graph through a multi-objective comprehensive decision-making algorithm.

[0032] Preferably, the human-computer interaction module comprises:

[0033] Model interaction unit, used to support the adjustment, verification and optimization of model parameters for analytical models and decision models;

[0034] Result interaction unit, responsible for displaying historical modeling results and data management;

[0035] The data management unit is used to perform structured processing on multi-source heterogeneous data and supports data cleaning, data conversion and data integration.

[0036] Preferably, the input parameters of the Bayesian network model used by the ellipticity change prediction unit include surrounding rock grade, ring width, ring seam width, lining thickness, convergence value and convergence difference, and the output parameter is the cumulative ellipticity.

[0037] Preferably, the expert knowledge graph constructed by the second analysis unit contains entities including tunnels, segments, key parts of segments, analysis indicators, treatment indicators and ellipticity processing solutions corresponding to each indicator, and the relationships include spatial position relationships, mathematical relationships and causal relationships.

[0038] Preferably, the multi-objective comprehensive decision-making unit adopts the TOPSIS method to calculate the optimization value of each construction node, and outputs the ellipticity processing solution according to the optimization value and the processing experience in the expert experience database.

[0039] Preferably, a model optimization module is also included, for:

[0040] Collect the ellipticity analysis of each project and build a model optimization training set;

[0041] Based on the optimized training set data, the expert knowledge graph and recommendation decision model are calibrated;

[0042] Output optimized analysis model and decision model.

[0043] The intelligent analysis and processing decision method for tunnel segment ovality includes the following steps:

[0044] S1. Collect data from each node in the whole life cycle of tunnel construction through the data acquisition module to build a multi-source heterogeneous data set of the tunnel;

[0045] S2. Build an ellipticity change prediction model based on 3D point cloud data and construction data, obtain various influencing factors that cause ellipticity change based on Bayesian network theory, and deduce the ellipticity change process during the tunnel operation life through a time series reasoning framework;

[0046] S3. Build an expert knowledge graph based on multi-source heterogeneous data. The entities of the expert knowledge graph include tunnels, segments, key parts of segments, analysis indicators, treatment indicators and ellipticity treatment solutions corresponding to each indicator. The relationships include spatial position relationships, mathematical relationships and causal relationships, and build a corresponding rule base.

[0047] S4. Based on the dynamic change of ellipticity and expert knowledge graph, a multi-objective comprehensive decision-making algorithm is used to obtain the recommended decision for the ellipticity treatment solution;

[0048] S5. Interactively display the intelligent analysis results and optimize the analysis model through deep learning.

[0049] The beneficial effects of the present invention are mainly reflected in the following aspects:

[0050] From a macro perspective, the system of the present invention realizes data-driven management of the entire life cycle of tunnel projects. By collecting multi-source heterogeneous data from various stages such as design, construction, and monitoring, the system establishes a comprehensive data foundation, providing rich information support for subsequent analysis and decision-making. This full-life cycle data management method can not only discover and solve problems in a timely manner, but also provide valuable experience feedback for future engineering design and construction.

[0051] At the micro level, the intelligent analysis module of the present invention combines advanced data analysis technology with tunnel engineering expertise to achieve accurate analysis and prediction of segment ovality problems. In particular, the ovality analysis method based on three-dimensional point cloud data greatly improves the accuracy and efficiency of measurement. At the same time, the ovality change prediction model based on the Bayesian network can consider the complex interaction of multiple influencing factors and provide more accurate and reliable prediction results.

[0052] Another important innovation of the present invention is the combination of expert knowledge graphs and data-driven analysis methods. By constructing a knowledge graph containing entities such as tunnels, segments, key parts, and analysis indicators, the system can simulate the expert's reasoning process and provide knowledge support for ellipticity processing. This method can not only make full use of existing engineering experience, but also continuously optimize and improve as new knowledge accumulates.

[0053] In terms of decision support, the present invention adopts a multi-objective comprehensive decision-making algorithm, which can balance multiple objectives such as effect, cost, and safety, and provide more scientific and reasonable treatment plan suggestions. The advantage of this decision-making method is that it can consider multiple aspects of the problem, avoiding the one-sidedness that may occur in traditional decision-making methods.

[0054] In addition, the human-computer interaction module of the present invention facilitates the application and optimization of the system. Through the intuitive visual interface, engineers can better understand and utilize the analysis results of the system. At the same time, the interactive parameter adjustment and model optimization functions enable the system to continuously adapt to new situations and needs and maintain long-term effectiveness.

[0055] In general, the system of the present invention realizes closed-loop management of data collection, analysis, decision-making and optimization through the collaborative work of multiple functional modules. This systematic and intelligent management method can not only improve the processing efficiency and quality of tunnel segment ovality problems, but also provide new ideas and tools for the overall management of tunnel engineering. By reducing human errors and improving the scientificity and timeliness of decision-making, the system of the present invention is expected to significantly improve the safety and economy of tunnel engineering and make important contributions to the development of modern tunnel engineering. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a logic block diagram of the whole system of the present invention.

[0057] Figure 2 It is a logic block diagram of the data acquisition module of the present invention.

[0058] Figure 3 It is a logic block diagram of the intelligent analysis module of the present invention.

[0059] Figure 4The following is a logic block diagram of a module recommended for the processing scheme of the present invention.

[0060] Figure 5 It is a logic block diagram of the human-computer interaction module of the present invention. DETAILED DESCRIPTION

[0061] See also Figure 1-5 The present invention provides a tunnel segment ovality intelligent analysis and processing decision system and method thereof. The system comprises a data acquisition module 1, an intelligent analysis module 2, a processing solution recommendation module 3 and a human-computer interaction module 4.

[0062] The data acquisition module 1 is used to collect data from each node of the entire life cycle of the tunnel and construct a multi-source heterogeneous data set for the tunnel. Preferably, the module can collect various types of data including but not limited to tunnel design data, monitoring data, construction process data and expert experience data. For example, tunnel design data may include engineering quantity tables, building crossing tables, stratum lithology information, etc.; monitoring data may include ellipticity measurement data, deformation monitoring data, etc.; construction process data may include construction plans, segment assembly data, etc.; and expert experience data may include expert basic information, expert knowledge maps, etc. By collecting these multi-source heterogeneous data, the system of the present invention can fully grasp various information in the process of tunnel construction and operation, and provide a rich data basis for subsequent analysis and decision-making.

[0063] The intelligent analysis module 2 is connected to the data acquisition module 1 in communication, and is used to receive the multi-source heterogeneous data sent by the data acquisition module 1. Based on these data, the intelligent analysis module 2 first analyzes the ovality of the pipe segment using the three-dimensional point cloud data. Compared with the traditional two-dimensional measurement method, this analysis method based on three-dimensional point cloud data can more accurately reflect the geometric characteristics of the pipe segment and improve the accuracy of ovality analysis. For example, the system may use the least squares method to fit the point cloud data and calculate the ovality deviation.

[0064] Next, the intelligent analysis module 2 constructs an ellipticity change prediction model. In one embodiment of the present invention, this prediction model may be based on Bayesian network theory. The input parameters of the Bayesian network model may include surrounding rock grade, ring width, ring seam width, lining thickness, convergence value and convergence difference, etc., and the output parameter is the cumulative ellipticity. The advantage of this model is that it can consider the interaction between multiple influencing factors and improve the accuracy of the prediction.

[0065] The treatment solution recommendation module 3 is connected to the intelligent analysis module 2 in communication, and is used to recommend ellipticity treatment solutions based on the expert knowledge graph. The core of this module is to use the expert knowledge graph to represent and reason about complex domain knowledge. For example, the knowledge graph may contain entities such as tunnels, segments, key parts of segments, analysis indicators, treatment indicators, and their spatial position relationships, mathematical relationships, and causal relationships. Through this structured knowledge representation, the system can more effectively utilize expert experience and provide appropriate treatment solutions for specific problems.

[0066] The human-computer interaction module 4 is connected to the intelligent analysis module 2 and the processing solution recommendation module 3 for interactive display of the intelligent analysis results and supports the input and optimization of the analysis model. This module is designed to improve the usability and interpretability of the system. For example, it may provide an intuitive visualization interface to display the ellipticity analysis results and prediction trends; at the same time, it also allows users to adjust model parameters and perform what-if analysis to better understand and utilize the analysis results of the system.

[0067] In a preferred embodiment of the present invention, the data collection module 1 includes a tunnel design data collection unit 11, a monitoring data collection unit 12, a construction process data collection unit 13 and an expert experience data collection unit 14. This subdivided design enables the system to collect and process different types of data more specifically.

[0068] The tunnel design data acquisition unit 11 is specially used to collect key data in the design stage, such as engineering quantity tables, building crossing tables, stratum lithology information, support parameters, surrounding rock levels, curve parameters and line parameters. These data provide the system with basic design information of the tunnel and are an important basis for subsequent analysis.

[0069] The monitoring data acquisition unit 12 is responsible for collecting ellipticity measurement data, deformation monitoring data and BIM models. These real-time monitoring data can reflect the actual conditions of the tunnel during construction and operation, which is crucial for timely discovery and handling of problems. In particular, the introduction of the BIM model provides the system with rich three-dimensional information, which facilitates more accurate analysis and visualization.

[0070] The construction process data collection unit 13 collects the construction plan, segment assembly data and surrounding rock quality evaluation data. These data reflect the specific operation and environmental conditions during the construction process, which are of great significance for understanding the reasons for the change of ovality and predicting future trends.

[0071] The expert experience data collection unit 14 is used to collect basic information of experts, expert knowledge graphs and ellipticity processing scheme information. The design of this unit reflects the importance that the system of the present invention attaches to expert experience. By systematically collecting and organizing expert knowledge, the system can make full use of the experience and wisdom of human experts in the analysis and decision-making process.

[0072] The intelligent analysis module 2 includes a first analysis unit 21, an ellipticity change prediction unit 22 and a decision-making unit 23. This modular design enables the system to perform data analysis, prediction and decision-making step by step, thereby improving the flexibility and maintainability of the system.

[0073] The first analysis unit 21 extracts key segment data based on the BIM model and point cloud data, performs ellipticity analysis on the key segment data, and supports three-dimensional visualization. The innovation of this unit is that it combines BIM technology and point cloud data processing technology. For example, the system may use a point cloud registration algorithm to align the collected point cloud data with the BIM model, and then extract the geometric features of the key segment. The ellipticity analysis may use a least squares ellipse fitting algorithm, which is mathematically expressed as follows:

[0074]

[0075] Among them, x i is a point in the point cloud, a is the ellipse parameter vector, and F is the distance function from the point to the ellipse.

[0076] The ellipticity change prediction unit 22 is based on the Bayesian network model, identifies various factors that cause ellipticity changes according to the segment ellipticity data and construction data, and obtains the ellipticity change prediction during the tunnel operation life based on the time series reasoning framework. A possible implementation of the Bayesian network model is as follows:

[0077]

[0078] Where E is the change in ellipticity, F i The temporal reasoning framework may use dynamic Bayesian networks or hidden Markov models to capture the temporal trend of ellipticity.

[0079] The decision unit 23 analyzes the applicability of various ellipticity treatment solutions based on expert experience data and ellipticity change predictions, implements multi-objective comprehensive decision-making, and outputs a recommended decision solution. This unit may use a multi-criteria decision-making method, such as the analytic hierarchy process (AHP) or the TOPSIS method. For example, when using the TOPSIS method, the distance from each solution to the ideal solution and the negative ideal solution can be calculated, and then the solution with the largest relative proximity can be selected as the optimal solution.

[0080] Through this combination of modularization and algorithms, the system of the present invention can comprehensively analyze the tunnel segment ovality problem, provide accurate predictions and reasonable treatment plan suggestions, thereby improving the quality and efficiency of tunnel construction and operation. The treatment plan recommendation module 3 of the present invention includes a second analysis unit 31 and a multi-objective comprehensive decision-making unit 32. This structural design is intended to make full use of expert knowledge and realize a scientific decision-making process.

[0081] The second analysis unit 31 constructs a multi-source heterogeneous ellipticity processing solution knowledge graph based on temporal reasoning theory. This unit supports the definition and update operations of entities and relationships, which enables the knowledge graph to dynamically adapt to new situations and experiences. For example, in one embodiment of the present invention, the knowledge graph may use RDF (Resource Description Framework) triples to represent knowledge. A typical RDF triple may be as follows:

[0082] (Segment cracks lead to increased ovality)

[0083] (Grouting reinforcement, relief, segment cracks)

[0084] This representation method allows the system to flexibly describe complex domain knowledge and support efficient query and reasoning. Preferably, the system can also use an ontology language (such as OWL) to define richer semantic relationships, thereby improving the accuracy and reasoning ability of knowledge representation.

[0085] The multi-objective comprehensive decision unit 32 determines the final ellipticity processing solution based on the ellipticity dynamic change prediction results and the expert knowledge graph through a multi-objective comprehensive decision algorithm. In a preferred embodiment of the present invention, this unit may use the TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method to make multi-objective decisions. The core idea of ​​the TOPSIS method is to select the solution that is closest to the ideal solution and farthest from the negative ideal solution. Its mathematical expression is as follows:

[0086]

[0087] Among them, C i is the relative closeness of the ith solution, and are the distances from the solution to the negative ideal solution and the ideal solution respectively. The system will choose C i The largest solution is recommended as the final solution.

[0088] The human-computer interaction module 4 of the present invention comprises a model interaction unit 41, a result interaction unit 42 and a data management unit 43. This design aims to provide comprehensive human-computer interaction functions, making the system easier to use and more flexible.

[0089] The model interaction unit 41 supports the adjustment, verification and optimization of model parameters for analysis models and decision models. For example, the user may adjust the weights of various factors in the Bayesian network model or modify the importance coefficients of various objectives in the TOPSIS method through this unit. This interactive capability allows the system to be fine-tuned according to specific circumstances to improve the accuracy of analysis and decision-making.

[0090] The result interaction unit 42 is responsible for displaying the historical modeling results and data management. In one embodiment of the present invention, this unit may provide an interactive visualization interface, allowing the user to browse the historical analysis results, compare the ellipticity change trends in different periods, or view the effect evaluation of different treatment plans. This function helps the user to deeply understand the analysis results of the system and provide support for long-term decision-making.

[0091] The data management unit 43 performs structured processing on multi-source heterogeneous data and supports data cleaning, data conversion and data integration. This unit is particularly important when processing large amounts of heterogeneous data. For example, it may use an ETL (Extract, Transform, Load) process to process data from different sources. During data cleaning, the system may use statistical methods to detect and process outliers, such as using the Z-score method:

[0092]

[0093] Where X is the observed value, μ is the mean, and σ is the standard deviation. Usually, when |Z|>3, the observed value can be considered an outlier.

[0094] In the ellipticity change prediction unit 22 of the present invention, the input parameters of the Bayesian network model include surrounding rock grade, ring width, ring gap width, lining thickness, convergence value and convergence difference, and the output parameter is cumulative ellipticity. This parameter selection is based on a deep understanding of tunnel engineering practice. For example, the surrounding rock grade directly affects the stability of the tunnel, and thus affects the deformation of the segment; the ring width and the ring gap width are closely related to the geometric characteristics of the segment and affect its stress state; the lining thickness is related to the stiffness of the structure; and the convergence value and the convergence difference directly reflect the deformation state of the tunnel.

[0095] Preferably, the system may use a conditional probability table (CPT) to quantify the relationship between these parameters. For example, given the surrounding rock grade and lining thickness, the conditional probability distribution of the annular crack width may be as follows:

[0096]

[0097] This probabilistic model can capture the complex interactions between parameters and improve the accuracy of predictions.

[0098] The expert knowledge graph constructed by the second analysis unit 31 of the present invention includes entities such as tunnels, segments, key segments, analysis indicators, treatment indicators, and ellipticity treatment solutions corresponding to each indicator, and relationships include spatial position relationships, mathematical relationships, and causal relationships. This comprehensive knowledge representation method enables the system to simulate the expert's reasoning process.

[0099] For example, in a knowledge graph, there may be the following relationship chain:

[0100] (Segment joints are located at the key parts of the segment)

[0101] (The cracking of the pipe segment joints leads to an increase in ovality)

[0102] (Grouting reinforcement, relief, cracking of pipe segment joints)

[0103] (Grouting reinforcement, requirement: grouting pressure <0.5MPa)

[0104] This knowledge representation not only includes the logic of problem diagnosis, but also includes specific solutions and their constraints. By reasoning about this complex knowledge network, the system can provide targeted processing suggestions for specific situations.

[0105] The present invention realizes comprehensive intelligent analysis and processing decision-making of the tunnel segment ovality problem through the collaborative working mode of multiple modules, and provides strong technical support for improving the quality and efficiency of tunnel engineering. The multi-objective comprehensive decision-making unit 32 of the present invention uses the TOPSIS method to calculate the optimization value of each construction node, and outputs the ovality processing plan based on the optimization value and the processing experience in the expert experience database. The application of the TOPSIS method enables the system to make balanced decisions while considering multiple objectives.

[0106] In a preferred embodiment of the present invention, the specific steps of the TOPSIS method are as follows:

[0107] First, construct a decision matrix. Assuming there are m alternatives and n evaluation indicators, the decision matrix X can be expressed as:

[0108]

[0109] Among them, x ij It represents the evaluation value of the i-th scheme for the j-th indicator.

[0110] Next, normalize the decision matrix to obtain the normalized decision matrix R:

[0111]

[0112] Then, the weighted normalized decision matrix V is calculated:

[0113] v ij =w j r ij

[0114] Among them, w j is the weight of the jth indicator.

[0115] Next, determine the ideal solution and the negative ideal solution:

[0116]

[0117] in, and are the maximum and minimum values ​​of the j-th indicator respectively.

[0118] Compute the distance of each solution to the ideal solution and the negative ideal solution:

[0119]

[0120] Finally, the relative closeness of each solution is calculated:

[0121]

[0122] Relative proximity C i The larger the value, the better the solution. The system will select C i The largest solution is taken as the final ellipticity processing solution.

[0123] The present invention also includes a model optimization module 5 for continuously improving the performance of the system. This module collects the ellipticity analysis of each project and constructs a model optimization training set. Based on this optimized training set data, the model optimization module 5 corrects the expert knowledge graph and the recommended decision model, and finally outputs the optimized analysis model and decision model. In one embodiment of the present invention, the model optimization process may adopt an iterative method. For example, for the optimization of the Bayesian network model, the maximum likelihood estimation (MLE) method may be used to update the model parameters:

[0124]

[0125] Among them, θ is the model parameter, X i For observation data.

[0126] For the optimization of knowledge graph, the system may use association rule mining algorithms, such as Apriori algorithm, to discover potential knowledge rules from new data. For example, support and confidence can be calculated:

[0127]

[0128]

[0129] When the support and confidence of a rule exceed the preset threshold, the system will add it to the knowledge graph.

[0130] The present invention also provides a tunnel segment ovality intelligent analysis and processing decision method, comprising the following steps:

[0131] S1. Collect data from each node of the whole life cycle of tunnel construction through data acquisition module 1 to construct a multi-source heterogeneous data set of the tunnel. This step lays a data foundation for subsequent analysis and decision-making. In a preferred embodiment of the present invention, the collected data may include but is not limited to: tunnel design parameters (such as tunnel diameter, lining thickness, etc.), construction parameters (such as excavation method, support method, etc.), monitoring data (such as convergence deformation, stress strain, etc.) and environmental parameters (such as geological conditions, hydrological conditions, etc.).

[0132] S2. Construct an ellipticity change prediction model based on three-dimensional point cloud data and construction data. This step uses Bayesian network theory to obtain various influencing factors that cause ellipticity changes, and uses a temporal reasoning framework to deduce the ellipticity change process during the tunnel's operating life. For example, the system may first use the RANSAC algorithm to extract the segment contour from the point cloud data, and then use the least squares method to fit the ellipse and calculate the ellipticity. The Bayesian network may consider the following factors: surrounding rock grade, lining thickness, excavation method, support method, groundwater level, etc. The temporal reasoning framework may use a dynamic Bayesian network or a long short-term memory network (LSTM) to capture the trend of ellipticity changes over time.

[0133] S3. Construct an expert knowledge graph based on multi-source heterogeneous data. In one embodiment of the present invention, the process of constructing a knowledge graph may include the following steps: first, extract entities and relationships from text data (such as engineering reports, expert opinions, etc.), possibly using named entity recognition (NER) and relationship extraction technology; then, use ontology engineering methods to define concept hierarchies and relationship types; finally, store and manage the knowledge graph through a graph database (such as Neo4j).

[0134] S4. Based on the dynamic changes of ellipticity and the expert knowledge graph, a multi-objective comprehensive decision-making algorithm is used to obtain the recommended decision for the ellipticity treatment solution. In this step, the system may first infer in the knowledge graph based on the current state to generate a set of candidate treatment solutions. Then, the TOPSIS method is used to evaluate and rank these candidate solutions. The evaluation indicators may include: treatment effect, cost, construction difficulty, safety, etc.

[0135] S5. Interactively display the intelligent analysis results and optimize the analysis model through deep learning. In a preferred embodiment of the present invention, the interactive display may use Web-based 3D visualization technology to allow users to view the ellipticity analysis results from different angles and scales. Model optimization may use reinforcement learning methods to continuously adjust model parameters based on user feedback and actual results.

[0136] Through this method, the present invention can realize comprehensive, intelligent and dynamic analysis and processing of the tunnel segment ovality problem, providing strong technical support for the quality control and safety management of tunnel engineering.

[0137] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. Tunnel segment ovality intelligent analysis and processing decision system, characterized by: include: Data acquisition module for: Collect data from each node throughout the tunnel life cycle and build a multi-source heterogeneous data set for the tunnel; The intelligent analysis module is connected to the data acquisition module for: Receiving multi-source heterogeneous data sent by the data acquisition module; Analyze the ovality of the pipe segments based on 3D point cloud data; Construct a prediction model for ellipticity change; The processing solution recommendation module is connected to the intelligent analysis module for: Recommend ellipticity processing solutions based on expert knowledge graph; The human-computer interaction module is in communication with the intelligent analysis module and the processing solution recommendation module, and is used to: Interactively display the results of intelligent analysis; Supports the input and optimization of analysis models.

2. The system according to claim 1, characterized in that The data acquisition module comprises: Tunnel design data acquisition unit, used to collect engineering quantity tables, building crossing tables, stratum lithology information, support parameters, surrounding rock grade, curve parameters and line parameters; Monitoring data acquisition unit, used to collect ovality measurement data, deformation monitoring data and BIM model; Construction process data collection unit, used to collect construction plans, segment assembly data and surrounding rock quality evaluation data; The expert experience data collection unit is used to collect basic information of experts, expert knowledge graphs and ellipticity processing solution information.

3. The system according to claim 1, characterized in that The intelligent analysis module includes: The first analysis unit is used to extract key segment data based on the BIM model and point cloud data, perform ellipticity analysis on the key segment data, and support 3D visualization; The ellipticity change prediction unit is used to identify various factors that cause ellipticity changes based on the Bayesian network model, according to the segment ellipticity data and construction data, and obtain the ellipticity change prediction during the tunnel operation life based on the time series reasoning framework; The decision-making unit is used to analyze the applicability of various ellipticity treatment solutions based on expert experience data and ellipticity change predictions, implement multi-objective comprehensive decision-making, and output recommended decision-making solutions.

4. The system according to claim 1, characterized in that The processing solution recommendation module includes: The second analysis unit is used to build a knowledge graph of multi-source heterogeneous ellipticity processing solutions based on temporal reasoning theory, supporting the definition and update operations of entities and relationships; The multi-objective comprehensive decision-making unit is used to determine the final ellipticity processing solution based on the ellipticity dynamic change prediction results and expert knowledge graph through a multi-objective comprehensive decision-making algorithm.

5. The system according to claim 1, characterized in that The human-computer interaction module comprises: Model interaction unit, used to support the adjustment, verification and optimization of model parameters for analytical models and decision models; Result interaction unit, responsible for displaying historical modeling results and data management; The data management unit is used to perform structured processing on multi-source heterogeneous data and supports data cleaning, data conversion and data integration.

6. The system according to claim 3, characterized in that The input parameters of the Bayesian network model adopted by the ellipticity change prediction unit include surrounding rock grade, ring width, ring seam width, lining thickness, convergence value and convergence difference, and the output parameter is the cumulative ellipticity.

7. The system according to claim 4, characterized in that The expert knowledge graph constructed by the second analysis unit contains entities including tunnels, segments, key parts of segments, analysis indicators, treatment indicators and ellipticity processing solutions corresponding to each indicator, and relationships include spatial position relationships, mathematical relationships and causal relationships.

8. The system according to claim 4, characterized in that The multi-objective comprehensive decision-making unit adopts the TOPSIS method to calculate the optimization value of each construction node, and outputs an ellipticity processing solution based on the optimization value and the processing experience in the expert experience database.

9. The system according to claim 1, characterized in that Also includes model optimization modules for: Collect the ellipticity analysis of each project and build a model optimization training set; Based on the optimized training set data, the expert knowledge graph and recommendation decision model are calibrated; Output optimized analysis model and decision model.

10. The intelligent analysis and processing decision method for tunnel segment ovality is characterized by: The following steps are involved: S1. Collect data from each node in the whole life cycle of tunnel construction through the data acquisition module to build a multi-source heterogeneous data set of the tunnel; S2. Build an ellipticity change prediction model based on 3D point cloud data and construction data, obtain various influencing factors that cause ellipticity change based on Bayesian network theory, and deduce the ellipticity change process during the tunnel operation life through a time series reasoning framework; S3. Build an expert knowledge graph based on multi-source heterogeneous data. The entities of the expert knowledge graph include tunnels, segments, key parts of segments, analysis indicators, treatment indicators and ellipticity treatment solutions corresponding to each indicator. The relationships include spatial position relationships, mathematical relationships and causal relationships, and build a corresponding rule base. S4. Based on the dynamic change of ellipticity and expert knowledge graph, a multi-objective comprehensive decision-making algorithm is used to obtain the recommended decision for the ellipticity treatment solution; S5. Interactively display the intelligent analysis results and optimize the analysis model through deep learning.

Images