A station deep foundation pit construction dynamic mechanical property analysis and stability control method and system

Abstract

The application discloses a kind of station deep foundation pit construction dynamic mechanical property analysis and stability control method, comprising, fusion multi-source data establishes initial prediction model, and parallel simulation optimization generates recommended construction scheme, synchronously collects real-time mechanical monitoring data in construction, with the measured data and prediction result dynamic comparison to trigger model online update, and the updated model is used to re-simulate evaluation and dynamically adjust construction parameters or optimize soil reinforcement scheme;Its beneficial effect is, effectively alleviate the survey design construction monitoring link fragmentation and the problem of monitoring data after-alarm lag, significantly improve the dynamic mechanical response prediction accuracy and stability active control ability of whole process of construction.

Description

Technical Field

[0001] This invention relates to the field of geotechnical engineering construction technology, specifically to a method and system for dynamic mechanical property analysis and stability control in deep foundation pit construction for railway stations. Background Technology

[0002] In the construction of deep foundation pits for subway stations, especially when building large-span T-shaped transfer stations in urban centers, construction faces extremely complex challenges. These challenges stem from variable hydrogeological conditions, stress redistribution in the soil caused by deep foundation pit excavation, uncertainties in the dynamic interaction between the support structure and the soil, and the unique disturbance to the soil caused by the tunnel boring machine (TBM) entering and exiting the pit. Existing technologies typically isolate the investigation, design, construction, and monitoring stages. The design phase relies on simplified models and static calculations, making it difficult to accurately predict the dynamic mechanical response throughout the construction process; construction schemes are often selected based on experience, lacking systematic optimization based on parallel simulations of multiple schemes; monitoring data is mostly used for post-event alarms, failing to deeply couple with the prediction model to form a real-time feedback and control closed loop. This leads to a lag in the understanding and control of key issues such as the deformation mechanism of the foundation pit, the stress law of the support structure, especially the mechanical properties of the internal support system under complex planar shapes, and the soil reinforcement effect at the TBM's entry and exit points. This results in bottlenecks in improving construction safety and efficiency, and the guarantee mechanism for information-based construction is also inadequate. Summary of the Invention

[0003] This invention proposes a method for analyzing the dynamic mechanical properties and controlling the stability of deep foundation pits in railway stations, including: S1. Collect and integrate engineering geological survey data, hydrogeological survey data, indoor physical model test data and numerical simulation analysis data of the construction site, and establish an initial construction mechanical behavior prediction model calibrated by multi-source data. S2. Based on the initial construction mechanical behavior prediction model, parallel numerical simulations are performed on various deep foundation pit excavation methods, support structure construction process parameters, and internal support system layout and parameters. Recommended deep foundation pit excavation and support construction schemes for stations are generated through optimization analysis. S3. During the construction process according to the recommended construction plan, the informatization of construction monitoring and the collection of real-time mechanical monitoring data of the foundation pit soil, support structure and internal support system shall be implemented simultaneously. S4. Dynamically compare the real-time mechanical monitoring data with the prediction results of the initial construction mechanical behavior prediction model corresponding to the construction stage, and trigger online model updates based on the deviation analysis results; S5. Use the updated model to re-simulate and risk-assess subsequent excavation, support construction, and shield tunneling entry and exit procedures, and dynamically adjust subsequent construction parameters or initiate targeted soil reinforcement scheme optimization based on the assessment results.

[0004] Furthermore, in S1, the fusion process specifically adopts a data assimilation framework based on Bayes' theorem.

[0005] Furthermore, in S2, the parallel numerical simulation of various deep foundation pit excavation methods, support structure construction process parameters, and internal support system layout and parameters, and the generation of recommended deep foundation pit excavation and support construction schemes for stations through optimization analysis, specifically involves: using a multi-objective evolutionary algorithm to perform parallel numerical simulation and optimization analysis of the various deep foundation pit excavation methods, support structure construction process parameters, and internal support system layout and parameters, and generating recommended deep foundation pit excavation and support construction schemes for stations.

[0006] Furthermore, in S4, the tolerance range set based on the current confidence level of the model is an adaptive threshold.

[0007] Furthermore, in S5, the specific steps of initiating targeted soil reinforcement scheme optimization are as follows: when the evaluation results show that the tunnel boring machine is about to enter or leave the pit sidewall opening area, the soil reinforcement scheme optimization program for the tunnel boring machine's entry and exit from the opening is initiated.

[0008] Furthermore, between S3 and S4, there is also a function to collect activity signals of microcracks inside the concrete of the support structure.

[0009] Furthermore, when the dynamic mechanical characteristics analysis and stability control method for deep foundation pit construction of the station is applied to the deep foundation pit of a large-span T-shaped transfer station, the optimized design of the layout and parameters of the internal support system in S2 is particularly crucial.

[0010] Furthermore, after generating the dynamic adjustment instruction in S5, the instruction, the basis for adjustment, and the risk assessment conclusion are recorded in the tamper-proof distributed database in the form of a structured process log.

[0011] Furthermore, this application also proposes a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the aforementioned method for dynamic mechanical characteristic analysis and stability control of deep foundation pit construction for railway stations.

[0012] Furthermore, this application also proposes a dynamic mechanical property analysis and stability control system for deep foundation pit construction of railway stations, used to implement the aforementioned method for dynamic mechanical property analysis and stability control of deep foundation pit construction of railway stations. The system includes: The data fusion and initial modeling module is used to collect and fuse engineering geological survey data, hydrogeological survey data, indoor physical model test data and numerical simulation analysis data of the construction site, and establish an initial construction mechanical behavior prediction model calibrated by multi-source data. The construction scheme optimization module is used to perform parallel numerical simulations on various deep foundation pit excavation methods, support structure construction process parameters, and internal support system layout and parameters based on the initial construction mechanical behavior prediction model. Through optimization analysis, it generates recommended deep foundation pit excavation and support construction schemes for stations. The real-time monitoring data acquisition module is used to simultaneously implement information-based construction monitoring and collect real-time mechanical monitoring data of the foundation pit soil, support structure and internal support system during the construction process according to the recommended construction plan. The dynamic feedback and model update module is used to dynamically compare the real-time mechanical monitoring data with the prediction results of the initial construction mechanical behavior prediction model corresponding to the construction stage, and trigger online model updates based on the deviation analysis results. The risk assessment and dynamic control module is used to re-simulate and assess the risks of subsequent excavation, support construction and shield tunneling entry and exit procedures using the updated model, and dynamically adjust subsequent construction parameters or initiate targeted soil reinforcement scheme optimization based on the assessment results.

[0013] This invention discloses a method for dynamic mechanical characteristic analysis and stability control during deep foundation pit construction in railway stations. The method includes the following steps: establishing an initial prediction model by integrating multi-source data; generating recommended construction schemes through parallel simulation and optimization; synchronously collecting real-time mechanical monitoring data during construction; dynamically comparing measured data with prediction results to trigger online model updates; and using the updated model to re-simulate and evaluate, and dynamically adjust construction parameters or optimize soil reinforcement schemes. This method effectively alleviates the problems of fragmentation between surveying, design, construction, and monitoring stages, as well as the lag in post-event alarms from monitoring data, significantly improving the accuracy of dynamic mechanical response prediction and the proactive stability control capability throughout the entire construction process. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating the dynamic mechanical properties analysis and stability control method for deep foundation pit construction of railway stations proposed in this invention. Figure 2 This is a schematic diagram of the sub-process for establishing the initial prediction model in the data assimilation framework based on Bayes' theorem in this invention; Figure 3 This is a schematic diagram of the sub-process of generating recommended solutions through parallel numerical simulation and optimization analysis of multiple schemes in this invention; Figure 4 This is a schematic diagram of a sub-process of the closed loop of online model update and dynamic feedback control in this invention. Detailed Implementation

[0015] refer to Figure 1 This invention proposes a method for analyzing the dynamic mechanical properties and controlling the stability of deep foundation pits in railway stations, referencing... Figure 1 ,include: S1. Collect and integrate engineering geological survey data, hydrogeological survey data, indoor physical model test data and numerical simulation analysis data of the construction site, and establish an initial construction mechanical behavior prediction model calibrated by multi-source data.

[0016] Among them, engineering geological survey data and hydrogeological survey data are used to provide stratigraphic distribution, soil physical and mechanical property parameters, groundwater depth and permeability; indoor physical model test data are based on earth pressure distribution, support pile wall deformation mode and internal force change data of internal support members obtained from scaled-down foundation pit model tests; numerical simulation analysis data are the output results of simulation calculations of the entire construction process based on preliminary design parameters; fusion processing establishes a unified numerical analysis framework, uses survey data as the initial input of the model, and uses physical model test data and previous numerical simulation results as calibration benchmarks (refer to Table 1). Through back analysis or data assimilation techniques, the soil constitutive parameters and structural interface parameters in the numerical model are adjusted, so that the model output matches the multi-source benchmark data and establishes an initial construction mechanical behavior prediction model.

[0017]

[0018] Specifically, a sensor network is deployed to collect data, using engineering geological survey data and hydrogeological survey data as initial inputs. This data is then combined with indoor physical model test data and numerical simulation analysis data, and fused using a data assimilation framework based on Bayes' theorem (see reference). Figure 2 The soil constitutive model parameters are treated as random variables. Their prior distribution is determined based on engineering geological survey data. Indoor physical model test data and previous numerical simulation analysis data are used as observation evidence inputs to calculate the posterior probability distribution of the parameters. The mean or maximum posterior probability estimate of the posterior distribution is taken as the calibrated parameters to establish an initial construction mechanical behavior prediction model.

[0019] In a specific implementation scenario, during the preparation stage of deep foundation pit construction at a station, engineering geological survey data and hydrogeological investigation data of the site are obtained through geological drilling. At the same time, scaled-down foundation pit model tests are conducted in the laboratory to obtain indoor physical model test data, and preliminary numerical simulations are performed using finite element software to obtain numerical simulation analysis data. The above four types of data are input into a data assimilation framework based on Bayes' theorem to calibrate key parameters such as the elastic modulus and friction angle of the site soil, thereby outputting an initial construction mechanical behavior prediction model calibrated by multi-source data.

[0020] S2. Based on the initial construction mechanical behavior prediction model, parallel numerical simulations are performed on various deep foundation pit excavation methods, support structure construction process parameters, and internal support system layout and parameters. Recommended deep foundation pit excavation and support construction schemes for stations are generated through optimization analysis.

[0021] The various deep foundation pit excavation methods include layered and segmented excavation steps, excavation depth and width of each layer, and excavation exposure time; the construction process parameters of the support structure involve the reinforcement ratio of the support piles or diaphragm walls, concrete strength grade, and construction joint treatment method; the layout and parameters of the internal support system include the number of support rows, the planar layout of each support row, horizontal spacing, member cross-sectional dimensions and material strength, prestressing application value, and construction sequence; parallel numerical simulation is achieved by running different parameter combinations of the entire construction process simulation simultaneously through a computing cluster; optimization analysis is achieved by setting multiple objectives such as controlling the maximum lateral displacement of the foundation pit within the allowable value, minimizing the amount of support material used, and shortening the critical construction period, and using optimization algorithms to select the scheme with the best comprehensive performance from a large number of simulation results.

[0022] Specifically, a multi-objective evolutionary algorithm was used to conduct parallel numerical simulation and optimization analysis on various deep foundation pit excavation methods, construction process parameters of support structures, and layout and parameters of internal support systems (see reference). Figure 3 The key point displacement control values ​​of the foundation pit, the maximum axial force of the internal support system, and the direct cost of the project are set as multiple competitive objectives. An iterative search is carried out in the design space composed of excavation parameters, support parameters, and bracing parameters. After evolution and convergence, a set of Pareto optimal solutions is obtained as a recommended set of construction schemes for the excavation and support of deep foundation pits in railway stations. When applied to the deep foundation pits of large-span T-shaped transfer stations, the focus is on the layout of the corner supports, the stiffness of the member sections, the connection node construction with the walers, and the reinforcement measures of the corner soil to optimize the layout and parameters of the internal support system.

[0023] In a specific implementation scenario, for the deep foundation pit of a large-span T-shaped transfer station, based on the established initial construction mechanical behavior prediction model, a parameter combination including excavation depth, number of supports and corner support structure parameters is constructed. Hundreds of parallel numerical simulations are performed using a computing cluster, and displacement control and cost objectives are iteratively optimized through a multi-objective evolutionary algorithm. Finally, a customized recommended construction scheme including corner triangular trusses to replace single diagonal braces and local soil reinforcement measures is output.

[0024] S3. During the construction process according to the recommended construction plan, the informatization of construction monitoring and the collection of real-time mechanical monitoring data of the foundation pit soil, support structure and internal support system shall be implemented simultaneously.

[0025] The informatization of construction monitoring is achieved through the deployment of a sensor network. The sensor network includes earth pressure cells and pore water pressure gauges buried in the soil, steel stress gauges and concrete strain gauges installed in the support structure, axial force gauges deployed on the internal support members, as well as inclinometer tubes, settlement observation points, and total station monitoring points for measuring the deformation of the soil and structure around the foundation pit. Real-time mechanical monitoring data is data collected by the above sensors and monitoring instruments at a predetermined frequency and transmitted in real time to the central data processing and analysis platform via a wireless transmission network.

[0026] Specifically, during the excavation and support construction of the foundation pit, earth pressure cells, pore water pressure gauges, steel stress gauges, concrete strain gauges, axial force gauges, and inclinometers are installed simultaneously, and total station monitoring points are set up. Data on soil pressure and pore water pressure in the foundation pit, internal forces and deformations of the support structure, axial forces of internal supports, and deformations of the surrounding soil are collected at predetermined frequencies. The data are transmitted in real time to the central data processing and analysis platform via a wireless transmission network. At the same time, an acoustic emission monitoring system is deployed between S3 and S4 to collect signals of microcrack activity inside the concrete of the support structure.

[0027] In one specific implementation scenario, before each layer of earthwork is excavated in the foundation pit, earth pressure cells and pore water pressure gauges are embedded in the soil at the corresponding locations. Steel stress gauges and concrete strain gauges are bound to the steel cages of the support piles or diaphragm walls. Axial force gauges are welded to the surface of the internal support members. Inclined tubes and settlement observation points are installed around the foundation pit. As construction progresses, the above-mentioned mechanical monitoring data are collected around the clock. At the same time, the acoustic emission monitoring system is activated to listen to the acoustic emission signals generated by the propagation of microcracks inside the concrete in real time. All data are aggregated to the central platform in real time.

[0028] S4. Dynamically compare the real-time mechanical monitoring data with the prediction results of the initial construction mechanical behavior prediction model corresponding to the construction stage, and trigger online model updates based on the deviation analysis results.

[0029] Among them, the corresponding construction stage is the spatiotemporal location of the same excavation depth and support setting status in the numerical model according to the actual construction progress; dynamic comparison is to compare the measured value of the same physical quantity at the same location with the model prediction value to calculate the relative deviation; the deviation analysis results include single-point deviation, spatial distribution pattern of deviation, and temporal development trend; online model update is a mechanism triggered when the system identifies systematic deviation or deviation exceeds the tolerance range set based on the current confidence level of the model. This mechanism uses parameter identification algorithm to reversely adjust the key uncertainty parameters of soil elastic modulus, friction angle, or support boundary conditions in the model driven by real-time monitoring data, so as to keep the model prediction capability in sync with the actual engineering progress.

[0030] Specifically, an adaptive threshold is set as the tolerance range based on the current confidence level of the model (refer to Table 2 and...). Figure 4The initial value of this threshold is set considering the discreteness of engineering survey data and the residual error level of the initial model calibration. As construction progresses and the model is continuously updated and corrected in multiple construction stages, the tolerance range is narrowed accordingly. The real-time mechanical monitoring data is compared with the prediction results of the initial construction mechanical behavior prediction model for the corresponding construction stage. At the same time, the spatial location clustering degree of acoustic emission events and the energy release rate are compared with the high stress area of ​​the structure predicted by the model. When the deviation exceeds the adaptive threshold or the acoustic emission activity characteristics indicate the accumulation of hidden damage, the model is triggered to update online and adjust the key uncertainty parameters in the model in reverse.

[0031]

[0032] In one specific implementation scenario, when the foundation pit is excavated to the third layer, the real-time collected data on soil displacement and support axial force of this layer are compared with the predicted values ​​of the corresponding third-layer excavation conditions in the initial model. It is found that the deviation of the support axial force exceeds the adaptive threshold under the current confidence level. At the same time, the acoustic emission monitoring system captures high-energy signals of high-frequency clusters at the support nodes. The system then triggers the online model update mechanism, using real-time data to reverse identify and correct the elastic modulus of the soil and the support boundary constraint parameters in this area, so that the model output matches the current actual state.

[0033] S5. Use the updated model to re-simulate and risk-assess subsequent excavation, support construction, and shield tunneling entry and exit procedures, and dynamically adjust subsequent construction parameters or initiate targeted soil reinforcement scheme optimization based on the assessment results.

[0034] Among these, the re-simulation is a forward-looking calculation of the construction procedures that have not yet been carried out based on the latest calibrated model state; the risk assessment is to analyze the overall stability of the foundation pit, whether the internal forces of the support components exceed the limits, and the risk of settlement of adjacent buildings in subsequent steps; dynamic adjustment includes adjusting the excavation rate of the next layer of earthwork, optimizing the installation timing and prestress of the next layer of internal support, and modifying the strength of local support; the initiation of targeted soil reinforcement scheme optimization is to start the soil reinforcement scheme optimization program for the shield machine's entry and exit from the tunnel entrance when the assessment results show that the shield machine is about to enter or leave the tunnel entrance area on the side wall of the foundation pit, and call the updated model to quickly simulate and compare the mechanical response of the soil around the tunnel entrance under different reinforcement methods, and generate customized tunnel entrance reinforcement decisions.

[0035] Specifically, the updated model was used to re-simulate and re-assess the risks of subsequent excavation, support construction, and shield tunneling entry and exit procedures (see reference). Figure 4When the assessment results indicate that the tunnel boring machine is about to enter or leave the pit sidewall opening area, the updated model is called to quickly simulate the mechanical response of the soil around the opening under different reinforcement methods such as high-pressure jet grouting, freezing, and chemical grouting. The effects of different reinforcement strengths, ranges, and depths are compared and analyzed. Under the premise of ensuring the stability of the soil at the opening and controlling the shield thrust and surface settlement to meet the requirements, the solution with the lowest reinforcement cost or the shortest construction period is sought, and a customized opening reinforcement decision is generated. Based on the assessment results, subsequent construction parameters are dynamically adjusted or targeted soil reinforcement schemes are optimized, and a dynamic adjustment command is generated. The command, adjustment basis, and risk assessment conclusion are recorded in the form of a structural condition log in a tamper-proof distributed database.

[0036] In one specific implementation scenario, on the eve of the tunnel boring machine's (TBM) entry into the tunnel, the high-precision model updated from the previous steps is used to simulate and evaluate the tunnel entry process. It is found that the strength of the soil reinforcement at the tunnel entrance designed in the original plan is too weak under the current excavation and unloading conditions. The system then starts an optimization program for the soil reinforcement scheme for the TBM's entry and exit conditions. It quickly compares the mechanical response of different grouting ranges and reinforcement depths, outputs an optimized scheme that increases the grouting strength and reduces the reinforcement range, issues a dynamic adjustment command based on this scheme, and encapsulates the command, the adjustment basis based on real-time monitoring data, and the tunnel entry risk assessment conclusion into a structural condition log, which is written into a tamper-proof distributed database for permanent storage.

[0037] Furthermore, in S1, the fusion process specifically adopts a data assimilation framework based on Bayes' theorem.

[0038] The data assimilation framework based on Bayes' theorem is a probabilistic statistical inference method used to quantify and fuse uncertainties in multi-source heterogeneous data. The fusion process treats the soil constitutive model parameters as random variables, determines their prior probability distribution based on engineering geological survey data, uses indoor physical model test data and previous numerical simulation analysis data as observational evidence input, calculates the posterior probability distribution of the parameters, and takes the mean or maximum posterior probability estimate of the posterior distribution as the calibrated and determined parameters.

[0039] Specifically, when establishing the initial construction mechanical behavior prediction model, for key uncertain parameters such as soil elastic modulus and friction angle, the prior probability density function is first constructed based on engineering geological survey data. Then, indoor physical model test data and numerical simulation analysis data are introduced as likelihood functions. Bayes' theorem is used to calculate the posterior probability density function. By solving for the maximum point or expected value of the posterior distribution, the key parameters are quantitatively calibrated and determined.

[0040] In one specific implementation scenario, during the preparation stage of deep foundation pit construction at the station, a uniform prior distribution is set for the silty clay layer in the site based on the range of compression modulus given in the survey report. Then, the earth pressure data measured by the centrifuge model test and the results of the finite element forward modeling are used as observation samples, substituted into the Bayesian update equation, and the posterior Gaussian distribution of the compression modulus is calculated. The mean value is then substituted into the numerical model to complete the Bayesian calibration of the initial model.

[0041] Furthermore, in S2, the parallel numerical simulation of various deep foundation pit excavation methods, support structure construction process parameters, and internal support system layout and parameters, and the generation of recommended deep foundation pit excavation and support construction schemes for stations through optimization analysis, specifically involves: using a multi-objective evolutionary algorithm to perform parallel numerical simulation and optimization analysis of the various deep foundation pit excavation methods, support structure construction process parameters, and internal support system layout and parameters, and generating recommended deep foundation pit excavation and support construction schemes for stations.

[0042] The multi-objective evolutionary algorithm is an optimization algorithm that performs a global search based on the natural selection mechanism within a multi-dimensional design space consisting of excavation parameters, support parameters, and bearing parameters. The parallel numerical simulation and optimization analysis is a process in which a computing cluster simultaneously runs simulations of the entire construction process with hundreds or even thousands of different parameter combinations, and uses the simulation results as the basis for evaluating individual fitness, and obtains the Pareto optimal solution set through iterative evolution.

[0043] Specifically, the displacement control values ​​of key points in the foundation pit, the maximum axial force of the internal support system, and the direct cost of the project are set as multiple competitive objectives that need to be optimized simultaneously. The algorithm generates an initial population within the design space, with each individual representing a specific combination of excavation and support parameters. The fitness of individuals in the population is evaluated through a parallel computing cluster. After selection, crossover, and mutation operations, a new generation of population is generated. After several generations of evolution, a set of Pareto optimal solutions that cannot be improved simultaneously is converged, which serves as the recommended construction scheme set.

[0044] In one specific implementation scenario, for the deep foundation pit of a large-span T-shaped transfer station, a parameter coding group is constructed, which includes the number of excavation layers, the number of support channels, the stiffness of the corner support, and the scope of local reinforcement. Using a high-performance computing platform, 200 schemes are simultaneously simulated in three dimensions using finite element methods. With the maximum lateral displacement of the foundation pit, the maximum axial force at the corner, and the total cost as optimization objectives, after 50 generations of genetic algorithm iterations, 10 Pareto optimal schemes are output for decision-makers to make a final choice based on safety reserves and cost limits.

[0045] Furthermore, in S4, the tolerance range set based on the current confidence level of the model is an adaptive threshold.

[0046] The adaptive threshold is a dynamic criterion used to determine whether the deviation between measured data and predicted data triggers online model updates. The tolerance range set based on the current confidence level of the model refers to the mechanism in which the initial value of the threshold is set considering the dispersion of engineering survey data and the residual error level of the initial model calibration. As construction progresses and the model is continuously updated and corrected in multiple construction stages, the model confidence level increases and the tolerance range narrows accordingly.

[0047] Specifically, in the early stages of foundation pit excavation, due to the limited monitoring data available for calibration and the low confidence level of the model, the adaptive threshold is set to a relatively wide range, allowing for larger prediction deviations without triggering frequent updates. As construction progresses, monitoring data from multiple periods continuously drive the model to make inverse corrections, reducing the uncertainty of model parameters and increasing confidence. The system automatically narrows the tolerance range to a smaller interval, thus becoming sensitive to small but systematic deviations and ensuring high-precision tracking of the model.

[0048] In one specific implementation scenario, during the first excavation stage of the foundation pit, the model confidence level is low, and the adaptive threshold is set to 15% for displacement deviation or 20% for axial force deviation. The system only triggers updates for large deviations exceeding this wide range. When the excavation reaches the third layer, the model has undergone two successful corrections, and the confidence level has increased. The system automatically narrows the adaptive threshold to 8% for displacement deviation or 10% for axial force deviation. Once the deviation between the real-time monitoring data and the predicted value exceeds this narrow range, the highly sensitive online model update program is immediately initiated.

[0049] Furthermore, in S5, the specific steps of initiating targeted soil reinforcement scheme optimization are as follows: when the evaluation results show that the tunnel boring machine is about to enter or leave the pit sidewall opening area, the soil reinforcement scheme optimization program for the tunnel boring machine's entry and exit from the opening is initiated.

[0050] The soil reinforcement scheme optimization program for the tunnel boring machine (TBM) entry and exit conditions is a special optimization logic that calls the updated model to quickly simulate and compare the mechanical response of the soil around the tunnel entrance under different reinforcement methods when the high-risk timing node of the TBM's imminent crossing of the tunnel entrance is identified. The targeted soil reinforcement scheme optimization is an intervention mechanism that is automatically triggered when the risk indicators in the assessment results exceed the preset safety margin.

[0051] Specifically, using the updated high-precision model, the mechanical response of the soil around the tunnel entrance under different reinforcement methods such as high-pressure jet grouting, freezing, and chemical grouting is rapidly simulated. The effects of different reinforcement intensities, ranges, and depths are compared and analyzed. Under the premise of ensuring the stability of the soil at the tunnel entrance and controlling the shield thrust and surface settlement to meet the requirements, the solution with the lowest reinforcement cost or the shortest construction period is sought, and customized tunnel entrance reinforcement decisions are generated.

[0052] In one specific implementation scenario, on the eve of the tunnel boring machine's (TBM) entry into the tunnel, the dynamic risk assessment module uses the updated model from the previous steps to simulate the tunnel entry process. It finds that the shear strength of the soil reinforced at the tunnel entrance is insufficient to resist the thrust of the TBM under the current excavation and unloading conditions. The system then automatically starts the soil reinforcement scheme optimization program for the TBM's entry and exit conditions, quickly compares the mechanical responses of grouting reinforcement and freezing reinforcement, and outputs an optimized decision to increase the grouting strength and reduce the reinforcement range, guiding the on-site timely adjustment of reinforcement construction.

[0053] Furthermore, between S3 and S4, there is also a function to collect activity signals of microcracks inside the concrete of the support structure.

[0054] The acquisition of microcrack activity signals inside the concrete of the support structure is achieved by deploying an acoustic emission monitoring system; the acoustic emission monitoring system is used to acquire transient elastic wave signals emitted by the generation and propagation of microcracks; the microcrack activity signals include the spatial location clustering degree and energy release rate characteristics of acoustic emission events.

[0055] Specifically, an array of acoustic emission sensors is deployed in areas of high internal force or stress concentration in the concrete support structure to continuously monitor acoustic emission signals generated by rapid energy release within the concrete. The acoustic emission events are located in three-dimensional space using a time-difference positioning algorithm, and the energy release rate per unit time is statistically analyzed. The acoustic emission activity characteristics are correlated and compared with the high-stress areas of the structure predicted by the model. When the acoustic emission activity characteristics indicate the possible accumulation of hidden damage or precursors to brittle failure, even if the traditional deformation monitoring data has not exceeded the limits, the system triggers a high-level warning and immediately initiates the review and update of key areas of the model and preventive adjustments to the construction plan.

[0056] In one specific implementation scenario, after the second concrete support of the deep foundation pit is poured and subjected to force, the acoustic emission sensor deployed at the support node continuously collects signals. On the third day, it is found that the spatial clustering degree of acoustic emission events in the node area increases sharply and the energy release rate exceeds the threshold, while the axial force gauge data is still within the allowable range. Based on this, the system determines that there is micro-damage accumulation, triggers a high-level warning in advance, and instructs to slow down the excavation speed in the area to avoid sudden brittle fracture of the support node.

[0057] Furthermore, when the dynamic mechanical characteristics analysis and stability control method for deep foundation pit construction of the station is applied to the deep foundation pit of a large-span T-shaped transfer station, the optimized design of the layout and parameters of the internal support system in S2 is particularly crucial.

[0058] Among them, the deep foundation pit of the large-span T-shaped transfer station has complex planar shape, stress concentration at the corners, and interlaced irregular supports. The optimization design of the internal support system layout and parameters is particularly critical. In the parallel numerical simulation and optimization analysis of S2, it is necessary to carry out special parametric modeling and densified search for the layout of the corner supports, the stiffness of the member sections, the connection node construction with the walers, and the reinforcement measures of the corner soil.

[0059] Specifically, when performing multi-objective evolutionary algorithm search, for the corner area of ​​the T-shaped foundation pit, a gene coding fragment containing multiple intervention measures such as replacing a single diagonal brace with a triangular truss at the corner, adding a cap beam to the corner, and grouting reinforcement of the corner soil is constructed separately. This fragment is given independent mutation and crossover probabilities to ensure that the optimal solution of the corner support system is not overwhelmed by the average effect of the global search, thereby selecting the optimal support arrangement specifically for the complex stress state of the corner.

[0060] In a specific implementation scenario, for a deep foundation pit of a large-span T-shaped transfer station, when constructing the design space of the multi-objective evolutionary algorithm, a combination code was specifically set for the support system within 5 meters of the corner, which includes three elements: triangular truss, corner bracing, and corner reinforcement. After algorithm iteration, the system output a recommended scheme that replaces the original single diagonal bracing with triangular trusses at the corner and is supplemented by local soil reinforcement, which effectively solves the problem of excessive deformation caused by stress concentration at the corner.

[0061] Furthermore, after generating the dynamic adjustment instruction in S5, the instruction, the basis for adjustment, and the risk assessment conclusion are recorded in the tamper-proof distributed database in the form of a structured process log.

[0062] The structured process log is a standardized data record that includes the specific content of dynamic adjustment instructions, the basis for real-time monitoring data deviations that trigger the instructions, and the risk assessment conclusions output by the updated model; the tamper-proof distributed database is a data storage system that only adds to and does not change, built on blockchain technology or hash chain storage.

[0063] Specifically, after the system dynamically adjusts subsequent construction parameters or initiates targeted soil reinforcement scheme optimization based on the evaluation results and generates dynamic adjustment instructions, the data recording module automatically extracts the operation code of the instruction, the corresponding real-time monitoring over-limit data screenshots, and the re-simulation output report of the updated model, packages them into a structured JSON format log with timestamps, and writes it to multiple nodes in the distributed network through consensus algorithms or hash encryption to ensure that no single node can tamper with historical decision records.

[0064] In a specific implementation scenario, after the system issues a dynamic adjustment instruction to slow down the excavation rate of the third layer, the instruction code, the data on the over-limit support axial force that triggered the instruction, and the model's predicted deformation trend diagram are immediately encapsulated into a working condition log. The log is then linked to the previous blockchain with a SHA-256 hash value and broadcast to three backup nodes in the project department's cloud for storage. This provides irrefutable data evidence for subsequent determination of responsibility for engineering quality accidents and review of solutions.

[0065] Furthermore, this application also proposes a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the aforementioned method for dynamic mechanical characteristic analysis and stability control of deep foundation pit construction for railway stations.

[0066] Furthermore, this application also proposes a dynamic mechanical property analysis and stability control system for deep foundation pit construction of railway stations, used to implement the aforementioned method for dynamic mechanical property analysis and stability control of deep foundation pit construction of railway stations. The system includes: The data fusion and initial modeling module is used to collect and fuse engineering geological survey data, hydrogeological survey data, indoor physical model test data and numerical simulation analysis data of the construction site, and establish an initial construction mechanical behavior prediction model calibrated by multi-source data. The construction scheme optimization module is used to perform parallel numerical simulations on various deep foundation pit excavation methods, support structure construction process parameters, and internal support system layout and parameters based on the initial construction mechanical behavior prediction model. Through optimization analysis, it generates recommended construction schemes for deep foundation pit excavation and support of the station. The real-time monitoring data acquisition module is used to simultaneously implement information-based construction monitoring and collect real-time mechanical monitoring data of the foundation pit soil, support structure and internal support system during the construction process according to the recommended construction plan. The dynamic feedback and model update module is used to dynamically compare the real-time mechanical monitoring data with the prediction results of the initial construction mechanical behavior prediction model corresponding to the construction stage, and trigger online model updates based on the deviation analysis results. The risk assessment and dynamic control module is used to re-simulate and assess the risks of subsequent excavation, support construction and shield tunneling entry and exit procedures using the updated model, and dynamically adjust subsequent construction parameters or initiate targeted soil reinforcement scheme optimization based on the assessment results.

[0067] Specifically, the output of the data fusion and initial modeling module is connected to the input of the construction scheme optimization module; the output of the construction scheme optimization module is connected to the control end of the real-time monitoring data acquisition module; the output of the real-time monitoring data acquisition module is connected to the input of the dynamic feedback and model update module; the output of the dynamic feedback and model update module is connected to the input of the risk assessment and dynamic control module; and the output of the risk assessment and dynamic control module is connected to the construction execution agency and data storage unit, forming a complete closed-loop system architecture of data perception-model update-decision control.

[0068] Specifically, the data fusion and initial modeling module is deployed on a central server and accesses the geological exploration database through a data interface; the construction scheme optimization module integrates a high-performance computing scheduling engine and calls a finite element solver cluster; the real-time monitoring data acquisition module includes a wireless gateway and signal conditioning circuits to aggregate data from the on-site sensor network; the dynamic feedback and model update module has a built-in parameter inversion algorithm library to perform deviation calculation and parameter identification; and the risk assessment and dynamic control module has a built-in rule engine to convert risk indicators into construction parameter adjustment instructions and send them to the on-site PLC controller.

[0069] In a specific implementation scenario, at the construction site of a deep foundation pit for a large transfer station, the data fusion and initial modeling module reads the preliminary exploration data to complete the model initialization. The construction scheme optimization module calls cloud computing power to generate the optimal excavation and support scheme to guide the first layer of earthwork excavation. The real-time monitoring data acquisition module collects data streams from hundreds of sensors buried inside and outside the pit. The dynamic feedback and model update module compares the data in real time and finds that the deep horizontal displacement deviation exceeds the limit and automatically corrects the model parameters. The risk assessment and dynamic control module finds that there is a risk of collapse in the next layer of excavation based on the updated model, and then issues control instructions to the excavator group control system to slow down the excavation rate and increase temporary steel supports, realizing automatic closed-loop control of the entire process.

[0070] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for analyzing the dynamic mechanical properties and controlling the stability of deep foundation pits in railway stations, characterized in that, include: S1. Collect and integrate engineering geological survey data, hydrogeological survey data, indoor physical model test data and numerical simulation analysis data of the construction site, and establish an initial construction mechanical behavior prediction model calibrated by multi-source data. S2. Based on the initial construction mechanical behavior prediction model, parallel numerical simulations are performed on various deep foundation pit excavation methods, support structure construction process parameters, and internal support system layout and parameters. Recommended deep foundation pit excavation and support construction schemes for the station are generated through optimization analysis. The various deep foundation pit excavation methods include layered and segmented excavation steps, the excavation depth and width of each layer, and the excavation exposure time. Support structure construction process parameters involve the reinforcement ratio of support piles or diaphragm walls, concrete strength grade, and construction joint treatment methods. The internal support system layout and parameters include the number of support rows, the planar layout of each support row, horizontal spacing, member cross-sectional dimensions and material strength, prestressing application value, and construction sequence. S3. During the construction process according to the recommended construction plan, the informatization of construction monitoring and the collection of real-time mechanical monitoring data of the foundation pit soil, support structure and internal support system shall be implemented simultaneously. S4. Dynamically compare the real-time mechanical monitoring data with the prediction results of the initial construction mechanical behavior prediction model corresponding to the construction stage, and trigger online model updates based on the deviation analysis results; S5. Use the updated model to re-simulate and risk-assess subsequent excavation, support construction, and shield tunneling entry and exit procedures, and dynamically adjust subsequent construction parameters or initiate targeted soil reinforcement scheme optimization based on the assessment results.

2. The method for dynamic mechanical characteristic analysis and stability control of deep foundation pit construction for railway stations according to claim 1, characterized in that, In S1, the fusion process specifically adopts a data assimilation framework based on Bayes' theorem.

3. The method for dynamic mechanical characteristic analysis and stability control of deep foundation pit construction for railway stations according to claim 1, characterized in that, In step S2, the parallel numerical simulation of various deep foundation pit excavation methods, construction process parameters of support structures, and layout and parameters of internal support systems, and the generation of recommended deep foundation pit excavation and support construction schemes for stations through optimization analysis, specifically involves using a multi-objective evolutionary algorithm to perform parallel numerical simulation and optimization analysis of the various deep foundation pit excavation methods, construction process parameters of support structures, and layout and parameters of internal support systems, and generating recommended deep foundation pit excavation and support construction schemes for stations.

4. The method for dynamic mechanical characteristic analysis and stability control of deep foundation pit construction for railway stations according to claim 1, characterized in that, In S4, the tolerance range set based on the current confidence level of the model is an adaptive threshold.

5. The method for dynamic mechanical characteristic analysis and stability control of deep foundation pit construction for railway stations according to claim 1, characterized in that, In S5, the specific steps of initiating targeted soil reinforcement scheme optimization are as follows: when the evaluation results show that the tunnel boring machine is about to enter or leave the pit sidewall opening area, the soil reinforcement scheme optimization program for the tunnel boring machine's entry and exit from the opening is initiated.

6. The method for dynamic mechanical characteristic analysis and stability control of deep foundation pit construction for railway stations according to claim 1, characterized in that, Between S3 and S4, there is also a method for collecting activity signals of microcracks inside the concrete of the support structure.

7. The method for dynamic mechanical characteristic analysis and stability control of deep foundation pit construction for railway stations according to claim 1, characterized in that, When the dynamic mechanical characteristics analysis and stability control method for deep foundation pit construction is applied to the deep foundation pit of a large-span T-shaped transfer station, the optimized design of the internal support system layout and parameters in S2 is particularly crucial.

8. The method for dynamic mechanical characteristic analysis and stability control of deep foundation pit construction for railway stations according to claim 1, characterized in that, After generating the dynamic adjustment instruction in S5, the instruction, the basis for adjustment, and the risk assessment conclusion are recorded in the tamper-proof distributed database in the form of a structured process log.

9. A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method for dynamic mechanical property analysis and stability control of deep foundation pit construction for railway stations as described in any one of claims 1 to 8.

10. A dynamic mechanical property analysis and stability control system for deep foundation pit construction of railway stations, used to implement the dynamic mechanical property analysis and stability control method for deep foundation pit construction of railway stations as described in any one of claims 1 to 8, the system comprising: The data fusion and initial modeling module is used to collect and fuse engineering geological survey data, hydrogeological survey data, indoor physical model test data and numerical simulation analysis data of the construction site, and establish an initial construction mechanical behavior prediction model calibrated by multi-source data. The construction scheme optimization module is used to perform parallel numerical simulations on various deep foundation pit excavation methods, support structure construction process parameters, and internal support system layout and parameters based on the initial construction mechanical behavior prediction model. Through optimization analysis, it generates recommended construction schemes for deep foundation pit excavation and support of the station. The real-time monitoring data acquisition module is used to simultaneously implement information-based construction monitoring and collect real-time mechanical monitoring data of the foundation pit soil, support structure and internal support system during the construction process according to the recommended construction plan. The dynamic feedback and model update module is used to dynamically compare the real-time mechanical monitoring data with the prediction results of the initial construction mechanical behavior prediction model corresponding to the construction stage, and trigger online model updates based on the deviation analysis results. The risk assessment and dynamic control module is used to re-simulate and assess the risks of subsequent excavation, support construction and shield tunneling entry and exit procedures using the updated model, and dynamically adjust subsequent construction parameters or initiate targeted soil reinforcement scheme optimization based on the assessment results.

Images