Limited drainage prevention evaluation construction and monitoring method for subsea tunnel

Abstract

The invention discloses a limited drainage evaluation construction and monitoring method for a subsea tunnel, and the evaluation method comprises the steps: determining a limited drainage scheme based on the characteristics of the subsea tunnel, simulating the construction according to the limited drainage scheme, and obtaining the monitoring data in different environments; collecting initial sample data, establishing an analysis and prediction feedback model, and training the analysis and prediction feedback model by using the initial sample data set; assigning the monitoring data to the analysis 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 sample data of the round; assigning the current round of sample data to the analysis prediction feedback model for analysis and evaluation; and adjusting the pressure release valve and / or the construction scheme according to the evaluation result until the evaluation is qualified. And according to the qualified drainage limiting scheme, carrying out drainage limiting construction and monitoring on the subsea tunnel.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of tunnel construction, in particular to a limited dewatering evaluation construction and monitoring method for a submarine tunnel. BACKGROUND

[0002] The construction environment of a submarine tunnel is complex, surrounded by high-pressure seawater, and the permeability of soil and rock may cause seawater to continuously infiltrate, increasing the difficulty of construction and potential safety risks; submarine tunnel water seepage not only affects construction progress, but also may cause structural deformation, corrosion and other safety hazards, and even may affect the long-term service life of the tunnel. Due to the harsh and complex environmental conditions of a submarine tunnel, the construction scheme is complex, which increases the difficulty of subsequent construction scheme evaluation. In the past, construction experts often evaluated the feasibility of the construction scheme. However, human evaluation is often limited by individual knowledge and experience levels, and it is difficult to consider complex multiple influencing factors in the evaluation process, which may easily overlook important influencing factors, resulting in insufficient and objective evaluation of the construction scheme, and thus potential risks in the construction process. In addition, during the manual evaluation process, the process is tedious, time-consuming and inefficient, which may easily affect project progress and decision-making efficiency. Furthermore, after the completion and use of a submarine tunnel, there is still a probability of water gushing and other sudden conditions, which may affect the safe use of the submarine tunnel. However, due to the complex and variable environment, manual monitoring is not effective. First, manual monitoring is labor-intensive and cannot cover everything. Even if the monitoring personnel detect a sudden condition, due to the reaction time required for manual operation, they cannot respond to the sudden condition in a timely and effective manner, resulting in irreversible losses. SUMMARY

[0003] In order to at least overcome the above-mentioned deficiencies in the prior art, the purpose of the present application is to provide a limited dewatering evaluation construction and monitoring method for a submarine tunnel.

[0004] The present application provides a limited dewatering evaluation method for a submarine tunnel, comprising: S1: determining a limited discharge scheme based on the characteristics of the submarine tunnel; S2: simulating construction according to the limited discharge scheme, and taking environmental data, water level data and tunnel structure data under different environmental conditions as monitoring data; S3: collecting initial sample data and establishing an analysis and prediction feedback model, and training the analysis and prediction feedback model using the initial sample data set; S4: assigning the monitoring data collected every 25 minutes to the analysis and prediction feedback model; S5: performing data cleaning on the monitoring data to remove outliers, and obtaining complete monitoring data; S6: standardize and normalize the complete monitoring data to obtain accurate monitoring data; S7: using a loss function , the accurate monitoring data is optimized to obtain the sample data of this round, wherein, represents the sample data amount, is the monitoring data, is the simulation prediction data corresponding to the monitoring data; S8: assign the sample data of this round to the analysis prediction feedback model for analysis and evaluation; S9: adjust the pressure relief valve and / or the construction scheme according to the evaluation result until the evaluation meets the standard.

[0005] Traditional construction scheme evaluation often needs to rely on experienced experts and construction personnel in the construction industry to evaluate the feasibility of the scheme, however, this method often has subjectivity and limitations, in the process of human evaluation, due to the complex geographical environment of the submarine tunnel, multiple factors need to be considered, which increases the difficulty of evaluation, so that the evaluation process takes a long time, and based on complex multiple influencing factors, important influencing factors are easily ignored in the evaluation process, which leads to potential risks in the construction process; The purpose of the present application is to abandon the human evaluation method and use a prediction model to evaluate the feasibility of the limited discharge construction scheme.

[0006] In the embodiments of the present application, the construction personnel formulates a preliminary limited discharge scheme according to the submarine tunnel discharge, lining permeability coefficient, lining deformation coefficient, water flow rate, surrounding rock stress, elastic modulus, Poisson's ratio, water pressure, water head height, hydraulic slope, historical hydrological data and regional climate and other parameters, establishes a three-dimensional finite element model, divides the finite element model with appropriate grid density, assigns the above parameters to the finite element model to determine the boundary conditions and initial conditions of the tunnel contour, simulates the construction of the limited discharge scheme in the finite element model, uses the Navier-Stokes equation to describe the flow behavior of fluid in the tunnel under different conditions, starts the simulation calculation program, and obtains the water flow distribution, pressure change and stress state under different conditions through iterative calculation; according to the simulation result, the accurate limited discharge scheme is determined.

[0007] The environmental data, water level data and tunnel structure data under different environmental conditions are uploaded to the analysis and prediction feedback model as monitoring data, to simulate the arrangement of environmental monitoring sensors, underground water level monitors and monitoring cameras in the tunnel, to simulate the real-time collection of monitoring data such as temperature data, humidity data and air pressure data, underground water level data and tunnel crack information data in the tunnel model by the above devices, and to transmit the monitoring data to the analysis and prediction feedback model every 25 minutes. After a series of screening and optimization processing in the analysis and prediction feedback model, the current tunnel model is predicted and evaluated, and the pressure relief valve and / or the emission control scheme are adjusted accordingly until the evaluation is qualified.

[0008] In the embodiments of the present application, by collecting complex and large amounts of monitoring data and performing data cleaning, standardization and normalization preprocessing steps, the data processing and analysis efficiency is improved, and possible human errors or omissions are avoided.

[0009] During the training process, the analysis and prediction feedback model can learn the complex relationship between the data, and obtain more accurate results by simulating and predicting the water level data. Compared with manual evaluation, the analysis and prediction feedback model can more comprehensively and systematically consider various influencing factors, improve the accuracy of prediction, and thus can more quickly feedback the evaluation results.

[0010] The analysis and prediction feedback model in the evaluation method of the embodiments of the present application can realize real-time dynamic adjustment and optimization, iteratively optimize the emission scheme and perform evaluation according to the continuously updated monitoring data feedback. This real-time dynamics makes the emission control scheme more adaptive and flexible, and better adapts to the complex and variable environment of the submarine tunnel.

[0011] In a possible implementation, in step S1, the characteristics of the submarine tunnel include tunnel discharge, lining permeability coefficient, lining deformation coefficient, water flow rate, surrounding rock stress, elastic modulus, Poisson's ratio, water pressure, water head height, hydraulic slope, historical hydrological data and regional climate data. A three-dimensional finite element model containing discharge and water pressure is constructed according to the characteristics of the submarine tunnel. The three-dimensional finite element model is used for simulation calculation to determine the limited emission scheme.

[0012] In a possible implementation, in step S2, the environmental data includes multiple sets of temperature data, humidity data and air pressure data collected by the environmental monitoring sensors to simulate the environment. The water level data are used to simulate the multiple sets of underground water level data collected by the underground water level monitors in the tunnel. The tunnel structure data are used to simulate the multiple sets of tunnel crack information data collected by the monitoring cameras in the tunnel.

[0013] In a possible implementation manner, the step S3 comprises: The initial sample data set is a large number of historical sample data sets under simulated construction conditions; The initial sample data set is divided into n initial sample data subsets according to characteristic attributes, and n+1 iteration training is performed on each initial sample data subset; The learning efficiency of different groups of initial sample data subsets is adjusted through a dynamic learning rate adjustment mechanism; The initial sample data subsets of different monitoring factors are strongly associated through a cross-association mechanism; The prediction evaluation result of the analysis and prediction feedback model is verified through a cross-validation mechanism until the verification result meets the prediction standard.

[0014] In a possible implementation manner, the step S5 comprises automatically identifying and filtering abnormal data by using a Z-score method and a machine learning model.

[0015] In a possible implementation manner, the abnormal monitoring values are preliminarily identified by using a Z-score method, and the machine learning model uses a ClusteringAlgorithms algorithm to further screen and filter the abnormal monitoring values.

[0016] In a possible implementation manner, the steps S8 and S9 comprise: If the evaluation result is unqualified and the unqualified influencing factor involves the pressure relief value, an alarm prompt 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 be within 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 prompt is triggered, and a new emission limiting scheme is developed for evaluation until the emission limiting scheme is qualified; After the evaluation is qualified, the sample data of this round is updated to the initial sample data set as training data.

[0017] In a possible implementation manner, a construction method for a submarine tunnel limited drainage comprises the following steps: S81: According to the qualified limited drainage simulation construction scheme, a waterproof layer is arranged between the primary support layer and the secondary lining layer of the submarine tunnel; S82: According to the qualified limited drainage simulation construction scheme, a drainage zone and a water stop zone are arranged in the submarine tunnel; S83: According to the qualified limited drainage simulation construction scheme, a side wall water storage pool, a water diversion channel and a central drainage ditch are arranged in the submarine tunnel; S84: According to the qualified limited drainage simulation construction scheme, the circumferential drainage pipe and the longitudinal drainage pipe are arranged in the submarine tunnel; S85: The pressure relief valve is arranged at the water outlet of the longitudinal drainage pipe, and the pressure relief value of each pressure relief valve is adjusted according to the expected adjustment range of the feedback pressure relief value, and the construction is completed.

[0018] In the implementation of the embodiment, a plurality of waterproof layers are laid on the surface of the primary support layer, which is gradually expanded downward, special sealing materials are used to reinforce the tightness of the joints and corners between the waterproof layers, the tightness of the joints and corners between the waterproof layer and the primary support layer, and the tightness of the joints and corners between the waterproof layer and the secondary lining layer, special glue is used to seal the drainage belt and the water stop belt at the deformation joint and the construction joint, the airtightness of the drainage belt and the water stop belt is ensured, and a large amount of water leakage caused by displacement is prevented; the circumferential drainage pipe and the longitudinal drainage pipe are connected through the tee pipe, the water in the circumferential drainage pipe is guided to the longitudinal drainage pipe in the side wall, the water in the longitudinal drainage pipe is discharged to the water storage pool through the pressure relief valve, the central drainage channel is buried deep in the tunnel floor, and the water in the water storage pool is transported to the central drainage channel through the water conduit for discharge.

[0019] In one possible implementation, a monitoring method for limited drainage of a submarine tunnel includes the following steps: S101: A plurality of groups of environmental monitoring sensors are sequentially arranged at the haunches and side walls on the left and right sides of the submarine tunnel; S102: A plurality of groups of monitoring cameras are sequentially arranged at the vaults of the submarine tunnel; S103: A plurality of groups of underground water level monitors are sequentially arranged at the soffits and haunches of the submarine tunnel; S104: The monitoring data collected by real-time monitoring is assigned to an analysis and prediction feedback model for evaluation every 25 minutes; S105: If the evaluation is not qualified, the pressure relief valve and / or the construction scheme at the abnormal position are adjusted until the evaluation is qualified.

[0020] In the embodiments of the present application, the environmental monitoring sensor includes a stress-strain sensor, a temperature sensor and a humidity sensor, the environmental monitoring sensor, the underground water level monitor and the monitoring camera are all equipped with a communication module, through the monitoring camera and the AI image analysis technology, the tunnel structure abnormal situation is automatically identified; wherein, the images inside the tunnel are continuously captured through the monitoring camera, the noise of the images in each region is processed through a filtering algorithm, the image objects containing cracks are screened out through an object detection algorithm, the crack information in the image is deformed and the surrounding rock deformation is more obvious through a histogram equalization technology, the feature points in the crack region of the image are extracted, and the feature points of the image are further normalized, the boundary points of the feature points are processed through an edge detection algorithm, and then the length and width data of the tunnel cracks are obtained, and the real-time monitoring data is transmitted to the analysis prediction feedback model for evaluation every 25 minutes.

[0021] Compared with the prior art, the present application has the following beneficial effects: (1) The present application collects complex and large amount of monitoring data, and performs data cleaning, standardization, normalization and other preprocessing steps, thereby improving the data processing and analysis efficiency and avoiding possible human errors or omissions.

[0022] (2) The analysis prediction feedback model of the present application can learn the complex relationship between data, simulate and predict the water level data, and obtain more accurate results. Compared with manual evaluation, the analysis prediction feedback model can more comprehensively and systematically consider various influencing factors, improve the prediction accuracy, and thus can more quickly feedback the evaluation results.

[0023] (3) The analysis prediction feedback model in the evaluation method of the present application can realize real-time dynamic adjustment and optimization, iteratively optimize the discharge scheme and perform evaluation according to the continuously updated monitoring data feedback. This real-time dynamics makes the evaluation of the emission scheme more adaptive and flexible, and better adapts to the complex and variable environment of the submarine tunnel.

[0024] (4) After the construction is completed, the submarine tunnel is monitored in real time by laying monitoring devices, the monitoring devices transmit the environmental data, hydrological data and tunnel structure data to the analysis prediction feedback model for evaluation through the communication module, and determine whether the water discharge of the submarine tunnel exceeds the preset standard range. This greatly reduces the proportion of manual operation in the entire monitoring measurement, greatly eliminates the influence of human intervention on the data, ensures the authenticity of the monitoring data, and thus quickly feeds back the monitoring results to play a role of twice the result with half the effort for the subsequent tunnel maintenance work. BRIEF DESCRIPTION OF DRAWINGS

[0025] The accompanying drawings used to provide further understanding of the embodiments of the present application, constitute a part of the present application, and do not constitute a limitation of the embodiments of the present application. In the drawings: Figure 1 Flow chart of the method for evaluating the limited waterproofing of the seabed tunnel according to the present application; Figure 2 Flow chart of the method for constructing the limited waterproofing of the seabed tunnel according to the present application; Figure 3 Flow chart of the method for monitoring the limited waterproofing of the seabed tunnel according to the present application. DETAILED DESCRIPTION

[0026] In order to make the objects, technical solutions and advantages of the present application clearer, further detailed description will be made to the present application in combination with embodiments and drawings, and the schematic embodiments and their descriptions are only used to explain the present application, and not as a limitation to the present application.

[0027] In the description of the present application, it should be understood that the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise" and the like indicate the orientation or positional relationship shown in the drawings, and are only used to facilitate the description of the present application and simplify the description, and therefore cannot be understood as indicating or implying that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation to the present application.

[0028] In addition, the terms "first" and "second" are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features referred to. Therefore, the features defined with "first" and "second" can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "multiple" is two or more, unless otherwise specifically limited.

[0029] In the present application, unless otherwise specifically defined and limited, the terms "mounting", "connection", "connection", "fixing" and the like should be understood broadly, for example, it can be fixedly connected, or it can be detachably connected, or it can be integrally connected; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium; it can be the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0030] In the present application, unless otherwise explicitly specified and limited, the first feature is "on" or "under" the second feature can include that the first and second features are in direct contact, or that the first and second features are not in direct contact but are in contact through another feature between them. Moreover, the first feature "on", "above" and "over" the second feature includes that the first feature is directly above and obliquely above the second feature, or only indicates that the first feature is higher in horizontal height than the second feature. The first feature "under", "below" and "under" the second feature includes that the first feature is directly below and obliquely below the second feature, or only indicates that the first feature is lower in horizontal height than the second feature.

[0031] In addition, the technical solutions among various embodiments can be combined with each other, but it must be based on that a person skilled in the art can realize, when the combination of technical solutions appears contradictory or unachievable, it should be considered that the combination of technical solutions does not exist, and is not within the scope disclosed by the present application.

[0032] In the subsequent description, suffixes such as "module", "component", "assembly" or "unit" are used only for the convenience of description of the present application, and have no specific meaning in themselves. Therefore, they can be mixedly used.

[0033] The present application will be further described in detail through specific embodiments in conjunction with the accompanying drawings.

[0034] Please refer to Figure 1 The flowchart of the seabed tunnel limited drainage evaluation method provided by the embodiments of the present application further includes the following steps S1-S9.

[0035] S1: determining a limited drainage scheme based on the characteristics of the seabed tunnel; S2: simulating construction according to the limited drainage scheme, taking environmental data, water level data and tunnel structure data under different environmental conditions as monitoring data; S3: collecting initial sample data and establishing an analysis and prediction feedback model, and training the analysis and prediction feedback model using the initial sample data set; S4: assigning the monitoring data collected every 25 minutes to the analysis and prediction feedback model; S5: performing data cleaning to remove outliers from the monitoring data to obtain complete monitoring data; S6: performing standardization and normalization processing on the complete monitoring data to obtain accurate monitoring data; S7: using a loss function performing data optimization on the accurate monitoring data to obtain the current sample data, wherein, represents the sample data amount, monitoring data, simulated prediction data corresponding to the monitoring data; S8: Assign the sample data of this round to the analysis prediction feedback model for analysis and evaluation; S9: Adjust the pressure relief valve and / or construction scheme according to the evaluation result until the evaluation meets the standard.

[0036] In the embodiment of the application, the environmental data, water level data and tunnel structure data are packaged and integrated into a unified data set, and the data is uploaded in time synchronization using the timestamp method, so that the monitoring data of different influencing factors can be compared and analyzed at the same time point; in the embodiment, the loss function The accurate monitoring data is optimized to obtain the sample data of this round, wherein, represents the sample data amount, monitoring data, simulated prediction data corresponding to the monitoring data, which is an existing algorithm, and the specific calculation logic is not described here; in the embodiment, the n-fold strategy is adopted, the entire initial sample data set is divided into n parts, n-1 parts are used for model training, and the remaining 1 part is used for test verification, and the cycle is repeated n+1 times. This method helps to prevent the overfitting phenomenon of the model; in the evaluation stage, the evaluation result of the monitoring sample data of this round is obtained by evaluation calculation, and a condition triggering mechanism is set to compare the evaluation result with the evaluation result of the historical sample data under similar or same environmental conditions. If the evaluation result deviates significantly from the historical trend, an alarm prompt is triggered.

[0037] In one possible implementation, in the step S1, the characteristics of the submarine tunnel include tunnel discharge, lining permeability coefficient, lining deformation coefficient, water flow rate, surrounding rock stress, elastic modulus, Poisson's ratio, water pressure, water head height, hydraulic slope, historical hydrological data and regional climate data. A three-dimensional finite element model containing discharge and water pressure is constructed according to the characteristics of the submarine tunnel. Through the three-dimensional finite element model, a simulation calculation is performed to determine the limited discharge scheme.

[0038] In the embodiment of the application, the construction survey personnel obtain the parameters of the submarine tunnel construction section such as discharge, lining permeability coefficient, lining deformation coefficient, water flow rate, surrounding rock stress, elastic modulus, Poisson's ratio, water pressure, water head height, hydraulic slope, historical hydrological data and regional climate data through field measurement and construction standards, and then establish a three-dimensional finite element model after formulating a preliminary limited discharge scheme.

[0039] In the implementation of the embodiment, the three-dimensional finite element model is meshed with appropriate mesh density, the parameters are assigned to the finite element model to determine the boundary conditions and initial conditions of the tunnel profile, the limited discharge scheme is simulated in the finite element model, the Navier-Stokes equation is used to describe the flow behavior of fluid 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; and the accurate limited discharge scheme is determined according to the simulation results.

[0040] In a possible implementation, in the step S2, the environmental data includes a plurality of groups of temperature data, humidity data and air pressure data in the tunnel collected by the environmental monitoring sensor. The water level data are used to simulate a plurality of groups of underground water level data in the tunnel collected by the underground water level monitor. The tunnel structure data are used to simulate a plurality of groups of tunnel crack information data in the tunnel collected by the monitoring camera.

[0041] In the embodiment, the environmental data, the water level data and the tunnel structure data under different environmental conditions are taken as the monitoring data and uploaded to the analysis and prediction feedback model to simulate the environmental monitoring sensor, the underground water level monitor and the monitoring camera arranged in sections in the tunnel, simulate the temperature data, the humidity data and the air pressure data, the underground water level data and the tunnel crack information data and other monitoring data collected by the above devices in the tunnel model in real time, deliver the monitoring data to the analysis and prediction feedback model every 25 minutes, and make a prediction evaluation on the limited discharge scheme of the current tunnel model after a series of screening and optimization processing in the analysis and prediction feedback model.

[0042] In a possible implementation, the step S3 includes: The initial sample data set is a historical sample data set under a large number of simulated construction conditions; The initial sample data set is divided into n initial sample data subsets according to characteristic attributes, and each initial sample data subset is iteratively trained n+1 times; The learning efficiency of different groups of initial sample data subsets is adjusted through a dynamic learning rate adjustment mechanism; The initial sample data subsets of different monitoring factors are strongly associated through a cross-correlation mechanism; The prediction evaluation result of the analysis and prediction feedback model is verified through a cross-validation mechanism until the verification result meets the prediction standard.

[0043] In the embodiment of the present application, the analysis prediction feedback model is trained by a large number of historical sample data, which can improve the generalization ability of the analysis prediction feedback model, thereby better adapting to the evaluation and prediction ability under different construction conditions and hydrogeological conditions; the n-fold strategy is adopted to divide the entire initial sample data set into n parts, n-1 parts are used for model training, and the remaining 1 part is used for test verification, and the cycle is repeated n+1 times, which helps to prevent the overfitting phenomenon of the model; the learning efficiency of the analysis prediction feedback model is adjusted in real time through the dynamic learning rate adjustment mechanism, which helps to speed up the learning convergence speed and avoid falling into a local optimal solution; through the cross-correlation mechanism, the analysis prediction feedback model can comprehensively consider the influence relationship between different factors on the limited discharge scheme and the tunnel structure, thereby improving the prediction ability of the model; through cross-validation mechanism, the stability and robustness of the analysis prediction feedback model can be enhanced; through a large number of analysis prediction training, the analysis prediction feedback model outputs the environmental expectation value, the environmental threshold value, the water level expectation value, the water level threshold value, the tunnel structure expectation value, the tunnel structure threshold value, the pressure relief expectation value and the pressure relief threshold value under different environmental conditions in the tunnel, when the absolute value of the difference between the input data and the corresponding expectation value data is greater than the corresponding threshold value, the evaluation result is unqualified, the unqualified input data is uploaded to the display device as an influence factor and an alarm prompt is triggered; in the embodiment, the environmental expectation value is the temperature expectation value, the humidity expectation value and the air pressure expectation value, the environmental threshold value is the temperature threshold value, the humidity threshold value and the air pressure threshold value, the tunnel structure expectation value is the tunnel crack length expectation value and the tunnel crack width expectation value, and the tunnel structure threshold value is the tunnel crack length threshold value and the tunnel crack width threshold value.

[0044] In a possible implementation, the step S5 comprises using a Z-score method and a machine learning model to automatically identify and filter abnormal data.

[0045] In the embodiment of the present application, the Z-score method is used to preliminarily identify abnormal monitoring values and null values in the monitoring data, and after the null values are removed, the machine learning model uses the ClusteringAlgorithms algorithm to further screen and filter the abnormal monitoring values. Since the monitoring data is uploaded every 25 minutes, a large amount of monitoring data has been accumulated in this process. Even if there are abnormal values and null values in the data, the completeness of the entire monitoring data set will not be affected after the abnormal values and null values are removed, thereby the evaluation result of the analysis prediction feedback model will not be affected.

[0046] In a possible implementation, the steps S8 and S9 comprise: If the evaluation result is unqualified and the unqualified influencing factor involves the pressure relief value, an alarm prompt is triggered, and the expected adjustment range of the pressure relief value is automatically fed back through the intelligent feedback mechanism of the analysis 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 the qualified evaluation result; If the evaluation result is unqualified and the unqualified influencing factor does not involve the pressure relief value, an alarm prompt is triggered, and a new limited drainage scheme is formulated for evaluation until the limited drainage scheme is evaluated to be qualified; After the evaluation is qualified, the sample data of the current round is updated to the initial sample data set as training data.

[0047] In the evaluation stage, the evaluation result of the current round of monitoring is obtained according to the evaluation calculation of the current round of monitoring sample data. If the evaluation result is unqualified, an alarm prompt is triggered, and the existing abnormality detection algorithm is used to determine whether the pressure relief value is abnormal. When the pressure relief value is normal, the limited drainage scheme needs to be manually adjusted or formulated, and the analysis prediction feedback model is used to evaluate the new limited drainage scheme again until the evaluation is qualified. When the pressure relief value is abnormal, the expected adjustment range of the pressure relief value is automatically fed back through the intelligent feedback mechanism, the pressure relief value is adjusted to the expected range, and the evaluation is performed again. If the evaluation result is qualified again, the current round of evaluation is completed. If the evaluation result is unqualified again, the limited drainage scheme needs to be manually adjusted or formulated, and the analysis prediction feedback model is used to evaluate again until the evaluation is qualified. After the evaluation is qualified, the sample data of the current round is updated to the initial sample data set as training data to train the analysis prediction feedback model.

[0048] Please refer to Figure 2 The flowchart of the seabed tunnel limited drainage construction method provided by the embodiment of the present application further includes the following steps S81-S85.

[0049] S81: According to the qualified limited drainage simulation construction scheme, a waterproof layer is arranged between the primary support layer and the secondary lining layer of the seabed tunnel; S82: According to the qualified limited drainage simulation construction scheme, a drainage zone and a water stop zone are arranged in the seabed tunnel; S83: According to the qualified limited drainage simulation construction scheme, a side wall water storage pool, a water diversion channel, and a central drainage ditch are arranged in the seabed tunnel; S84: According to the qualified limited drainage simulation construction scheme, a ring-shaped drainage pipe and a longitudinal drainage pipe are arranged in the seabed tunnel; S85: A pressure relief valve is arranged at the water outlet of the longitudinal drainage pipe, and the pressure relief value of each pressure relief valve is adjusted according to the expected adjustment range of the feedback pressure relief value, and the construction is completed.

[0050] In a possible implementation, the waterproof layer is an EVA waterproof plate and a non-woven geotextile, the non-woven geotextile is laid between the primary support layer and the EVA waterproof plate, and the EVA waterproof plate is arranged between the non-woven geotextile and the secondary lining layer; The waterstop is a rubber waterstop arranged at the deformation joint and the circumferential construction joint of the subsea tunnel. The circumferential drainage pipe is arranged at the wall back of the arch wall, the longitudinal drainage pipe is arranged at the wall back of the side wall, the circumferential drainage pipe and the longitudinal drainage pipe are communicated, and the water outlet of the longitudinal drainage pipe faces the side wall reservoir. The side wall reservoir is a plurality of groups of reservoirs arranged at the left and right side walls of the tunnel. The side wall reservoir is communicated with the water diversion channel and the central drainage ditch.

[0051] In the embodiment, a plurality of non-woven geotextiles are laid on the surface of the primary support layer, and then the EVA waterproof plate is laid on the surface of the non-woven geotextile after the non-woven geotextiles are gradually expanded downward, special sealing materials are used to reinforce the tightness of the joints and corners between the surfaces of the non-woven geotextiles, the tightness of the joints and corners between the surfaces of the non-woven geotextiles and the EVA waterproof plate, the tightness of the joints and corners between the non-woven geotextiles and the primary support layer, and the tightness of the joints and corners between the EVA waterproof plate and the secondary lining layer, special glue is used to seal the drainage belts and the waterstops at the deformation joints and the construction joints, the airtightness of the drainage belts and the waterstops is ensured, and a large amount of water leakage caused by displacement is prevented, the circumferential drainage pipe and the longitudinal drainage pipe are communicated through the tee pipe, the water in the circumferential drainage pipe is guided to the longitudinal drainage pipe in the side wall, the amount of water is controlled through the pressure relief valve, the water in the longitudinal drainage pipe is discharged to the reservoir, the central drainage ditch is buried in the deep part of the tunnel floor, and the water in the reservoir is transported to the central drainage ditch through the water diversion channel for discharge.

[0052] Please refer to Figure 3 The flowchart of the subsea tunnel limited drainage monitoring method provided in the embodiment further includes the steps S101-S105.

[0053] S101: A plurality of groups of environmental monitoring sensors are sequentially arranged at the haunches and the side walls on the left and right sides of the subsea tunnel. S102: A plurality of groups of monitoring cameras are sequentially arranged at the vaults of the subsea tunnel. S103: sequentially setting multiple groups of underground water level monitors at the arch springings and arch soles of the submarine tunnel; S104: assigning the monitoring data collected in real time to the analysis and prediction feedback model for evaluation every 25 minutes; S105: if the evaluation is unqualified, adjusting the pressure relief valve and / or the construction scheme at the abnormal position until the evaluation is qualified.

[0054] In the embodiments of the present application, the environmental monitoring sensor includes a stress-strain sensor, a temperature sensor and a humidity sensor, the environmental monitoring sensor, the underground water level monitor and the monitoring camera are all equipped with a communication module, through the monitoring camera and the AI image analysis technology, the tunnel structure abnormality is automatically identified; wherein, the images inside the tunnel are continuously captured through the monitoring camera, the images in each region are denoised using a filtering algorithm, the image objects containing cracks are screened out through an object detection algorithm, the crack information in the image is deformed and the surrounding rock deformation is more obvious through a histogram equalization technology, the feature points in the crack region of the image are extracted, the feature points of the image are further normalized, the boundary points of the feature points are processed through an edge detection algorithm, and then the length and width data of the tunnel cracks are obtained, and the real-time monitoring data are transmitted to the analysis and prediction feedback model for evaluation every 25 minutes.

[0055] When the embodiments of the present application are implemented, the monitoring data collected by the monitoring device are pushed to the analysis and prediction feedback model at regular intervals, the analysis and prediction feedback model evaluates and feeds back according to the monitoring data, if the evaluation is unqualified, an alarm prompt is triggered, when the alarm prompt content is that the pressure relief value is abnormal, the monitoring personnel go to the abnormal position in the submarine tunnel according to the feedback prompt, and manually observe and adjust the pressure relief valve until the evaluation is qualified, if the pressure relief valve is manually adjusted and the pressure relief value is still in an abnormal state, the detection personnel need to judge whether the water leakage problem is caused by pipe body rupture and / or lining surrounding rock deformation according to the monitoring data of the water level monitor, the monitoring camera and the environmental monitoring sensor combined with manual observation, if so, the limited drainage construction scheme at the abnormal position needs to be re-optimized and adjusted or the surrounding rock and / or the ruptured pipeline at the water leakage position needs to be repaired, and then the evaluation is performed, after the evaluation is qualified, the construction adjustment and optimization at the abnormal position of the tunnel are performed; otherwise, it is judged that the annular drainage pipe and / or the longitudinal drainage pipe is blocked, and then the drainage pipe needs to be dredged manually, and the water leakage value of the pressure relief valve is adjusted to be within the expected standard range of the feedback; when the alarm prompt content is not related to the abnormal pressure relief value, the limited drainage scheme at the abnormal position of the tunnel needs to be re-developed or adjusted, and then the evaluation is performed, after the evaluation is qualified, the construction rectification in the tunnel is performed.

[0056] The above detailed description of the specific embodiments of the present application has been given to understand the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above description is only a specific embodiment of the present application and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for limited dewatering assessment of a subsea tunnel, characterized in that, The method comprises the following steps: S1: determining a limited discharge scheme based on the characteristics of the submarine tunnel; S2: simulating construction according to the limited discharge scheme, taking environmental data, water level data, and tunnel structure data under different environmental conditions as monitoring data; S3: collecting initial sample data and establishing an analysis and prediction feedback model, and training the analysis and prediction feedback model using the initial sample data set; S4: assigning the monitoring data collected every 25 minutes to the analysis and prediction feedback model; S5: performing data cleaning to remove abnormal values from the monitoring data to obtain complete monitoring data; S6: standardizing and normalizing the complete monitoring data to obtain accurate monitoring data; S7: using a loss function , wherein the accurate monitoring data is data optimized, and a sample data of this round is obtained, , wherein the sample data is represented by a quantity, , wherein the monitoring data is represented by a quantity, , wherein the monitoring data corresponds to the simulation prediction data. S8: assigning the current sample data to the analysis and prediction feedback model for analysis and evaluation; S9: adjusting the pressure relief valve and / or the construction scheme according to the evaluation results until the evaluation meets the standard.

2. A method for limited dewatering assessment of a subsea tunnel according to claim 1, characterized in that, In the step S1, the characteristics of the submarine tunnel include tunnel discharge, lining permeability coefficient, lining deformation coefficient, water flow rate, surrounding rock stress, elastic modulus, Poisson's ratio, water pressure, water head height, hydraulic slope, historical hydrological data, and regional climate data; A three-dimensional finite element model containing discharge and water pressure is constructed according to the characteristics of the submarine tunnel; The limited discharge scheme is determined by simulating calculation through the three-dimensional finite element model.

3. A method for limited dewatering assessment of a subsea tunnel according to claim 1, characterized in that, In the step S2, the environmental data includes multiple sets of temperature data, humidity data, and air pressure data collected by environmental monitoring sensors to simulate the environment inside the tunnel; The water level data are used to simulate the multiple sets of underground water level data collected by underground water level monitors inside the tunnel; The tunnel structure data are used to simulate the multiple sets of tunnel crack information data collected by monitoring cameras.

4. The method for limited dewatering assessment of a subsea tunnel according to claim 1, wherein, The step S3 comprises: The initial sample data set is a large set of historical sample data sets under simulated construction conditions; The initial sample data set is divided into n initial sample data subsets according to characteristic attributes, and each initial sample data subset is iteratively trained n+1 times; The learning efficiency of different groups of initial sample data subsets is adjusted through a dynamic learning rate adjustment mechanism; The initial sample data subsets of different monitoring factors are strongly associated through a cross-correlation mechanism; The prediction evaluation results of the analysis and prediction feedback model are verified through a cross-validation mechanism until the verification results meet the prediction standard.

5. The method for limited dewatering evaluation of a subsea tunnel according to claim 1, wherein, The step S5 comprises using the Z-score method and a machine learning model to automatically identify and filter abnormal data.

6. A method for limited dewatering assessment of a subsea tunnel according to claim 5, characterized in that, Abnormal monitoring values are preliminarily identified through the Z-score method, and the machine learning model uses the ClusteringAlgorithms algorithm to further screen and filter the abnormal monitoring values.

7. The method for limited dewatering assessment of a subsea tunnel according to claim 1, wherein, The steps S8 and S9 comprise: 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 the qualified evaluation. If the evaluation result is unqualified and the unqualified influencing factor does not involve the pressure relief value, an alarm prompt is triggered, a limited drainage scheme is re-formulated for evaluation, and the process is repeated until the limited drainage scheme evaluation is qualified. After the evaluation is qualified, the current sample data is updated to the initial sample data set as training data.

8. A construction method for a limited drainage waterproofing for a subsea tunnel, based on the implementation of the limited drainage waterproofing assessment method according to any one of claims 1 to 7, characterized in that, The method comprises the following steps: S81: According to the qualified limited drainage simulation construction scheme, a waterproof layer is arranged between the primary support layer and the secondary lining layer of the submarine tunnel; S82: According to the qualified limited drainage simulation construction scheme, a drainage zone and a water stop zone are arranged in the submarine tunnel; S83: According to the qualified limited drainage simulation construction scheme, a side wall water storage pool, a water diversion channel and a central drainage ditch are arranged in the submarine tunnel; S84: According to the qualified limited drainage simulation construction scheme, a circumferential drainage pipe and a longitudinal drainage pipe are arranged in the submarine tunnel; S85: A pressure relief valve is arranged at the water outlet of the longitudinal drainage pipe, and the pressure relief value of each pressure relief valve is adjusted according to the expected adjustment range of the feedback pressure relief value, and the construction is completed.

9. The construction method for limited drainage of a subsea tunnel according to claim 8, wherein, The method comprises: The waterproof layer is an EVA waterproof board and a non-woven geotextile, the non-woven geotextile is laid between the primary support layer and the EVA waterproof board, and the EVA waterproof board is arranged between the non-woven geotextile and the secondary lining layer; The water stop zone is a rubber water stop zone arranged at the deformation joint and the circumferential construction joint of the submarine tunnel; The circumferential drainage pipe is arranged on the back of the arch wall, the longitudinal drainage pipe is arranged on the back of the side wall, the circumferential drainage pipe and the longitudinal drainage pipe are connected, and the water outlet of the longitudinal drainage pipe faces the side wall water storage pool; The side wall water storage pool is a plurality of groups of water storage pools arranged on the left and right side walls of the tunnel; The side wall water storage pool is connected through the water diversion channel and the central drainage ditch.

10. A monitoring method for the limited drainage of a subsea tunnel, based on the implementation of the evaluation method of the limited drainage according to any one of claims 1 to 9, characterized in that, The method comprises the following steps: S101: A plurality of groups of environmental monitoring sensors are sequentially arranged at the haunches and side walls on the left and right sides of the submarine tunnel; S102: A plurality of groups of monitoring cameras are sequentially arranged at the vaults of the submarine tunnel; S103: A plurality of groups of underground water level monitors are sequentially arranged at the soles and haunches of the submarine tunnel; S104: Every 25 minutes, the monitoring data collected by real-time monitoring is assigned to the analysis and prediction feedback model for evaluation; S105: If the evaluation is unqualified, the pressure relief valve and / or the construction scheme at the abnormal position are adjusted until the evaluation is qualified.

Images