A method for limited dewatering assessment, construction and monitoring of a subsea tunnel

Abstract

The application discloses a limited drainage evaluation construction and monitoring method for a submarine tunnel, and the evaluation method comprises the following steps: determining a limited drainage scheme based on the characteristics of the submarine tunnel, simulating construction according to the limited drainage scheme, and obtaining monitoring data under different environments; collecting initial sample data and establishing an analysis and prediction feedback model, training the analysis and prediction feedback model by using the initial sample data set; assigning the monitoring data to the analysis and prediction feedback model; cleaning the monitoring data to obtain complete monitoring data; processing the complete monitoring data to obtain accurate monitoring data; processing the accurate monitoring data by using a loss function to obtain this round of sample data; assigning the this round of sample data to the analysis and prediction feedback model for analysis and evaluation; and adjusting the pressure relief valve and / or the construction scheme according to the evaluation result until the evaluation is qualified. According to the qualified limited drainage scheme, limited drainage construction and monitoring are performed on the submarine tunnel.

Description

Technical Field

[0001] This invention relates to the field of tunnel construction technology, specifically a method for limited drainage assessment, construction, and monitoring of submarine tunnels. Background Technology

[0002] The construction environment for undersea tunnels is complex, surrounded by high-pressure seawater. The permeability of the soil and rock can lead to continuous seawater infiltration, increasing construction difficulty and potential safety risks. Water seepage in undersea tunnels not only affects construction progress but can also cause structural deformation, corrosion, and other safety hazards, potentially impacting the tunnel's long-term service life. Due to the harsh and variable environmental conditions of undersea tunnels, and the complexity of construction plans, subsequent feasibility assessments are challenging. In the past, feasibility assessments were often conducted by construction experts. However, human assessments are often limited by individual knowledge and experience, and it is difficult to consider the complex and multifaceted influencing factors, easily overlooking important ones. This leads to insufficient and subjective evaluation of the construction plan, resulting in potential risks during construction. Furthermore, manual evaluation is often cumbersome, time-consuming, and inefficient, easily affecting project progress and decision-making efficiency. Moreover, even after the undersea tunnel is completed and put into use, there is still a chance of sudden situations such as water inrush, affecting the safe use of the undersea tunnel. However, due to the complex and changeable environment, manual monitoring alone is not effective. Firstly, relying solely on manual monitoring is labor-intensive and cannot cover all aspects. Even if monitoring personnel detect a sudden situation, the reaction time required for manual operation prevents timely and effective responses, resulting in irreversible losses. Summary of the Invention

[0003] In order to at least overcome the above-mentioned shortcomings in the prior art, the purpose of this application is to provide a limited drainage assessment construction and monitoring method for submarine tunnels.

[0004] This application provides a limited waterproofing and drainage assessment method for submarine tunnels, including:

[0005] S1: Determine the emission limit scheme based on the characteristics of the undersea tunnel;

[0006] S2: Simulate construction based on the limited emission scheme, and use environmental data, water level data and tunnel structure data under different environmental conditions as monitoring data;

[0007] S3: Collect initial sample data and establish an analysis, prediction and feedback model; use the initial sample dataset to train the analysis, prediction and feedback model.

[0008] S4: Assign the monitoring data collected every 25 minutes to the analysis, prediction and feedback model;

[0009] S5: Perform data cleaning on the monitoring data to remove outliers and obtain complete monitoring data;

[0010] S6: Standardize and normalize the complete monitoring data to obtain accurate monitoring data;

[0011] S7: Using the loss function The precise monitoring data is then optimized to obtain the sample data for this round, wherein... Indicates the amount of sample data. For monitoring data, The simulated prediction data corresponding to the monitoring data;

[0012] S8: Assign the sample data from this round to the analysis, prediction, and feedback model for analysis and evaluation;

[0013] S9: Adjust the pressure relief valve and / or construction plan according to the evaluation results until the evaluation meets the standards.

[0014] Traditional construction scheme evaluation often relies on senior experts and construction personnel in the construction industry to assess the feasibility of the scheme. However, this method is often subjective and limited. Due to the complex geographical environment of the undersea tunnel, multiple factors need to be considered during the human evaluation process, which increases the difficulty of the evaluation and makes the evaluation process time-consuming. Moreover, based on the complex multiple influencing factors, important influencing factors are easily overlooked during the evaluation process, which may lead to potential risks during construction. The purpose of this application is to abandon the human evaluation method and use a predictive model to evaluate the feasibility of the emission restriction construction scheme.

[0015] In this embodiment, construction personnel formulate a preliminary limited discharge plan based on parameters such as the drainage volume, lining permeability coefficient, lining deformation coefficient, water flow velocity, surrounding rock stress, elastic modulus, Poisson's ratio, water pressure, hydraulic head, hydraulic gradient, historical hydrological data, and regional climate of the undersea tunnel. A three-dimensional finite element model is established, and the finite element model is meshed using an appropriate mesh density. The aforementioned parameters are assigned to the finite element model to determine the boundary conditions and initial conditions of the tunnel outline. The limited discharge plan is simulated in the finite element model. The Navier-Stokes equations are used to describe the fluid flow behavior in the tunnel under different conditions. The simulation calculation program is started, and the water flow distribution, pressure changes, and stress state under different conditions are obtained through iterative calculation. Based on the simulation results, a precise limited discharge plan is determined.

[0016] Environmental data, water level data, and tunnel structure data under different environmental conditions are fed into the analysis and prediction feedback model as monitoring data. This is used to simulate the segmented deployment of environmental monitoring sensors, groundwater level monitors, and surveillance cameras within the tunnel. The simulation assumes that these devices collect real-time monitoring data such as temperature, humidity, air pressure, groundwater level, and tunnel crack information within the tunnel model. The monitoring data is transmitted to the analysis and prediction feedback model every 25 minutes. After a series of screening and optimization processes within the analysis and prediction feedback model, a predictive evaluation of the current drainage scheme for the tunnel model is made. Based on the current evaluation results, corresponding adjustments are made to the pressure relief valve and / or drainage scheme until the evaluation is satisfactory.

[0017] In the embodiments of this application, by collecting complex and large amounts of monitoring data and performing preprocessing steps such as data cleaning, standardization and normalization, the efficiency of data processing and analysis is improved and possible human errors or omissions are avoided.

[0018] During training, the predictive feedback model can learn the complex relationships between data. By simulating and predicting water level data, it can obtain more accurate results. Compared with manual assessment, the predictive feedback model can consider various influencing factors more comprehensively and systematically, improve the accuracy of prediction, and thus provide feedback assessment results more quickly.

[0019] The analysis, prediction, and feedback model in the evaluation method of this application embodiment can achieve real-time dynamic adjustment and optimization. Based on the continuously updated monitoring data feedback, the emission scheme is iteratively optimized and evaluated. This real-time dynamism makes the emission restriction scheme more adaptable and flexible, and better suited to the complex and ever-changing underwater tunnel environment.

[0020] In one possible implementation, in step S1, the characteristics of the submarine tunnel include tunnel drainage volume, lining permeability coefficient, lining deformation coefficient, water flow velocity, surrounding rock stress, elastic modulus, Poisson's ratio, water pressure, hydraulic head height, hydraulic gradient, historical hydrological data, and regional climate data.

[0021] A three-dimensional finite element model, including drainage volume and water pressure, was constructed based on the characteristics of the undersea tunnel.

[0022] A three-dimensional finite element model was used to perform simulation calculations to determine the emission limit scheme.

[0023] In one possible implementation, in step S2, the environmental data includes multiple sets of temperature data, humidity data, and air pressure data collected by environmental monitoring sensors within the tunnel.

[0024] The water level data is used to simulate multiple sets of groundwater level data collected inside the tunnel by the groundwater level monitor;

[0025] The tunnel structure data is used to simulate multiple sets of tunnel crack information data collected by surveillance cameras.

[0026] In one possible implementation, step S3 includes:

[0027] The initial sample dataset consists of a large number of historical sample datasets under simulated construction conditions;

[0028] The initial sample dataset is divided into n initial sample data subsets according to feature attributes, and each initial sample data subset is trained n+1 times.

[0029] The learning efficiency of different groups' initial sample data subsets is adjusted through a dynamic learning rate adjustment mechanism.

[0030] By using a cross-correlation mechanism, strong correlations are established between initial sample data subsets of different monitoring factors;

[0031] The prediction and evaluation results of the analysis, prediction, and feedback model are validated through cross-validation until the validation results meet the prediction criteria.

[0032] In one possible implementation, step S5 includes using the Z-score method with a machine learning model to automatically identify and filter outomas.

[0033] In one possible implementation, anomaly monitoring values ​​are initially identified using the Z-score method, and the machine learning model uses the ClusteringAlgorithms algorithm to further screen and filter the anomaly monitoring values.

[0034] In one possible implementation, steps S8 and S9 include:

[0035] If the assessment result is unqualified and the unqualified influencing factor involves the pressure relief value, an alarm will be triggered. Through the intelligent feedback mechanism of the analysis and prediction feedback model, the expected adjustment range of the pressure relief value will be automatically fed back.

[0036] Adjust the pressure relief value to the expected adjustment range and reassess it until the pressure relief value is adjusted to meet the assessment requirements.

[0037] If the evaluation result is unqualified and the unqualified influencing factor does not involve the pressure relief value, an alarm will be triggered, and a new emission restriction plan will be formulated and evaluated until the emission restriction plan is qualified.

[0038] If the evaluation is successful, the sample data from this round will be used as training data to update the initial sample dataset.

[0039] In one possible implementation, a construction method for limited drainage and waterproofing of submarine tunnels includes the following steps:

[0040] S81: Based on the qualified limited-quantity waterproofing and drainage simulation construction plan, a waterproof layer is laid between the initial support layer and the secondary lining layer of the submarine tunnel.

[0041] S82: Based on the qualified limited-quantity waterproofing and drainage simulation construction plan, drainage strips and waterstops shall be installed in the submarine tunnel;

[0042] S83: Based on the qualified limited-scale waterproofing and drainage simulation construction plan, sidewall water storage tanks, water diversion channels and central drainage ditches shall be set up in the submarine tunnel.

[0043] S84: Based on the qualified limited drainage simulation construction plan, circumferential drainage pipes and longitudinal drainage pipes shall be installed in the submarine tunnel.

[0044] S85: Install pressure relief valves at the outlet of the longitudinal drainage pipe, and adjust the pressure relief values ​​of each pressure relief valve according to the expected adjustment range of the feedback pressure relief value. Construction is then complete.

[0045] In this embodiment, multiple waterproof layers are laid on the surface of the initial support layer, gradually extending downwards. Special sealing materials are used to reinforce the tightness of the joints and corners between the waterproof layers, the joints and corners between the waterproof layer and the initial support layer, and the joints and corners between the waterproof layer and the secondary lining layer. Special adhesive is used to seal the drainage strips and waterstops at the expansion joints and construction joints to ensure the airtightness of the drainage strips and waterstops and prevent large amounts of water leakage due to displacement. The circumferential drainage pipe and the longitudinal drainage pipe are connected through a T-junction to ensure that the water in the circumferential drainage pipe is led to the longitudinal drainage pipe in the sidewall. The water flow in the longitudinal drainage pipe is controlled by a pressure relief valve to discharge the water to the reservoir. The central drainage channel is buried deep in the tunnel floor slab, and the water in the reservoir is transported to the central drainage channel for discharge through a water diversion channel.

[0046] In one possible implementation, a monitoring method for limited drainage and waterproofing of submarine tunnels includes the following steps:

[0047] S101: Multiple sets of environmental monitoring sensors are installed sequentially in sections at the arch waist and sidewalls on both sides of the undersea tunnel.

[0048] S102: Multiple sets of monitoring cameras are installed in sections at the arch of the undersea tunnel;

[0049] S103: Multiple sets of groundwater level monitors are installed sequentially at the arch foot and arch waist of the undersea tunnel;

[0050] S104: Every 25 minutes, the monitoring data collected in real time is assigned to the analysis, prediction and feedback model for evaluation;

[0051] S105: If the assessment fails, the pressure relief valve and / or construction plan at the abnormal location shall be adjusted until the assessment passes.

[0052] In this embodiment, the environmental monitoring sensor includes a stress-strain sensor, a temperature sensor, and a humidity sensor. The environmental monitoring sensor, groundwater level monitor, and surveillance camera are all equipped with communication modules. Through the surveillance camera and AI image analysis technology, abnormal conditions in the tunnel structure are automatically identified. Specifically, the surveillance camera continuously captures images inside the tunnel, and a filtering algorithm is used to reduce noise in the images of each area. An object detection algorithm is used to filter out image objects containing cracks. Histogram equalization technology is used to make the crack information deformation and surrounding rock deformation in the image more obvious. Feature points are extracted from the crack areas in the image, and then the feature points of the image are further normalized. The boundary points of the feature points are processed by an edge detection algorithm, thereby obtaining the length and width data of the tunnel cracks. The real-time monitoring data is transmitted to the analysis and prediction feedback model for evaluation every 25 minutes.

[0053] Compared with the prior art, the present invention has the following advantages:

[0054] (1) This invention improves the efficiency of data processing and analysis by collecting complex and large amounts of monitoring data and performing preprocessing steps such as data cleaning, standardization, and normalization, and avoids possible human errors or omissions.

[0055] (2) The analytical prediction feedback model of the present invention can learn the complex relationship between data and obtain more accurate results by simulating and predicting water level data. Compared with manual assessment, the analytical prediction feedback model can consider various influencing factors more comprehensively and systematically, improve the accuracy of prediction, and thus provide feedback assessment results more quickly.

[0056] (3) The analysis, prediction and feedback model in the evaluation method of the present invention can realize real-time dynamic adjustment and optimization. Based on the continuously updated monitoring data feedback, the emission scheme is iteratively optimized and evaluated. This real-time dynamism makes the evaluation of emission restriction schemes more adaptable and flexible, and better adapts to the complex and ever-changing underwater tunnel environment.

[0057] (4) After the construction is completed, the submarine tunnel is monitored in real time by setting up monitoring devices. The monitoring devices transmit environmental data, hydrological data and tunnel structure data to the analysis and prediction feedback model through the communication module for evaluation, and determine whether the water output of the submarine tunnel exceeds the preset standard range. This measure greatly reduces the proportion of manual operation in the entire monitoring and measurement, greatly eliminates the impact of human intervention on the data, ensures the authenticity of the monitoring data, and then quickly feeds back the monitoring results, which plays a role in the subsequent maintenance work of the tunnel. Attached Figure Description

[0058] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and form part of this application, do not constitute a limitation thereof. In the drawings:

[0059] Figure 1 This is a flowchart of the limited waterproofing and drainage assessment method for submarine tunnels according to the present invention;

[0060] Figure 2 This is a flowchart of the limited drainage and waterproofing construction method for submarine tunnels according to the present invention;

[0061] Figure 3 This is a flowchart of the limited drainage and waterproofing monitoring method for submarine tunnels according to the present invention. Detailed Implementation

[0062] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0063] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0064] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0065] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0066] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of the second feature includes the first feature directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0067] Furthermore, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of the present invention.

[0068] In the following description, suffixes such as "module," "part," "component," or "unit" are used only for the purpose of illustrative purposes and have no specific meaning in themselves. Therefore, they can be used interchangeably.

[0069] The present invention will now be described in further detail with reference to specific embodiments and accompanying drawings.

[0070] Please refer to the following: Figure 1 This is a schematic diagram of the limited waterproofing and drainage assessment method for submarine tunnels provided in this application embodiment. Further, a limited waterproofing and drainage assessment method for submarine tunnels may specifically include the contents described in steps S1-S9.

[0071] S1: Determine the emission limit scheme based on the characteristics of the undersea tunnel;

[0072] S2: Simulate construction according to the aforementioned limited emission scheme, and use environmental data, water level data, and tunnel structure data under different environmental conditions as monitoring data;

[0073] S3: Collect initial sample data and establish an analysis, prediction and feedback model, and train the analysis, prediction and feedback model using the initial sample dataset;

[0074] S4: Assign the monitoring data collected every 25 minutes to the analysis, prediction and feedback model;

[0075] S5: Perform data cleaning on the monitoring data to remove outliers and obtain complete monitoring data;

[0076] S6: Standardize and normalize the complete monitoring data to obtain accurate monitoring data;

[0077] S7: Using the loss function The precise monitoring data is then optimized to obtain the sample data for this round, wherein... Indicates the amount of sample data. For monitoring data, The simulated prediction data corresponding to the monitoring data;

[0078] S8: Assign the sample data of this round to the analysis and prediction feedback model for analysis and evaluation;

[0079] S9: Adjust the pressure relief valve and / or construction plan according to the evaluation results until the evaluation meets the standards.

[0080] In this embodiment, environmental data, water level data, and tunnel structure data are packaged and integrated into a unified dataset. Timestamps are used to synchronize and upload the data periodically, ensuring that monitoring data from different influencing factors can be compared and analyzed at the same point in time. In this embodiment, a loss function is utilized. Data optimization was performed on the precise monitoring data to obtain the sample data for this round, among which, Indicates the amount of sample data. For monitoring data, The method used is an existing algorithm for simulated prediction data corresponding to the monitoring data, and the specific calculation logic will not be described in detail here. In this embodiment, an n-fold strategy is adopted to divide the entire initial sample dataset into n parts. The n-1 parts are used for model training, and the remaining part is used for testing and verification. This process is repeated n+1 times. This method helps to prevent the model from overfitting. In the evaluation phase, the evaluation calculation is performed based on the monitoring sample data of this round to obtain the evaluation result of this round of monitoring. A condition triggering mechanism is set up to compare the evaluation result with the evaluation result of historical sample data under similar or the same environmental conditions. If the evaluation result deviates significantly from the historical trend, an alarm is triggered.

[0081] In one possible implementation, the characteristics of the submarine tunnel in step S1 include tunnel drainage volume, lining permeability coefficient, lining deformation coefficient, water flow velocity, surrounding rock stress, elastic modulus, Poisson's ratio, water pressure, water head height, hydraulic gradient, historical hydrological data, and regional climate data.

[0082] A three-dimensional finite element model, including drainage volume and water pressure, was constructed based on the characteristics of the submarine tunnel.

[0083] The emission limit scheme is determined by performing simulation calculations using the three-dimensional finite element model.

[0084] In this embodiment of the application, the construction surveyors obtained parameters such as drainage volume, lining permeability coefficient, lining deformation coefficient, water flow velocity, surrounding rock stress, elastic modulus, Poisson's ratio, water pressure, hydraulic head height, hydraulic gradient, historical hydrological data and regional climate data of the undersea tunnel construction section through on-site measurements and construction standards. Then, after formulating a preliminary limited discharge plan, a three-dimensional finite element model was established.

[0085] In this embodiment, an appropriate mesh density is used to mesh the three-dimensional finite element model. The above parameters are assigned to the finite element model to determine the boundary conditions and initial conditions of the tunnel outline. The construction of the limited discharge scheme is simulated in the finite element model. The Navier-Stokes equations are used to describe the fluid flow behavior in the tunnel under different conditions. The simulation calculation program is started, and the water flow distribution, pressure change and stress state under different conditions are obtained through iterative calculation. Based on the simulation results, the accurate limited discharge scheme is determined.

[0086] In one possible implementation, in step S2, the environmental data includes multiple sets of temperature data, humidity data, and air pressure data collected by environmental monitoring sensors inside the tunnel.

[0087] The water level data is used to simulate multiple sets of groundwater level data collected in the tunnel by the groundwater level monitor;

[0088] The tunnel structure data is used to simulate multiple sets of tunnel crack information data collected by surveillance cameras.

[0089] In this embodiment, environmental data, water level data, and tunnel structure data under different environmental conditions are fed as monitoring data to the analysis and prediction feedback model. This is used to simulate the segmented deployment of environmental monitoring sensors, groundwater level monitors, and surveillance cameras within the tunnel. The simulation shows that the above devices collect monitoring data such as temperature, humidity, air pressure, groundwater level, and tunnel crack information within the tunnel model in real time. The monitoring data is transmitted to the analysis and prediction feedback model every 25 minutes. After a series of screening and optimization processes within the analysis and prediction feedback model, a predictive evaluation of the current tunnel model's emission restriction scheme is made.

[0090] In one possible implementation, step S3 includes:

[0091] The initial sample dataset consists of a large number of historical sample datasets under simulated construction conditions;

[0092] The initial sample dataset is divided into n initial sample data subsets according to feature attributes, and each initial sample data subset is trained n+1 times.

[0093] The learning efficiency of the initial sample data subsets for different groups is adjusted through a dynamic learning rate adjustment mechanism.

[0094] Through a cross-correlation mechanism, strong correlations are established between the initial sample data subsets of different monitoring factors;

[0095] The prediction and evaluation results of the analysis, prediction, and feedback model are verified through a cross-validation mechanism until the verification results meet the prediction criteria.

[0096] In this embodiment, training the analysis and prediction feedback model with a large amount of historical sample data improves its generalization ability, thus better adapting to assessment and prediction under different construction and hydrogeological conditions. The n-fold strategy divides the initial sample dataset into n parts, using n-1 parts for model training and the remaining part for testing and validation, repeating this process n+1 times. This method helps prevent overfitting. A dynamic learning rate adjustment mechanism adjusts the learning efficiency of the analysis and prediction feedback model in real time, accelerating convergence and avoiding local optima. A cross-correlation mechanism helps the model comprehensively consider the influence of different factors on the drainage scheme and tunnel structure, thereby improving its predictive ability. A cross-validation mechanism enhances the overall performance of the model. The stability and robustness of the analysis and prediction feedback model are described. Through extensive analysis and prediction training, the analysis and prediction feedback model outputs environmental expected values, environmental threshold values, water level expected values, water level threshold values, tunnel structure expected values, tunnel structure threshold values, pressure relief expected values, and pressure relief threshold values ​​under different environmental conditions within the tunnel. When the absolute value of the difference between a set of input data and its corresponding expected value data is greater than its corresponding threshold value, the evaluation result is unqualified. The unqualified input data is sent as an influencing factor to the display device for display, and an alarm prompt is triggered. In this embodiment, the environmental expected values ​​are the expected values ​​of temperature, humidity, and air pressure; the environmental threshold values ​​are the threshold values ​​of temperature, humidity, and air pressure; the tunnel structure expected values ​​are the expected values ​​of tunnel crack length and tunnel crack width; and the tunnel structure threshold values ​​are the threshold values ​​of tunnel crack length and tunnel crack width.

[0097] In one possible implementation, step S5 includes using the Z-score method with a machine learning model to automatically identify and filter out anomalous data.

[0098] In this embodiment, the Z-score method is used to initially identify abnormal monitoring values ​​and null values ​​in the monitoring data. After removing null values, the machine learning model uses the ClusteringAlgorithms algorithm to further filter and screen the abnormal monitoring values. Since the monitoring data is uploaded every 25 minutes, a large amount of monitoring data has been accumulated and collected during this process. Even if abnormal values ​​and null values ​​appear in the data, removing the abnormal values ​​and null values ​​will not affect the integrity of the entire monitoring dataset, and thus will not affect the evaluation results of the analysis, prediction and feedback model.

[0099] In one possible implementation, steps S8 and S9 include:

[0100] If the evaluation result is unqualified and the unqualified influencing factor involves the pressure relief value, an alarm is triggered. Through the intelligent feedback mechanism of the analysis and prediction feedback model, the expected adjustment range of the pressure relief value is automatically fed back.

[0101] Based on the expected adjustment range of the pressure relief value, adjust the pressure relief value to the expected adjustment range and re-evaluate until the pressure relief value is adjusted to meet the evaluation requirements;

[0102] If the evaluation result is unqualified and the unqualified influencing factor does not involve the pressure relief value, an alarm will be triggered, and a new emission restriction plan will be formulated and evaluated until the emission restriction plan is qualified.

[0103] If the evaluation is successful, the sample data from this round will be used as training data to update the initial sample dataset.

[0104] In this embodiment, during the evaluation phase, evaluation calculations are performed based on the current round of monitoring sample data to obtain the evaluation results for this round of monitoring. If the evaluation result is unqualified, an alarm is triggered. Simultaneously, an existing anomaly detection algorithm is used to determine whether the pressure relief value for this round is abnormal. If the pressure relief value is normal, manual readjustment or a new emission restriction plan needs to be formulated. The analysis and prediction feedback model is then used to evaluate the new emission restriction plan again until the evaluation is qualified. If the pressure relief value is abnormal, an intelligent feedback mechanism automatically provides feedback on the expected adjustment range of the pressure relief value. After adjusting the pressure relief value to the expected range, the evaluation is performed again. If the second evaluation result is qualified, the current round of evaluation is completed. If the second evaluation result is unqualified, manual readjustment or a new emission restriction plan needs to be formulated. The analysis and prediction feedback model is then used to evaluate again until the evaluation is qualified. After the evaluation is qualified, the current round of sample data is used as training data to update the initial sample dataset for training the analysis and prediction feedback model.

[0105] Please refer to the following: Figure 2This is a schematic diagram of the construction method for limited waterproofing and drainage of submarine tunnels provided in the embodiments of this application. Further, a construction method for limited waterproofing and drainage of submarine tunnels may specifically include the contents described in steps S81-S85.

[0106] S81: Based on the qualified limited-quantity waterproofing and drainage simulation construction plan, a waterproof layer is laid between the initial support layer and the secondary lining layer of the submarine tunnel.

[0107] S82: Based on the qualified limited-quantity waterproofing and drainage simulation construction plan, drainage strips and waterstops are installed in the submarine tunnel;

[0108] S83: Based on the qualified limited-scale waterproofing and drainage simulation construction plan, a sidewall water storage tank, a water diversion channel and a central drainage ditch shall be set in the submarine tunnel.

[0109] S84: Based on the qualified limited-quantity drainage simulation construction plan, circumferential drainage pipes and longitudinal drainage pipes shall be installed in the submarine tunnel;

[0110] S85: Install a pressure relief valve at the outlet of the longitudinal drainage pipe, and adjust the pressure relief value of each pressure relief valve according to the expected adjustment range of the pressure relief value fed back, and the construction is completed.

[0111] In one possible implementation, the waterproof layer is an EVA waterproof membrane and a non-woven geotextile, with the non-woven geotextile laid between the initial support layer and the EVA waterproof membrane, and the EVA waterproof membrane disposed between the non-woven geotextile and the secondary lining layer.

[0112] The waterstop is a rubber waterstop installed at the deformation joints and circumferential construction joints of the submarine tunnel;

[0113] The circumferential drainage pipe is installed on the back of the arch wall, and the longitudinal drainage pipe is installed on the back of the side wall. The circumferential drainage pipe and the longitudinal drainage pipe are connected, and the outlet of the longitudinal drainage pipe faces the water storage tank of the side wall.

[0114] The sidewall water storage tanks are multiple sets of water storage tanks installed on the left and right sidewalls of the tunnel.

[0115] The side wall water storage tank is connected to the water intake channel and the central drainage ditch.

[0116] In this embodiment, multiple non-woven geotextiles are laid on the surface of the initial support layer. After gradually extending downwards, an EVA waterproof membrane is laid on the surface of the non-woven geotextiles. Special sealing materials are used to reinforce the tightness of the joints and corners between the surfaces of the non-woven geotextiles, the joints and corners between the non-woven geotextiles and the EVA waterproof membrane, the joints and corners between the non-woven geotextiles and the initial support layer, and the joints and corners between the EVA waterproof membrane and the secondary lining layer. For airtightness, special adhesive is used to seal the drainage strips and waterstops at the expansion joints and construction joints to ensure their airtightness and prevent large amounts of water leakage due to displacement. The circumferential drainage pipes and longitudinal drainage pipes are connected by a T-junction to ensure that the water in the circumferential drainage pipes is led to the longitudinal drainage pipes in the side wall. The water flow rate is controlled by a pressure relief valve to discharge the water in the longitudinal drainage pipes to the reservoir. The central drainage ditch is buried deep in the tunnel floor and water from the reservoir is transported to the central drainage ditch through a water diversion channel.

[0117] Please refer to the following: Figure 3 This is a schematic diagram of the limited drainage monitoring method for submarine tunnels provided in the embodiments of this application. Further, a monitoring method for limited drainage in submarine tunnels may specifically include the contents described in steps S101-S105.

[0118] S101: Multiple sets of environmental monitoring sensors are sequentially installed in sections at the arch waist and sidewalls on both sides of the submarine tunnel.

[0119] S102: Multiple sets of monitoring cameras are installed in sections at the arch of the undersea tunnel;

[0120] S103: Multiple sets of groundwater level monitors are installed sequentially at the arch foot and arch waist of the submarine tunnel;

[0121] S104: Every 25 minutes, the monitoring data collected in real time is assigned to the analysis, prediction and feedback model for evaluation;

[0122] S105: If the assessment fails, the pressure relief valve and / or construction plan at the abnormal location shall be adjusted until the assessment passes.

[0123] In this embodiment, the environmental monitoring sensor includes a stress-strain sensor, a temperature sensor, and a humidity sensor. The environmental monitoring sensor, groundwater level monitor, and surveillance camera are all equipped with communication modules. Through the surveillance camera and AI image analysis technology, abnormal conditions in the tunnel structure are automatically identified. Specifically, the surveillance camera continuously captures images inside the tunnel, and a filtering algorithm is used to reduce noise in the images of each area. An object detection algorithm is used to filter out image objects containing cracks. Histogram equalization technology is used to make the crack information deformation and surrounding rock deformation in the image more obvious. Feature points are extracted from the crack areas in the image, and then the feature points of the image are further normalized. The boundary points of the feature points are processed by an edge detection algorithm, thereby obtaining the length and width data of the tunnel cracks. The real-time monitoring data is transmitted to the analysis and prediction feedback model for evaluation every 25 minutes.

[0124] In this embodiment, a timed push technology is used to periodically push the monitoring data collected by the monitoring device to the analysis, prediction, and feedback model. The analysis, prediction, and feedback model evaluates the monitoring data and provides feedback. If the evaluation is unsatisfactory, an alarm is triggered. When the alarm indicates an abnormal pressure relief value, monitoring personnel go to the abnormal location in the undersea tunnel according to the feedback prompt to manually observe and adjust the pressure relief valve until the evaluation is satisfactory. If the pressure relief valve is manually adjusted but the pressure relief value remains abnormal, the inspection personnel need to combine the monitoring data from the water level monitor, surveillance camera, and environmental monitoring sensor with manual observation to determine whether it is due to... If the drainage problem is caused by pipe rupture and / or deformation of the surrounding rock lining, the drainage restriction construction plan at the abnormal location needs to be manually optimized and adjusted, or the surrounding rock and / or ruptured pipe at the drainage location needs to be repaired. Then, an assessment should be conducted. If the assessment is satisfactory, the construction adjustment and optimization of the abnormal location in the tunnel should be carried out. Otherwise, if it is determined that the annular drainage pipe and / or longitudinal drainage pipe are blocked, the drainage pipe needs to be manually cleared and the drainage value of the pressure relief valve adjusted to the expected standard range. When the alarm message indicates that there is no abnormal pressure relief value, the drainage restriction plan at the abnormal location in the tunnel needs to be revised or adjusted, and an assessment should be conducted. If the assessment is satisfactory, construction rectification should be carried out in the tunnel.

[0125] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A limited waterproofing and drainage assessment method for submarine tunnels, characterized in that, Includes the following steps: S1: Determine the emission limit scheme based on the characteristics of the undersea tunnel; S2: Simulate construction according to the limited emission scheme, and use environmental data, water level data, and tunnel structure data under different environmental conditions as monitoring data; S3: Collect initial sample data and establish an analysis and prediction feedback model, and train the analysis and prediction feedback model using the initial sample dataset; S4: Assign the monitoring data collected every 25 minutes to the analysis and prediction feedback model; S5: Clean the monitoring data to remove outliers and obtain complete monitoring data; S6: Standardize and normalize the complete monitoring data to obtain accurate monitoring data; S7: Using the loss function The precise monitoring data is then optimized to obtain the sample data for this round, wherein... Indicates the amount of sample data. For monitoring data, S8: Assign the sample data of this round to the analysis and prediction feedback model for analysis and evaluation; S9: Adjust the pressure relief valve and / or construction plan according to the evaluation results until the evaluation meets the standards; In step S1, the characteristics of the subsea tunnel include tunnel drainage volume, lining permeability coefficient, lining deformation coefficient, water flow velocity, surrounding rock stress, elastic modulus, Poisson's ratio, water pressure, hydraulic head, hydraulic gradient, historical hydrological data, and regional climate data. A three-dimensional finite element model containing drainage volume and water pressure is constructed based on the characteristics of the subsea tunnel. Simulation calculations are performed using the three-dimensional finite element model to determine a limited discharge scheme. Steps S8 and S9 include: if the evaluation result is unqualified and the unqualified influencing factor involves the pressure relief value, an alarm is triggered, and the expected adjustment range of the pressure relief value is automatically fed back through the intelligent feedback mechanism of the analysis and prediction feedback model; according to the expected adjustment range of the pressure relief value, the pressure relief value is adjusted to the expected adjustment range and re-evaluated until the pressure relief value is adjusted to be qualified; if the evaluation result is unqualified and the unqualified influencing factor does not involve the pressure relief value, an alarm is triggered, and a new emission restriction scheme is formulated and evaluated until the emission restriction scheme is qualified; after the evaluation is qualified, the sample data of this round is used as training data and updated to the initial sample dataset.

2. The limited waterproofing and drainage assessment method for submarine tunnels according to claim 1, characterized in that, In step S2, the environmental data includes multiple sets of temperature, humidity, and air pressure data collected by environmental monitoring sensors within the tunnel; the water level data is used to simulate multiple sets of groundwater level data collected by groundwater level monitors within the tunnel; and the tunnel structure data is used to simulate multiple sets of tunnel crack information data collected by surveillance cameras within the tunnel.

3. The limited waterproofing and drainage assessment method for submarine tunnels according to claim 1, characterized in that, Step S3 includes: the initial sample dataset is a large number of historical sample datasets under simulated construction conditions; the initial sample dataset is divided into n initial sample data subsets according to feature attributes, and each initial sample data subset is trained for n+1 iterations; the learning efficiency of the initial sample data subsets in different groups is adjusted through a dynamic learning rate adjustment mechanism; strong correlation is established between the initial sample data subsets of different monitoring factors through a cross-correlation mechanism; and the prediction evaluation results of the analysis, prediction and feedback model are verified through a cross-validation mechanism until the verification results meet the prediction criteria.

4. The limited waterproofing and drainage assessment method for submarine tunnels according to claim 1, characterized in that, Step S5 includes using the Z-score method and machine learning model to automatically identify and filter abnormal data.

5. The limited waterproofing and drainage assessment method for submarine tunnels according to claim 4, characterized in that, The Z-score method is used to initially identify abnormal monitoring values, and the machine learning model uses the ClusteringAlgorithms algorithm to further screen and filter the abnormal monitoring values.

6. A construction method for limited drainage and waterproofing in submarine tunnels, based on the implementation of the limited drainage and waterproofing assessment method according to any one of claims 1 to 5, characterized in that, The process includes the following steps: S81: Based on the approved limited-quantity waterproofing and drainage simulation construction plan, a waterproof layer is laid between the initial support layer and the secondary lining layer of the submarine tunnel; S82: Based on the approved limited-quantity waterproofing and drainage simulation construction plan, drainage strips and waterstops are installed in the submarine tunnel; S83: Based on the approved limited-quantity waterproofing and drainage simulation construction plan, sidewall water storage tanks, water diversion channels, and central drainage ditches are installed in the submarine tunnel; S84: Based on the approved limited-quantity waterproofing and drainage simulation construction plan, circumferential drainage pipes and longitudinal drainage pipes are installed in the submarine tunnel; S85: Pressure relief valves are installed at the outlets of the longitudinal drainage pipes, and the pressure relief values ​​of each pressure relief valve are adjusted according to the expected adjustment range of the pressure relief value, thus completing the construction.

7. A construction method for limited drainage and waterproofing in submarine tunnels according to claim 6, characterized in that, include: The waterproof layer consists of an EVA waterproof membrane and a non-woven geotextile. The non-woven geotextile is laid between the initial support layer and the EVA waterproof membrane, and the EVA waterproof membrane is positioned between the non-woven geotextile and the secondary lining layer. The waterstop is a rubber waterstop installed at the expansion joints and circumferential construction joints of the submarine tunnel. The circumferential drainage pipe is installed on the back of the arch wall, and the longitudinal drainage pipe is installed on the back of the side wall. The circumferential drainage pipe and the longitudinal drainage pipe are connected, and the outlet of the longitudinal drainage pipe faces the side wall water storage tank. The side wall water storage tank consists of multiple sets of water storage tanks installed on the left and right side walls of the tunnel. The side wall water storage tank is connected to the central drainage ditch through the water diversion channel.

8. A monitoring method for limited drainage and waterproofing in submarine tunnels, based on the implementation of the limited drainage and waterproofing assessment method according to any one of claims 1 to 5, characterized in that, The process includes the following steps: S101: Installing multiple sets of environmental monitoring sensors in sections along the arch waist and sidewalls on both sides of the undersea tunnel; S102: Installing multiple sets of monitoring cameras in sections along the arch top of the undersea tunnel; S103: Installing multiple sets of groundwater level monitors in sections along the arch foot and arch waist of the undersea tunnel; S104: Assigning the monitoring data collected in real time to the analysis, prediction, and feedback model for evaluation every 25 minutes; S105: If the evaluation is unsatisfactory, adjusting the pressure relief valve and / or construction plan at the abnormal location until the evaluation is satisfactory.

Images