A method, system, equipment, and procedure for dynamic management and control of geological risks in construction tunnels.

Abstract

This invention discloses a method, system, equipment, and program for dynamic management and control of geological risks in construction tunnels, relating to the field of risk assessment technology. The method includes: collecting construction tunnel data, wherein the construction tunnel data includes preliminary geological data, real-time construction data, and environmental data; establishing an initial three-dimensional geological model based on the preliminary geological data; dynamically updating the initial three-dimensional geological model to obtain a three-dimensional geological model; performing risk prediction based on the three-dimensional geological model to obtain prediction results; and quantitatively assessing the risk based on the prediction results to determine whether a risk warning should be issued. When a risk warning is issued, a matching process is performed in a contingency plan database based on the characteristics of the current risk, and the contingency plan with the highest matching degree is selected for execution. Through this disclosed processing scheme, accurate and robust forward-looking predictions of unexcavated geological conditions can be achieved.

Description

Technical Field

[0001] This invention relates to the field of risk assessment technology, and in particular to a method, system, equipment and procedure for dynamic management and control of geological risks in construction tunnels. Background Technology

[0002] In the construction of tunnels and underground engineering projects, geological risk management is a crucial aspect of ensuring construction safety. Traditional methods rely on manual experience and a single data source, making it difficult to achieve accurate risk identification and dynamic assessment.

[0003] In recent years, with the development of geological exploration technology, construction monitoring technology, geophysical exploration technology, and digital technology, multi-source data fusion and intelligent analysis have gradually become a research hotspot. However, existing technologies still suffer from problems such as data fragmentation, static models, and a single early warning mechanism, resulting in low risk management efficiency and high false alarm and false negative rates.

[0004] The specific solutions of existing technologies mainly include: (1) Single data source early warning: single indicator threshold early warning based on geological radar or TSP advanced geological forecast data, lacking multi-factor comprehensive judgment. (2) Static geological model: three-dimensional geological model established using previous geological exploration data, which is not dynamically updated during construction, resulting in a large deviation between the model and the actual geological conditions. (3) Human decision support: risk management mainly relies on the experience of engineers, lacking data-driven quantitative decision-making tools.

[0005] While the aforementioned geological risk control methods for construction tunnels can ensure construction safety to a certain extent, several shortcomings have been found in their structure / methods during actual use, preventing them from achieving optimal results. These shortcomings can be summarized as follows: 1) Data silo problem: Data formats are different in each stage of surveying, design and construction, making it difficult to integrate and utilize them.

[0006] 2) Model lag: Static geological models cannot reflect the actual geological conditions revealed during construction, resulting in a lag in risk assessment.

[0007] 3) Low accuracy of early warning: The false alarm and missed alarm rates of early warning based on single indicator threshold are high.

[0008] 4) Insufficient decision support: Lack of quantitative decision recommendations based on multi-source data fusion.

[0009] Therefore, it is evident that the existing methods for geological risk management in construction tunnels still have inconveniences and shortcomings in practice, and urgently need further improvement. Creating a new method for geological risk management in construction tunnels has become a pressing goal for the industry. Summary of the Invention

[0010] In view of this, the present disclosure provides a method for dynamic management and control of geological risks in construction tunnels, which at least partially solves the problems existing in the prior art.

[0011] In a first aspect, this disclosure provides a method for dynamic management and control of geological risks in construction tunnels, the method comprising the following steps: Collect construction tunnel data; wherein, the construction tunnel data includes preliminary geological data, real-time construction data, and environmental data; An initial three-dimensional geological model was established based on the aforementioned preliminary geological data; The initial three-dimensional geological model is dynamically updated to obtain a three-dimensional geological model. Risk prediction is performed based on the aforementioned three-dimensional geological model, and prediction results are obtained. Based on the prediction results, the risk is quantitatively assessed to determine whether a risk warning should be issued. When a risk warning is issued, the system matches the current risk with the response plan in the contingency plan database based on the characteristics of the risk and selects the response plan with the highest matching degree.

[0012] According to a specific implementation of an embodiment of this disclosure, the step of dynamically updating the initial three-dimensional geological model to obtain a three-dimensional geological model includes: Acquire the latest real-time construction data of the tunneling process; wherein, the latest real-time construction data of the tunneling process includes: newly revealed tunnel face sketches and high-definition images and advanced geological prediction data; The latest real-time construction data of the tunneling is used as driving information to update the initial three-dimensional geological model, thus obtaining a three-dimensional geological model.

[0013] According to a specific implementation of this disclosure, the step of performing risk prediction based on the three-dimensional geological model to obtain prediction results includes: Using data from the excavated area as features and geological attributes as labels, multiple machine learning models were trained. Based on the aforementioned multiple machine learning models, the geological attribute probability prediction of the unexcavated area of ​​the construction tunnel is performed to obtain the geological trend field. Calculate the residuals between the machine learning predictions and actual values ​​at the excavated points, and generate a residual correction field. The geological trend field and the residual correction field are superimposed to generate the final geological attribute prediction probability field; The probability field for geological attribute prediction is quantified based on a Bayesian model to obtain robust geological prediction information and uncertainty measurement.

[0014] According to a specific implementation of an embodiment of this disclosure, the method further includes: The three-dimensional geological model is reverse-corrected and updated based on the predicted probability field of the geological attributes and the actual information of the tunneling.

[0015] According to a specific implementation of this disclosure, the plurality of machine learning models include: random forest, gradient boosting tree, support vector machine, multilayer perceptron, radial basis function network, K-nearest neighbor, XGBoost model, and Bayesian neural network.

[0016] According to a specific implementation of an embodiment of this disclosure, the step of quantitatively assessing the risk based on the prediction result and determining whether to issue a risk warning includes: Select key evaluation indicators; wherein, the key evaluation indicators include at least one of the following: geological complexity index, deformation rate, water inflow, and construction parameter anomaly. Calculate the objective weights of the key evaluation indicators respectively to obtain the objective weight vector; Define the risk level domain; the risk level domain includes: low, medium, high, and extremely high; Based on fuzzy logic, a membership function is constructed for each key evaluation indicator. Calculate the membership degree of each key assessment indicator to each level of risk in the risk level domain, and generate a fuzzy relation matrix; Based on the objective weight vector and the fuzzy relation matrix, the risk probability distribution is obtained; The causal dependency correction of the risk probability distribution is performed based on a Bayesian network to obtain the posterior probability distribution of the risk. The decision fusion method based on DS evidence theory fuses the risk probability distribution and the posterior probability distribution of the risk to obtain a comprehensive risk probability distribution. Calculate the overall risk probability distribution trend within a preset time window to obtain the risk level within the window; The moving average and standard deviation within the window are used to determine whether the risk level is high or extremely high; when the moving average and standard deviation exceed a preset threshold, a warning of the corresponding level is triggered.

[0017] According to a specific implementation of this disclosure, when a risk warning is issued, matching is performed in the contingency plan database based on the characteristics of the current risk, and the contingency plan with the highest matching degree is selected for execution, including: A contingency plan database is constructed based on historical information; each contingency plan description in the database includes at least one of the following: applicable geological conditions, risk type, contingency measures, expected effects, and historical application success rate. When an alert is triggered, obtain the current risk characteristics; Match the current risk characteristics with the descriptions of contingency plans in the contingency plan database based on their similarity, and output at least one contingency plan with the highest similarity. After adopting the response plan with the highest similarity, subsequent monitoring data will be used as feedback to evaluate the effectiveness of the response plan under the current geological risk type.

[0018] Secondly, this disclosure provides a dynamic management and control system for geological risks in construction tunnels, the system comprising: The data acquisition module is configured to collect construction tunnel data, which includes preliminary geological data, real-time construction data, and environmental data. A 3D geological BIM module is configured to establish an initial 3D geological model based on the aforementioned preliminary geological data; and to dynamically update the initial 3D geological model to obtain a 3D geological model. The risk assessment module is configured to perform risk prediction based on the three-dimensional geological model and obtain prediction results; The early warning module is configured to quantitatively assess the risk based on the prediction results and determine whether to issue a risk warning; wherein, when a risk warning is issued, the module matches the current risk characteristics in the emergency response plan library and selects the emergency response plan with the highest matching degree.

[0019] Thirdly, embodiments of this disclosure also provide an electronic device, the electronic device comprising: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor. When the instructions are executed by the at least one processor, they enable the at least one processor to perform the dynamic management and control method for geological risks of construction tunnels as described in the first aspect or any implementation thereof.

[0020] Fourthly, this disclosure also provides a computer program product, which includes a computing program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions, which, when executed by a computer, cause the computer to execute the dynamic management and control method for geological risks of construction tunnels in the first aspect or any implementation thereof.

[0021] The dynamic geological risk management method for construction tunnels disclosed in this embodiment provides a full-cycle, dynamic, intelligent, and closed-loop geological risk management method. Its advantages include: 1) breaking down data barriers and achieving spatiotemporal unification and deep fusion of multi-source heterogeneous data; 2) creating a dynamic geological model that grows with tunneling progress, enabling forward-looking prediction of unexcavated areas; 3) establishing a multi-indicator collaborative intelligent fusion assessment and early warning mechanism to improve early warning accuracy; and 4) forming a data-driven closed-loop decision-making process of risk perception, assessment, early warning, handling, and feedback optimization to improve management efficiency. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of a method for dynamic management and control of geological risks in construction tunnels provided in an embodiment of this disclosure; Figure 2 A flowchart illustrating a method for dynamic management and control of geological risks in construction tunnels, as provided in this embodiment of the disclosure; Figure 3 A schematic diagram of a dynamic geological risk management system for construction tunnels provided in this embodiment of the present disclosure; Figure 4 A schematic diagram of an electronic device provided in an embodiment of this disclosure. Detailed Implementation

[0023] The embodiments of this disclosure will now be described in detail with reference to the accompanying drawings.

[0024] The following specific examples illustrate the implementation of this disclosure. Those skilled in the art can easily understand other advantages and effects of this disclosure from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them. This disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this disclosure. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0025] It should be noted that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this disclosure, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. Furthermore, this device and / or this method can be implemented using other structures and / or functionalities besides one or more of the aspects set forth herein.

[0026] Furthermore, specific details are provided in the following description to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that the described aspects can be practiced without these specific details.

[0027] This invention provides a method for dynamic management and control of geological risks in construction tunnels. The method involves collecting construction tunnel data; establishing an initial three-dimensional geological model based on the preliminary geological data; dynamically updating the initial three-dimensional geological model to obtain a three-dimensional geological model; performing risk prediction based on the three-dimensional geological model to obtain prediction results; and quantitatively assessing the risks based on the prediction results to determine whether to issue a risk warning, thereby achieving accurate and robust forward-looking prediction of unexcavated geological conditions.

[0028] Figure 1 This is a schematic diagram of the dynamic management and control method for geological risks in construction tunnels provided in this embodiment of the disclosure.

[0029] Figure 2 To and Figure 1 The corresponding flowchart for dynamic management and control of geological risks in construction tunnels.

[0030] like Figure 1 As shown, in step S110, construction tunnel data is collected; wherein, the construction tunnel data includes preliminary geological data, real-time construction data, and environmental data.

[0031] More specifically, the first step is to collect preliminary geological data, real-time construction data, and environmental data.

[0032] Preliminary geological data includes: borehole core logging, geophysical reports, geological maps, and geostress test data.

[0033] Real-time construction data includes: advanced geological prediction (tunnel seismic wave prediction TSP, ground-penetrating radar), advanced horizontal drilling, geological sketches and high-definition images of the tunnel face, stress and strain monitoring of support structures, tunnel convergence and crown settlement monitoring, and tunnel boring machine parameters (thrust, torque, tunneling speed).

[0034] Environmental data include: groundwater level, water inflow into the cave, and rainfall.

[0035] Then, a unified spatiotemporal reference operation is performed.

[0036] (1) Spatial datum: The plane coordinate system of all data shall be unified to the CGCS2000 National Geodetic Coordinate System. If the original data is an engineering-independent coordinate system, the transformation parameters (seven or four parameters) between the coordinate system and CGCS2000 shall be established and recorded. The elevation datum shall be unified to the 1985 National Elevation Datum. For vertical data such as tunnel burial depth and borehole depth, the starting surface (such as the ground or tunnel floor) shall be clearly defined, and a conversion relationship shall be established with the 1985 National Elevation Datum.

[0037] (2) Time base: All data timestamps are uniformly in Coordinated Universal Time (UTC) or the specified engineering time, and time zone information is recorded.

[0038] Finally, to address the issue of semantic confusion in the data, the semantic encoding of various data types is unified. An encoding rule of "hierarchical classification + spatiotemporal identification" is designed.

[0039] Project code: Identifies a specific tunnel project, such as "TBM01".

[0040] Exploration point code (e.g., borehole): .in, For project code; Represents drilling; It indicates distance, such as "12+345" which means 12 kilometers and 345 meters from the starting point; It is a sequence number. For example: TBM01-ZK-12+345-001 means TBM01 project, the borehole located at mileage K12+345, and numbered 001.

[0041] Geological structure coding (e.g., faults): .in, Represents a fault; It indicates the orientation code, such as "NE45" indicating a dip of 45° northeast.

[0042] Monitoring point code: .in, Represents stress monitoring; For location descriptions, such as "GD" (vault).

[0043] Face drawing code: .in, Representative sketch, Represents the date.

[0044] By establishing mandatory spatiotemporal benchmarks and coding rules, the heterogeneity problem of multi-source data was fundamentally solved, and a structured geological-construction spatiotemporal database was constructed, laying the data foundation for subsequent fusion analysis and overcoming the drawbacks of data fragmentation in traditional methods.

[0045] More specifically, we now proceed to step S120.

[0046] In step S120, an initial three-dimensional geological model is established based on the aforementioned preliminary geological data.

[0047] More specifically, based on preliminary geological data, including borehole core logging, geophysical reports, geological maps, and geostress test data, an initial 3D geological BIM model is constructed in 3D geological modeling software before construction. ). The model includes the initial spatial distribution of strata and structures (such as faults and folds) inferred from exploration information.

[0048] Next, proceed to step S130.

[0049] In step S130, the initial three-dimensional geological model is dynamically updated to obtain a three-dimensional geological model.

[0050] In this embodiment of the invention, the step of dynamically updating the initial three-dimensional geological model to obtain a three-dimensional geological model includes: acquiring the latest real-time construction data of the tunneling progress; wherein, the latest real-time construction data of the tunneling progress includes: newly exposed face sketches and high-definition images and advanced geological prediction data; using the latest real-time construction data of the tunneling progress as driving information to update the initial three-dimensional geological model to obtain a three-dimensional geological model.

[0051] More specifically, after each tunneling cycle (i.e., the tunnel face advances one predetermined length), the newly exposed tunnel face sketch, high-resolution imagery, and advanced geological prediction data are used as driving information. This driving information is standardized (encoded according to the rules of step S110) and then compared with the model. The model compares the predicted geological information at the corresponding location to generate evidence driving model updates. For example, at the working face at kilometer marker K12+345, the model... The predicted surrounding rock was Class III with a few joints, but the actual sketch revealed a fracture zone with well-developed fissures and water seepage. This discrepancy between the prediction and reality was recorded as evidence. The geological sketch and image data of the tunnel face were used to correct the model of the excavated area (updating the prediction to reality), while the geological sketch and image data of the tunnel face, together with the advanced geological prediction data, served as input information for predicting the geological conditions of the unexcavated area ahead.

[0052] Among them, the geological sketches and high-definition images of the working face mainly include: real geological information such as lithology, joints, faults, and groundwater revealed after excavation; advanced geological prediction data mainly include: indirect detection information of unexcavated areas such as TSP and ground-penetrating radar.

[0053] Next, proceed to step S140.

[0054] In step S140, risk prediction is performed based on the three-dimensional geological model to obtain the prediction result.

[0055] In this embodiment of the invention, the step of risk prediction based on the three-dimensional geological model to obtain prediction results includes: training multiple machine learning models using data from the excavated area as features and geological attributes as labels; performing geological attribute probability predictions on the unexcavated areas of the construction tunnel based on the multiple machine learning models to obtain a geological trend field; calculating the residuals between the machine learning predictions and actual values ​​at the excavated points to generate a residual correction field; superimposing the geological trend field and the residual correction field to generate a final geological attribute prediction probability field; and quantifying the geological attribute prediction probability field based on a Bayesian model to obtain robust geological prediction information and uncertainty measures.

[0056] In this embodiment of the invention, the method further includes: performing reverse correction and updating of the three-dimensional geological model based on the predicted probability field of the geological attributes and the actual information of the excavation.

[0057] In this embodiment of the invention, the plurality of machine learning models include: random forest, gradient boosting tree, support vector machine, multilayer perceptron, radial basis function network, K-nearest neighbor, XGBoost model and Bayesian neural network.

[0058] More specifically, risk prediction based on the aforementioned three-dimensional geological model mainly includes the following steps: 1. Generate machine learning trend fields The location coordinates, burial depth, nearby geophysical anomalies (derived from advanced geological predictions), and historical lithology of each unit in the excavated area are used as features, while the actual revealed geological attributes (such as rock mass quality grade and whether it is a fault) are used as labels. Multiple machine learning models (such as Random Forest, RF, etc.) are trained on these models. Then, these trained models are used to predict the probability of geological attributes for each grid point in the unexcavated area ahead, forming a trend field reflecting large-scale geological patterns. .

[0059] Furthermore, training multiple machine learning models aims to learn the complex relationships in geological data from different perspectives, generating more reliable trend fields and quantifying uncertainties. These models are typically of different types to complement each other. The preferred number of machine learning models is 3-5, which can be selectively chosen and adjusted based on the complexity of the geological conditions, data characteristics (sample size, quality, dimensionality), and core prediction objectives of the specific project, avoiding computational efficiency reduction due to model redundancy.

[0060] The following is a detailed description through examples.

[0061] Scenario 1: Moderate data volume, complex geological conditions, and a pursuit of robust prediction.

[0062] The project has accumulated a certain amount of excavation data (e.g., hundreds of tunnel face logging units), and the geological conditions are varied, including faults and contact zones with different lithologies. The goal is to obtain stable and reliable rock mass quality classification (e.g., RQD, BQ) predictions.

[0063] The chosen machine learning combination is: Random Forest + XGBoost + Support Vector Machine.

[0064] Random forests, as a cornerstone model, have strong resistance to overfitting and can provide feature importance assessment, providing a robust baseline for prediction and helping to understand which features (such as burial depth, distance from faults, and geophysical anomalies) are most critical to prediction.

[0065] As a performance enhancement model, XGBoost / LightGBM can achieve higher prediction accuracy. It refines errors through gradient boosting strategies and is particularly good at capturing complex data patterns and interaction effects.

[0066] Support Vector Machines (SVMs), as a diversity model, differ from tree-based models in their maximum margin classification principle based on kernel functions. When data is linearly separable in a high-dimensional space, SVMs offer a unique predictive perspective, increasing the diversity of the model ensemble.

[0067] Scenario 2: The amount of data is relatively small, or the feature dimensions are high.

[0068] The project has just started and the data sample is limited; or it has introduced a lot of multi-dimensional features (such as multiple geophysical parameters, a complete set of tunneling machine parameters, microseismic monitoring data, etc.).

[0069] The chosen machine learning combination is: Support Vector Machine + Simple Neural Network + LASSO Regression.

[0070] In small sample sizes, by choosing an appropriate kernel function (such as RBF), Support Vector Machines (SVMs) often outperform deep neural networks and are less prone to overfitting.

[0071] Simple neural networks employ a shallow multilayer perceptron (e.g., 1-2 hidden layers). This complements the kernel method of SVM. Overfitting can be prevented through strong regularization (e.g., Dropout, L2), and higher-order interactions between features can be learned automatically.

[0072] As a strongly linear model, LASSO regression not only provides predictions but also has feature selection capabilities. It can automatically compress the coefficients of a large number of irrelevant or redundant features to zero, helping to identify the most critical predictor variables, improving the interpretability of the model, and providing a reference for feature selection for other models.

[0073] Scenario 3: Massive data volume with distinct sequence or spatial features. In TBM tunneling, there is a massive amount of continuous time-series data (thrust, torque, rotational speed, etc. per second); or it is necessary to explicitly consider the spatial sequence dependence of geological properties along the tunnel axis.

[0074] The chosen machine learning combination is: gradient boosting tree + recurrent neural network / long short-term memory network + spatial statistical model.

[0075] Gradient boosting trees are the main force in processing high-dimensional static features and nonlinear relationships, and can effectively integrate the statistical features (such as mean and variance) of time series data with other geological features.

[0076] RNN / LSTM is used to process sequential data. Tunnel mileage can be viewed as a sequence, with the features and geological attributes of each mileage point as time steps, allowing the model to learn the dependency between the geological conditions ahead and historical sequence patterns (such as the changing trends of tunneling parameters over a past distance and the sequence of exposed lithology).

[0077] Ordinary kriging / indicator kriging can directly model the spatial autocorrelation of geological properties. It can predict unknown points without relying on complex features, using only the spatial location and attribute values ​​of known points, providing machine learning models with a purely spatial structure-based predictive perspective and greatly enhancing the diversity of model ensembles.

[0078] 2. Generate a geostatistical residual correction field For each known grid point that has been excavated ( ), calculate the residual between the machine learning prediction and the actual value for each grid point:

[0079] in, The residual between the machine learning prediction and the actual value for each grid point represents the degree to which the purely data-driven trend field prediction deviates from reality. This represents the actual value at the already excavated points; These are machine learning predictions for the already excavated points. and It can be a specific numerical value (such as the rock mass integrity index) or a category probability (such as the probability of belonging to a fault).

[0080] A set of discrete, spatially distributed residual samples is obtained. .

[0081] Furthermore, based on the residual sample point set And its spatial coordinates, calculate the experimental variability function (i.e., the spatial variability characteristic model):

[0082] in, The lag distance is The experimental variogram at time t is used to measure the distance between two points that are vectors apart. The average degree of variation of the residual values ​​between two points; Indicates the lag distance; For all intervals The number of point pairs; For the first The spatial coordinates of a known, excavated sample point; For in position The residual value at the location; In distance point The residual value of another excavated sample point.

[0083] Subsequently, the experimental variation function curves were fitted using theoretical models (such as spherical models and exponential models) to obtain parameters describing the autocorrelation range of the residual space: range, sill value, and nugget value.

[0084] Finally, Kriging interpolation is used to generate the residual correction field.

[0085] Based on grid points of all unexcavated areas Given the spatial coordinates of the point, calculate the residual estimate at that point. When the residuals are continuous values, ordinary kriging is used; when the residuals are class probabilities, indicator kriging is used.

[0086] Taking ordinary kriging as an example, its essence is a spatially optimal linear unbiased estimator. The residual estimate at point A is the weighted sum of the residuals at known points:

[0087] in, For any grid point in the unexcavated area The residual estimate at the location; In the first A known, excavated point The residual value calculated at the location; For the corresponding to the first Known points Kriging weights, It is obtained by solving the Kriging equations.

[0088] Ultimately, a continuous residual correction field covering the entire unexcavated area is obtained. ,in This field characterizes the systematic local correction of the trend field inferred from the known point error space structure.

[0089] 3. Trend and Residual Overlay Trend field With residual correction field Adding them together yields the final probability field for predicting the geological properties of the unexcavated area. :

[0090] 4. Bayesian model average quantization prediction uncertainty Applying multiple different machine learning models to the same grid point Prediction results Weighted average based on posterior probability:

[0091] in, For robust geological prediction information (posterior prediction distribution); The number of machine learning models; Circular index; For the first The posterior model probability of each model; For a specific number A model and all existing data that have been observed Under these conditions, geological properties The predicted distribution.

[0092] The posterior probability of the model is calculated using the following formula. :

[0093] in, For the first Bayesian information criterion values ​​for each model; For the first Bayesian information criterion values ​​for each model; This represents the total number of machine learning models. To ensure that the sum of all weights is 1: .

[0094] Each model is calculated based on the following formula. of value :

[0095] in, For the already excavated verification set (All excavated points) The number of samples in ) For the model In the validation set The mean square error on; For the model The number of valid parameters.

[0096] The calculation model is based on the following formula. In the validation set mean square error :

[0097] in, For the validation set The total number of sample points in the sample; For the validation set The Middle Spatial coordinates of the excavated sample points; This consists of all excavated sample points with authentic geological records. The set that constitutes; For in position The actual geological attribute values ​​observed through on-site geological sketching; For the first A machine learning model for location Predicted values ​​of geological properties.

[0098] The uncertainty measure is calculated based on the following formula, which includes both the uncertainty of each model itself and the discrepancy between models:

[0099] in, To the unexcavated point Regarding geological attributes The overall forecast uncertainty measure; To know all the data Under these conditions, geological properties The posterior variance; The total number of machine learning models participating in the fusion; Indexed by serial number; For the first The posterior probability of each model; For the first Each model itself relative to points Uncertainty (variance) in forecasting; For the first Each model to point Predicted expected value (mean) of geological attributes; This represents the final expected value (BMA mean) after fusing all models.

[0100] The final expected value of the fused models is calculated using the following formula. :

[0101] This step reduces the risk of model selection and provides a quantitative indicator of prediction reliability, which is crucial for subsequent risk assessment.

[0102] 5. Model Iterative Updates and Closed-Loop Feedback The model was replaced with real data from the excavated area, including geological sketches and image data of the tunnel face. The model attributes corresponding to the already excavated portion.

[0103] At the same time, the probability field for predicting the geological attributes of the unexcavated area will be used. and its uncertainty As the best estimate of the unexcavated portion, update the model. The corresponding region. At this point, the model has completed one iteration, from... Updated to .

[0104] Assimilate all data and prediction results from this cycle into the model. Updated model. This will serve as the new base for the next tunneling cycle. When new geological sketches and imagery data of the tunnel face and advanced geological prediction data are input, cycle steps 2-5 will be performed to achieve continuous growth and optimization of the model. → → …).

[0105] Compared to traditional static models or single prediction methods, this method can integrate data patterns and spatial correlations, enabling the model to automatically and dynamically approximate real geological conditions as tunneling progresses. It also provides an uncertainty measure for the prediction results, offering more reliable and cutting-edge geological information input for risk assessment.

[0106] Next, proceed to step S150.

[0107] In step S150, the risk is quantitatively assessed based on the prediction results to determine whether to issue a risk warning.

[0108] In this embodiment of the invention, the step of quantitatively assessing the risk based on the prediction results and determining whether to issue a risk warning includes: selecting key assessment indicators; wherein the key assessment indicators include at least one of geological complexity index, deformation rate, water inflow, and construction parameter anomaly; calculating the objective weights of the key assessment indicators respectively to obtain an objective weight vector; defining a risk level domain; the risk level domain includes: low, medium, high, and extremely high; constructing a membership function for each key assessment indicator based on fuzzy logic membership; calculating the membership degree of each key assessment indicator to each level of risk in the risk level domain, and generating... A fuzzy relation matrix is ​​formed; based on the objective weight vector and the fuzzy relation matrix, a risk probability distribution is obtained; the risk probability distribution is corrected for causal dependence using a Bayesian network to obtain a posterior probability distribution of the risk; the risk probability distribution and the posterior probability distribution of the risk are fused based on DS evidence theory fusion decision to obtain a comprehensive risk probability distribution; the trend of the comprehensive risk probability distribution within a preset time window is calculated to obtain the risk level within the window; the moving average and standard deviation of the risk level within the window are determined to be high or extremely high; wherein, when the moving average and standard deviation exceed a preset threshold, a warning of the corresponding level is triggered.

[0109] More specifically, a quantitative assessment of risk includes the following steps: 1. Selection and normalization of evaluation indicators Select n key indicators, such as geological complexity index, deformation rate, water inflow, and anomaly degree of construction parameters. Normalize these indicators to eliminate dimensions.

[0110] in, The normalized standardized index value; For the first The evaluation sample in the first The original observations on each risk assessment indicator; For the first One indicator.

[0111] 2. Calculating objective weights based on the entropy weight method The calculation of the first based on the following method The entropy value of each indicator:

[0112] in, For the first The information entropy of each risk assessment indicator has a value range of [0,1]. The sample size (the sample size must be more than 3 times the number of indicators); For the first Under the first indicator, the first The feature weight of each sample .

[0113] The index weights are calculated based on the following formula. :

[0114] Obtain the objective weight vector .

[0115] 3. Calculation of membership degree of indicators based on fuzzy logic Define the risk level domain V = {low (L), medium (M), high (H), extremely high (VH)}.

[0116] Construct a membership function for each indicator (e.g., trapezoidal function), calculate the membership degree of each indicator's measured value to each level of risk, and form a fuzzy relation matrix. .

[0117] (1) Constructing the membership function For each risk assessment indicator For each level in the risk level domain, design a membership function. Defined the indicator value The probability of belonging to each level in the risk level domain.

[0118] In this embodiment of the invention, the membership function is a trapezoidal function, a triangular function, or a semi-trapezoidal function. The parameters of the function are determined based on engineering specifications and standards (such as the deformation rate thresholds for different surrounding rock grades in tunnel design specifications), historical data statistics (analyzing historical accident data or expert experience to determine the typical range of each indicator under different risk levels), or expert experience.

[0119] (2) Calculate the membership vector of a single indicator For the current evaluation sample, the first Each indicator is represented by its measured value. Substitute this into all m (4 in this example) membership functions defined for this indicator. The calculations are performed in the middle.

[0120] The calculation yielded:

[0121]

[0122]

[0123]

[0124] Output vector It describes the distribution of the measured values ​​of the indicator across the four risk levels.

[0125] To ensure that the sum of the membership degrees of each indicator at different levels is 1, for each Perform normalization to obtain the normalized result. .

[0126] (3) Assemble the fuzzy relation matrix For all For each evaluation indicator, repeat step (2) to calculate... Membership vectors .

[0127] These row vectors Stack them in order to form a complete fuzzy relation matrix. .

[0128]

[0129] Objective weight vector With fuzzy relation matrix Synthesizing, a preliminary risk level fuzzy vector is obtained. ( (This is a fuzzy synthesis operator), which, after normalization, can be considered as a risk probability distribution based on index weighting. .

[0130] 4. Causal Dependency Correction Based on Bayesian Networks Construct a Bayesian network, with nodes including risk levels. and various indicators Arrows represent causal or dependency relationships.

[0131] Input the measured values ​​of each current indicator as evidence, and obtain the posterior probability distribution of risk through network inference. .

[0132] 5. DS Evidence Theory Fusion Decision Making Will and Basic probability allocation as two independent sources of evidence and Fusion is performed using Dempster's synthesis rules:

[0133] in, Let be the value of the fused basic probability assignment function on proposition A; Indicate the degree of conflict of evidence; To identify a subset of frame Θ, representing high or extremely high risk, Θ = V = {low (L), medium (M), high (H), extremely high (VH)}; The possible conclusion supported by the first source of evidence (fuzzy comprehensive evaluation); Possible conclusions supported by a second source of evidence (Bayesian network reasoning); Based on causal probabilistic reasoning, it is assumed that the risk state conforms to the proposition. Reliability; Based on causal probabilistic reasoning, it is assumed that the risk state conforms to the proposition. The reliability.

[0134] 6. Smart alert for sliding windows The early warning is not based on the risk value at a single moment, but on the trend of the risk probability within a time window.

[0135] Set a sliding time window (the time window length is 4-12 hours, preferably 6 hours).

[0136] Calculate the moving average and standard deviation of the probability of a high or higher risk level within the window. When the current value exceeds the mean plus N times the standard deviation, a warning of the corresponding level is triggered. N is obtained based on historical data training.

[0137] The above methods can effectively filter out instantaneous data spikes, reduce false alarms, and ensure the robustness of early warning systems.

[0138] Next, proceed to step S160.

[0139] In step S160, when a risk warning is issued, the system matches the current risk in the emergency response plan database based on the characteristics of the current risk and selects the emergency response plan with the highest matching degree.

[0140] In this embodiment of the invention, when a risk warning is issued, matching the current risk characteristics with the emergency response plan database and selecting the emergency response plan with the highest matching degree includes: constructing an emergency response plan database based on historical information; wherein each plan description in the database includes at least one of the following: applicable geological conditions, risk type, response measures, expected effects, and historical application success rate; when a warning is triggered, obtaining the current risk characteristics; performing similarity matching between the current risk characteristics and the plan descriptions in the database, and outputting at least one emergency response plan with the highest similarity; after adopting the emergency response plan with the highest similarity, using subsequent monitoring data as feedback to evaluate the effectiveness of the emergency response plan under the current geological risk type.

[0141] More specifically, a contingency plan database is built based on historical information. Each plan includes fields such as applicable geological conditions, risk type, disposal measures (such as adjustment of support parameters, scope of advanced grouting, and changes in process and method), expected effects, and historical application success rate.

[0142] When the system issues an early warning, it simultaneously transforms the current risk characteristics (such as risk location, level, and dominant cause) and the contingency plan descriptions in the contingency plan library into a vector.

[0143] First, the top-K most relevant candidate solutions are retrieved using vector similarity.

[0144] Then, using a large language model (LLM) to deeply understand the current risk context and the details of the candidate solutions, the candidate solutions are intelligently rearranged, adaptively modified, or combined, and 1-3 optimal recommended solutions and their reasons are output.

[0145] After the recommended plan is adopted and implemented, subsequent monitoring data (such as deformation stability and water inrush control effectiveness) will serve as feedback input to the system. The system will assess the effectiveness of the plan under the current geological risk type by comparing changes in risk indicators before and after the intervention, and update the plan's weights or success probabilities accordingly to optimize the parameters of the risk assessment model. This achieves a closed-loop self-learning process of assessment-decision-feedback-optimization, enabling the system's control capabilities to continuously improve over time.

[0146] In this embodiment of the invention, the method further includes digital twin visualization and collaborative management, which can intuitively present complex data and analysis results to support efficient decision-making and multi-party collaboration.

[0147] Develop a 3D visualization control platform (digital twin cockpit) that integrates with the above steps.

[0148] Dynamic model rendering: Displays updated dynamic geological models in real time, highlighting high-risk areas with different colors / transparencies.

[0149] Risk information overlay: Overlay risk heat maps and warning location markers in a 3D scene.

[0150] Multi-dimensional data linkage: Displays monitoring curves, early warning details, LLM-recommended handling solutions, and historical case comparisons.

[0151] Multi-role collaboration: Early warning information and handling procedures can be pushed to the mobile terminals of relevant responsible persons (construction, supervision, owner) in real time, and the handling status can be tracked.

[0152] The dynamic management method for geological risks in construction tunnels proposed in this invention has the following advantages: Deep data integration eliminates information silos: Through unified spatiotemporal benchmarks and coding rules, standardized integration of full-cycle, multi-source data is achieved, providing a reliable data source for precise analysis.

[0153] Dynamic model growth and accurate predictions: The dynamic update framework of machine learning and geostatistics enables the geological model to automatically iterate and optimize as construction progresses, achieving advanced and quantitative predictions of unexcavated geological conditions.

[0154] Intelligent risk assessment and robust early warning: Through a multi-level fusion assessment model combining entropy weight method, fuzzy logic, Bayesian network and DS evidence theory, the scientific nature of risk assessment and the accuracy and robustness of early warning are greatly improved by integrating multi-indicator information and causal logic and combining a sliding window mechanism.

[0155] Data-driven decision support with closed-loop self-optimization: Leveraging LLM intelligent matching and recommendation of solutions, and providing historical data for comparison, transforms decision-making from experience-driven to data-driven. Real-time feedback continuously optimizes the system, forming a self-improving closed-loop capability.

[0156] Visualized and intuitive control, improved collaboration efficiency: The digital twin cockpit presents all risk information and analysis results intuitively, supporting real-time synchronization and collaboration across multiple terminals, which greatly improves the intuitiveness and efficiency of on-site risk control.

[0157] Figure 3 The present invention illustrates a dynamic management and control system 300 for geological risks in construction tunnels, comprising a data acquisition module 310, a three-dimensional geological BIM module 320, a risk assessment module 330, and an early warning module 340.

[0158] The data acquisition module 310 is used to collect construction tunnel data; wherein, the construction tunnel data includes preliminary geological data, real-time construction data and environmental data; The 3D geological BIM module 320 is used to establish an initial 3D geological model based on the aforementioned preliminary geological data; and to dynamically update the initial 3D geological model to obtain a 3D geological model. The risk assessment module 330 is used to perform risk prediction based on the three-dimensional geological model and obtain the prediction results; The early warning module 340 is used to quantitatively assess the risk based on the prediction results and determine whether to issue a risk warning; wherein, when a risk warning is issued, the module matches the current risk in the emergency response plan library according to the characteristics of the current risk and selects the emergency response plan with the highest matching degree.

[0159] See Figure 4 This disclosure also provides an electronic device 40, which includes: At least one processor; and, The memory is communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which enables the at least one processor to perform the dynamic management and control method for geological risks of construction tunnels in the foregoing method embodiments.

[0160] This disclosure also provides a computer program product, which includes a computing program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions that, when executed by a computer, cause the computer to perform the dynamic management and control method for geological risks of construction tunnels in the aforementioned method embodiments.

[0161] The following is for reference. Figure 4 The diagram illustrates a structural schematic of an electronic device 40 suitable for implementing embodiments of the present disclosure. The electronic devices in the embodiments of the present disclosure may include, but are not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 4 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments disclosed herein.

[0162] like Figure 4 As shown, electronic device 40 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 401, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 402 or a program loaded from storage device 408 into random access memory (RAM) 403. The RAM 403 also stores various programs and data required for the operation of electronic device 40. The processing unit 401, ROM 402, and RAM 403 are interconnected via bus 404. Input / output (I / O) interface 405 is also connected to bus 404.

[0163] Typically, the following devices can be connected to I / O interface 405, including: input devices 406 such as touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 407 such as liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 408 such as magnetic tapes, hard disks, etc.; and communication devices 409. Communication device 409 allows electronic device 40 to communicate wirelessly or wiredly with other devices to exchange data. Although electronic device 40 with various devices is shown in the figure, it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed alternatively.

[0164] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 409, or installed from a storage device 408, or installed from a ROM 402. When the computer program is executed by the processing device 401, it performs the functions defined in the methods of embodiments of this disclosure.

[0165] It should be noted that the computer-readable medium described in this disclosure can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this disclosure, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this disclosure, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0166] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device.

[0167] The aforementioned computer-readable medium carries one or more programs that, when executed by the electronic device, cause the electronic device to: acquire at least two Internet Protocol (IP) addresses; send a node evaluation request including the at least two IP addresses to a node evaluation device, wherein the node evaluation device selects an IP address from the at least two IP addresses and returns it; and receive the IP address returned by the node evaluation device; wherein the acquired IP address indicates an edge node in a content delivery network.

[0168] Alternatively, the aforementioned computer-readable medium carries one or more programs that, when executed by the electronic device, cause the electronic device to: receive a node evaluation request including at least two Internet Protocol (IP) addresses; select an IP address from the at least two IP addresses; and return the selected IP address; wherein the received IP address indicates an edge node in the content delivery network.

[0169] Computer program code for performing the operations of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0170] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0171] The units described in the embodiments of this disclosure can be implemented in software or in hardware. The name of a unit does not necessarily limit the unit itself; for example, the first acquisition unit can also be described as "a unit that acquires at least two Internet Protocol addresses".

[0172] It should be understood that the various parts of this disclosure can be implemented in hardware, software, firmware, or a combination thereof.

[0173] The above description is merely a specific embodiment of this disclosure, but the scope of protection of this disclosure is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for dynamic management and control of geological risks in construction tunnels, characterized in that, The method includes the following steps: Collect construction tunnel data; wherein, the construction tunnel data includes preliminary geological data, real-time construction data, and environmental data; An initial three-dimensional geological model was established based on the aforementioned preliminary geological data; The initial three-dimensional geological model is dynamically updated to obtain a three-dimensional geological model. Risk prediction is performed based on the aforementioned three-dimensional geological model, and prediction results are obtained. Based on the prediction results, the risk is quantitatively assessed to determine whether a risk warning should be issued. When a risk warning is issued, the system matches the current risk with the response plan in the contingency plan database based on the characteristics of the risk and selects the response plan with the highest matching degree.

2. The method for dynamic management and control of geological risks in construction tunnels according to claim 1, characterized in that, The dynamic updating of the initial three-dimensional geological model to obtain a three-dimensional geological model includes: Acquire the latest real-time construction data of the tunneling process; wherein, the latest real-time construction data of the tunneling process includes: newly revealed tunnel face sketches and high-definition images and advanced geological prediction data; The latest real-time construction data of the tunneling is used as driving information to update the initial three-dimensional geological model, thus obtaining a three-dimensional geological model.

3. The method for dynamic management and control of geological risks in construction tunnels according to claim 1, characterized in that, The risk prediction based on the three-dimensional geological model yields prediction results, including: Using data from the excavated area as features and geological attributes as labels, multiple machine learning models were trained. Based on the aforementioned multiple machine learning models, the geological attribute probability prediction of the unexcavated area of ​​the construction tunnel is performed to obtain the geological trend field. Calculate the residuals between the machine learning predictions and actual values ​​at the excavated points, and generate a residual correction field. The geological trend field and the residual correction field are superimposed to generate the final geological attribute prediction probability field; The probability field for geological attribute prediction is quantified based on a Bayesian model to obtain robust geological prediction information and uncertainty measurement.

4. The method for dynamic management and control of geological risks in construction tunnels according to claim 3, characterized in that, The method further includes: The three-dimensional geological model is reverse-corrected and updated based on the predicted probability field of the geological attributes and the actual information of the tunneling.

5. The method for dynamic management and control of geological risks in construction tunnels according to claim 3, characterized in that, The various machine learning models include: random forest, gradient boosting tree, support vector machine, multilayer perceptron, radial basis function network, K-nearest neighbor, XGBoost model, and Bayesian neural network.

6. The method for dynamic management and control of geological risks in construction tunnels according to claim 1, characterized in that, The step of quantitatively assessing the risk based on the prediction results and determining whether to issue a risk warning includes: Select key evaluation indicators; wherein, the key evaluation indicators include at least one of the following: geological complexity index, deformation rate, water inflow, and construction parameter anomaly. Calculate the objective weights of the key evaluation indicators respectively to obtain the objective weight vector; Define the risk level domain; the risk level domain includes: low, medium, high, and extremely high; Based on fuzzy logic, a membership function is constructed for each key evaluation indicator. Calculate the membership degree of each key assessment indicator to each level of risk in the risk level domain, and generate a fuzzy relation matrix; Based on the objective weight vector and the fuzzy relation matrix, the risk probability distribution is obtained; The causal dependency correction of the risk probability distribution is performed based on a Bayesian network to obtain the posterior probability distribution of the risk. The decision fusion method based on DS evidence theory fuses the risk probability distribution and the posterior probability distribution of the risk to obtain a comprehensive risk probability distribution. Calculate the overall risk probability distribution trend within a preset time window to obtain the risk level within the window; The moving average and standard deviation within the window are used to determine whether the risk level is high or extremely high; when the moving average and standard deviation exceed a preset threshold, a warning of the corresponding level is triggered.

7. The method for dynamic management and control of geological risks in construction tunnels according to claim 1, characterized in that, When a risk warning is issued, the system matches the current risk characteristics against a pool of emergency response plans and selects the response plan with the highest matching degree. This includes: A contingency plan database is constructed based on historical information; each contingency plan description in the database includes at least one of the following: applicable geological conditions, risk type, contingency measures, expected effects, and historical application success rate. When an alert is triggered, obtain the current risk characteristics; Match the current risk characteristics with the descriptions of contingency plans in the contingency plan database based on their similarity, and output at least one contingency plan with the highest similarity. After adopting the response plan with the highest similarity, subsequent monitoring data will be used as feedback to evaluate the effectiveness of the response plan under the current geological risk type.

8. A dynamic management and control system for geological risks in construction tunnels, characterized in that, The system includes: The data acquisition module is configured to collect construction tunnel data, which includes preliminary geological data, real-time construction data, and environmental data. A 3D geological BIM module is configured to establish an initial 3D geological model based on the aforementioned preliminary geological data; and to dynamically update the initial 3D geological model to obtain a 3D geological model. The risk assessment module is configured to perform risk prediction based on the three-dimensional geological model and obtain prediction results; The early warning module is configured to quantitatively assess the risk based on the prediction results and determine whether to issue a risk warning; wherein, when a risk warning is issued, the module matches the current risk characteristics in the emergency response plan library and selects the emergency response plan with the highest matching degree.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, cause the at least one processor to perform the dynamic management and control method for geological risks of construction tunnels as described in any one of claims 1 to 7.

10. A computer program product, characterized in that, The computer program product includes a computing program stored on a non-transitory computer-readable storage medium, the computer program including program instructions that, when executed by a computer, cause the computer to perform the dynamic management and control method for geological risks of construction tunnels as described in any one of claims 1 to 7.

Images