Construction safety rolling prediction and collaborative optimization control method and system

Abstract

The application discloses a construction safety rolling prediction and collaborative optimization control method and system, and belongs to the technical field of construction safety control. The method aims at the problems of control lag and poor collaboration of the prior art, and forms a dynamic control closed loop through the steps of establishing a construction mechanical behavior prediction model, and cyclically executing steps of sequence influence prediction and parameter optimization, optimized parameter execution and real-time monitoring, and rolling updating the model by using monitoring data in the whole construction process. The system comprises corresponding data acquisition and fusion, model management and updating, prediction and optimization, instruction execution and monitoring modules. The application realizes the prospective and self-adaptive fine control of construction safety, and improves the prediction accuracy, control timeliness and systematic management level.

Description

Technical Field

[0001] This invention relates to the field of construction safety control technology, specifically to a method for rolling prediction and collaborative optimization control of construction safety. Background Technology

[0002] In urban underground space development, subway tunnel boring or deep foundation pit excavation often needs to be carried out in complex environments adjacent to existing high-speed rail stations, main roads, large underground pipelines, bridge foundations, and other sensitive structures. Such construction can cause displacement and stress redistribution in the surrounding strata, seriously threatening the safety and normal use of nearby structures. Existing construction safety control methods mostly follow a relatively fixed design process: the construction plan and parameters are determined in the preliminary survey and design stage, and the construction phase is carried out according to the drawings, supplemented by monitoring and early warning. This method separates the prediction, decision-making, and execution stages in terms of time and information, making it difficult to cope with the uncertainty and dynamic changes in geological conditions during construction. Its drawbacks include limited accuracy in predicting environmental impacts beforehand, delayed adjustments during construction, and a lack of coordinated consideration among different construction parameters. This leads to either a reliance on conservative strategies resulting in cost waste or untimely responses leading to safety accidents. Summary of the Invention

[0003] This invention proposes a rolling prediction and collaborative optimization control method for construction safety, comprising: S1, acquiring hydrogeological data of the construction area and establishing an initial construction mechanical behavior prediction model based on this data; S2, before the start of the current construction step, predicting the impact of the construction step on the surrounding environment based on the construction mechanical behavior prediction model, and optimizing and generating a set of construction control parameters with the goal of controlling the impact within a safe threshold; S3, executing the current construction step according to the optimized construction control parameters, while simultaneously monitoring and acquiring environmental response data caused by the construction in real time; S4, updating the construction mechanical behavior prediction model using the real-time monitoring data acquired in step S3, and using the updated model for the prediction and optimization of the next construction step, returning to step S2.

[0004] Furthermore, the step of establishing an initial construction mechanical behavior prediction model also includes calibrating the initial model by integrating physical model test data and numerical simulation analysis data.

[0005] Furthermore, the optimization of generating a set of construction control parameters refers to the coordinated optimization of at least two of the shield tunneling parameters, grouting parameters, and foundation pit support parameters to simultaneously meet the control requirements of multiple monitoring targets.

[0006] Furthermore, updating the construction mechanical behavior prediction model specifically includes: comparing real-time monitoring data with model prediction values; if the deviation exceeds a preset tolerance, then correcting the soil mechanical parameters or boundary conditions in the model through back analysis.

[0007] Furthermore, the working conditions of the construction area are that the shield tunnel passes under or side through operating railway facilities, roads, underground culverts or building foundations; the construction mechanical behavior prediction model is a three-dimensional finite element model that considers the interaction between the shield, soil and structure; the construction control parameters include soil pressure, advance speed, grouting pressure and grouting volume.

[0008] Furthermore, the working conditions of the construction area are deep foundation pits or pit-within-pit excavations with multi-layer internal supports; the construction mechanical behavior prediction model is a three-dimensional finite element model that can simulate the sequence of earthwork excavation and support erection; the construction control parameters include the earthwork excavation dimensions of each layer and the pre-applied axial force of the supports.

[0009] Furthermore, the construction safety rolling prediction and collaborative optimization control method also includes: establishing a rapid prediction model based on simplified analytical theory; in step S2, the rapid prediction model is first used for preliminary screening, and then the construction mechanical behavior prediction model is used for fine optimization.

[0010] Furthermore, the construction safety rolling prediction and collaborative optimization control method also includes: automatically converting the optimized construction control parameters into control commands for construction equipment and issuing them; monitoring the deviation between the equipment execution parameters and the optimized values ​​in real time, and issuing an early warning if the deviation exceeds the limit.

[0011] Furthermore, the present invention also proposes a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the aforementioned construction safety rolling prediction and collaborative optimization control method.

[0012] Furthermore, this invention proposes a rolling prediction and collaborative optimization control system for subway construction safety, comprising: a data acquisition and fusion module for acquiring hydrogeological data, physical model test data, numerical simulation data, and real-time monitoring data at the construction site; a model management and update module for constructing, calibrating, and rollingly updating a construction mechanical behavior prediction model based on real-time monitoring data; a prediction and optimization module for predicting the environmental impact of the next construction step based on the updated model and generating a construction control parameter set using a multi-objective control optimization algorithm; and an instruction execution and monitoring module for converting the construction control parameter set into equipment control instructions, monitoring the execution process, and feeding back execution deviation information to the model management and update module.

[0013] This invention proposes a rolling prediction and collaborative optimization control method for construction safety, which transforms one-time static prediction into continuously updated dynamic prediction and cyclically executes steps of predicting the impact of steps and optimizing parameters, executing optimized parameters and real-time monitoring, and using monitoring data to continuously update the model to form a dynamic control closed loop. This significantly improves the cumulative control accuracy and timeliness of the long-term construction process and uncertain factors, and effectively alleviates the shortcomings of existing technologies such as control lag and poor coordination. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating a construction safety rolling prediction and collaborative optimization control method proposed in this invention. Figure 2 A schematic diagram of a three-dimensional finite element model mesh for a shield tunnel passing under an operating high-speed railway station; Figure 3 This is a graph comparing the deviation between instruction execution and monitoring. Detailed Implementation

[0015] This invention proposes a rolling prediction and collaborative optimization control method for construction safety, comprising: S1. Obtain hydrogeological data of the construction area and establish an initial construction mechanical behavior prediction model based on this data.

[0016] The hydrogeological data includes soil layer distribution and physical and mechanical parameters of each soil layer. The initial construction mechanical behavior prediction model is a numerical model established using finite element software based on the hydrogeological data to simulate construction mechanical behavior. Its technical functional boundary is to predict the redistribution of ground stress and deformation of surrounding structures caused by the entire construction process. Its dynamic feature is that the parameters are updated through back analysis as the construction progresses.

[0017] Specifically, hydrogeological data of the construction area are obtained, and based on the hydrogeological data, an initial three-dimensional coupled numerical model of the shield tunnel soil high-speed railway station foundation is established using finite element software as an initial construction mechanical behavior prediction model.

[0018] In a specific implementation scenario, where a shield tunnel passes under an operating high-speed railway station, hydrogeological data of the construction area is acquired, and a system is established based on this data. Figure 2 The three-dimensional finite element model shown considers the interaction between the tunnel boring machine, the soil, and the structure. The model realistically reproduces the reinforced concrete structure of the high-speed railway station's pile foundation and the large rainwater culvert.

[0019] S2. Before the start of the current construction step, based on the construction mechanical behavior prediction model, predict the impact of the construction step on the surrounding environment, and optimize and generate a set of construction control parameters with the goal of controlling the impact within a safe threshold.

[0020] The optimization of a set of construction control parameters refers to the coordinated optimization of at least two of the following parameters: shield tunneling parameters, grouting parameters, and foundation pit support parameters, to simultaneously meet the control requirements of multiple monitoring targets. The source of these construction control parameters is the search results of the multi-objective optimization algorithm based on the construction mechanical behavior prediction model.

[0021] Specifically, before the tunnel boring machine (TBM) begins excavation, the initial construction mechanical behavior prediction model is run to simulate the impact of the planned number of tunneling rings on the settlement of the high-speed railway station track slab and predict the settlement value. In order to control the settlement within a safe threshold, a multi-objective optimization algorithm is invoked. The multi-objective optimization algorithm attempts to adjust the combination of soil pressure, advance speed, and synchronous grouting pressure parameters for repeated simulation calculations, and finally optimizes and generates a set of construction control parameters, including specific soil pressure values, advance speed values, and synchronous grouting pressure values.

[0022] In a specific implementation scenario, where a shield tunnel passes under an operating high-speed railway station, the control objectives include keeping the settlement of the railway track slab within a safe threshold and keeping the longitudinal stress of the tunnel segment lining within an allowable range. The multi-objective optimization algorithm employs a multi-objective genetic algorithm, using four parameters—soil pressure, propulsion speed, articulation angle, and grouting pressure—as decision variables. In each iteration, the algorithm calls the construction mechanics behavior prediction model to calculate a set of track slab settlement and segment stress values ​​corresponding to these parameters. The algorithm uses the weighted sum of the two objectives—minimizing track slab settlement and segment stress—as its fitness function, searching within the entire feasible parameter domain. The final output set of recommended parameters represents a near-optimal solution that minimizes segment stress while satisfying settlement constraints. The parameter control range is shown in Table 1.

[0023] Table 1. Control Range of Multi-Objective Cooperative Optimization Parameters for Shield Tunnel Underpass Conditions

[0024] S3. Execute the current construction sequence according to the optimized construction control parameters, and at the same time monitor and obtain environmental response data caused by construction in real time.

[0025] The environmental response data is a physical quantity obtained through real-time monitoring that reflects the impact of construction on the surrounding environment. Its technical functional boundary is to characterize the actual environmental deformation and stress state under the current construction sequence, and its source is an automated sensor network deployed at the construction site.

[0026] Specifically, the current tunnel boring machine advance sequence is executed according to the optimized construction control parameters, while the settlement and tilt data of the track slab are monitored in real time by an automated static level and tilt sensor deployed on the high-speed railway subgrade.

[0027] In one specific implementation scenario, the system automatically converts the optimized construction control parameters into control commands for the construction equipment and sends them to the main controller of the tunnel boring machine (TBM). Simultaneously, it reads real-time data on the actual soil pressure and propulsion speed from the TBM's sensors. The system interface continuously displays a curve comparing the optimized and actual values, such as... Figure 3 As shown, once the actual value of a certain parameter is found to continuously deviate from the optimized value and exceed a certain limit, the system will automatically pop up a warning.

[0028] S4. Using the real-time monitoring data obtained in step S3, update the construction mechanical behavior prediction model, and use the updated model for the prediction and optimization of the next construction step, then return to step S2.

[0029] The process of updating the construction mechanics behavior prediction model involves comparing real-time monitoring data with the model's predicted values. If the deviation exceeds a preset tolerance, the soil mechanics parameters or boundary conditions in the model are corrected through back-analysis. The preset tolerance is a threshold value representing the relative error between the predicted and measured values, set according to the project's risk level.

[0030] Specifically, after the construction steps are completed, the measured settlement data is compared with the settlement value predicted by the model. If a certain deviation is found, the soil elastic modulus parameter in the initial construction mechanical behavior prediction model is fine-tuned by back-analysis using the measured data to make the model prediction value closer to the measured value, and an updated construction mechanical behavior prediction model is obtained. This updated model will be used for a new round of prediction and optimization before subsequent tunneling. Then, return to step S2, and repeat this cycle until the tunnel is completed.

[0031] In one specific implementation scenario, the preset model prediction tolerance is that the relative error between the predicted value and the measured value must not exceed a set proportion. Specific tolerance settings are shown in Table 2. During post-construction comparison, the deviation is automatically calculated. If the deviation is within the tolerance, the model is considered valid and can be used directly. If the deviation exceeds the tolerance, an update is triggered. The update is achieved through parameter inversion. Assuming the deviation mainly stems from improper values ​​of mechanical parameters in a certain soil layer, the inverse analysis program is automatically initiated to adjust the corresponding mechanical parameters in the model using the measured settlement data as the target. Through iteration, the model output value is made to approximate the measured value.

[0032] Table 2 Comparison of Prediction Tolerance and Back Analysis Parameters for Deep Foundation Pit Excavation Model

[0033] Furthermore, the step of establishing an initial construction mechanical behavior prediction model also includes calibrating the initial model by integrating physical model test data and numerical simulation analysis data.

[0034] The calibration of the initial model by integrating physical model test data and numerical simulation analysis data refers to comparing the settlement curves of key points measured by centrifuge tests and the results of fine numerical simulation with the prediction results of the initial model, and adjusting the soil constitutive relation parameters in the model to make the predicted trend and magnitude match the test and verification simulation results.

[0035] Specifically, in the early stages of the project, a scaled-down centrifuge model test was conducted for the underpass of the high-speed railway station to obtain experimental data on ground deformation caused by tunnel boring. Simultaneously, a refined numerical model was established for verification based on the same geological conditions. After establishing the initial construction mechanical behavior prediction model, the settlement curves of key points measured by the centrifuge test and the results of the refined numerical simulation were compared with the prediction results of the initial model. By adjusting the soil constitutive relation parameters in the initial model, the prediction trend and magnitude of the initial model were made to basically match the experimental and verification simulation results, thus completing model calibration. The calibrated model has a higher confidence level in predicting the deformation of the high-speed railway station caused by the underpass construction.

[0036] Furthermore, the optimization of generating a set of construction control parameters refers to the coordinated optimization of at least two of the shield tunneling parameters, grouting parameters, and foundation pit support parameters to simultaneously meet the control requirements of multiple monitoring targets.

[0037] The aforementioned collaborative optimization is a process in which a multi-objective optimization algorithm searches for parameter combinations that simultaneously meet the control requirements of multiple monitoring objectives within the parameter feasible domain, thereby changing the phenomenon of mutual constraints between parameters caused by independent control of a single parameter.

[0038] Specifically, in this embodiment, the control objectives include controlling the settlement of the high-speed railway station track slab within a safe threshold and controlling the longitudinal stress of the tunnel segment lining within an allowable range. The multi-objective optimization algorithm adopts a multi-objective genetic algorithm, and simultaneously uses four parameters—soil pressure, advance speed, articulation angle, and grouting pressure—as decision variables. In each iteration, the algorithm calls the construction mechanics behavior prediction model to calculate the track slab settlement value and segment stress value corresponding to a set of parameters. The algorithm uses the weighted sum of the two objectives—minimizing the track slab settlement and segment stress—as the fitness function, and searches within the entire parameter feasible region. The final output set of recommended parameters is the optimal solution that minimizes the segment stress while satisfying the settlement constraint.

[0039] Furthermore, updating the construction mechanical behavior prediction model specifically includes: comparing real-time monitoring data with model prediction values; if the deviation exceeds a preset tolerance, then correcting the soil mechanical parameters or boundary conditions in the model through back analysis.

[0040] The preset tolerance is a threshold value representing the relative error between the predicted and measured values, set according to the engineering risk level. The back analysis is an iterative process of adjusting the corresponding mechanical parameters in the model using measured settlement data as the target, so that the model output value approximates the measured value.

[0041] Specifically, the preset model prediction tolerance is that the relative error between the predicted and measured values ​​must not exceed a certain proportion. During post-construction comparison, the deviation is automatically calculated. If the deviation is within the tolerance, the model is considered valid and can be used directly. If the deviation exceeds the tolerance, an update is triggered. The update is achieved through parameter inversion. Assuming the deviation mainly stems from improper values ​​of mechanical parameters in a certain soil layer, an automatic inversion analysis program is initiated to adjust the corresponding mechanical parameters in the model using measured settlement data as the target. Through iteration, the model output value is brought closer to the measured value. The model prediction tolerance can be set according to the project risk level. For particularly important structures such as high-speed railway stations, a stricter tolerance ratio is set, while for general pipelines, a relatively lenient ratio can be set.

[0042] Furthermore, the working condition is that the shield tunnel passes under or side through operating railway facilities, roads, underground culverts or building foundations; the construction mechanics behavior prediction model is a three-dimensional finite element model that considers the interaction between the shield, soil and structure; the construction control parameters include soil pressure, advance speed, grouting pressure and grouting volume.

[0043] The three-dimensional finite element model that considers the interaction between the tunnel boring machine, the soil, and the structure is a mechanical simulation model that realistically restores the structural form and stress state of the adjacent structure in numerical simulation. Its technical functional boundary is to quantitatively predict the ground displacement transmission and additional structural response caused by ultra-close distance construction.

[0044] Specifically, the construction mechanics behavior prediction model is a three-dimensional finite element model considering the interaction between the tunnel boring machine (TBM), the soil, and the structure. The model realistically recreates the reinforced concrete structures of the high-speed railway station's pile foundation and large stormwater culvert. The construction control parameters include soil chamber pressure, advance speed, grouting pressure, and grouting volume. Traditional close-proximity TBM construction relies on expert experience and conservative strategies such as generally increasing soil chamber pressure and reducing advance speed, resulting in low efficiency and difficulty in controlling safety margins. This invention, through model-based quantitative prediction and parameter co-optimization, transforms experience-based decision-making into data-driven scientific decision-making. It finds more economical combinations of construction parameters while ensuring safety. This shift from qualitative experience-based decision-making to quantitative data-driven decision-making enables quantitative control and proactive intervention of risks in ultra-close-proximity construction, leading to a dual improvement in control accuracy and engineering economy. The overall technical solution is not readily apparent to those skilled in the art.

[0045] Furthermore, the working condition is the excavation of a deep foundation pit or pit within a pit with multiple internal supports; the construction mechanical behavior prediction model is a three-dimensional finite element model that can simulate the sequence of earthwork excavation and support erection; the construction control parameters include the earthwork excavation dimensions of each layer and the pre-applied axial force of the supports.

[0046] The three-dimensional finite element model that can simulate the sequence of earthwork excavation and support erection is a numerical model that reproduces the sequential effects of layered and block excavation of the foundation pit and support installation. Its technical functional boundary is to predict the dynamic changes of the displacement of the foundation pit sidewall and the axial force of the support.

[0047] Specifically, taking a substation interchange pit-within-a-pit project as an example, the construction mechanics behavior prediction model is a three-dimensional finite element model capable of simulating the entire process of layered and segmented excavation of the foundation pit, as well as the installation of steel supports and the application of axial force. The construction control parameters include the excavation dimensions of each layer and the axial force applied to the supports. Before excavation, based on the model updated from previous monitoring data, the changes in horizontal displacement of the pit sidewalls and axial force of the supports that may be caused by the next excavation are predicted. The optimization algorithm aims to control the displacement and axial force within the allowable range and to balance the stress on each support as much as possible, collaboratively optimizing the next excavation dimensions and the pre-stressing force of the supports. This method overcomes the drawback of traditional deep foundation pit construction where the support scheme is difficult to dynamically adjust once determined. By actively guiding the stress state of the foundation pit through rolling prediction and optimization, it avoids local stress concentration caused by improper excavation sequence, thereby significantly improving the overall stability and construction safety of deep foundation pits, especially complex pit-within-a-pit structures.

[0048] Furthermore, the construction safety rolling prediction and collaborative optimization control method also includes: establishing a rapid prediction model based on simplified analytical theory; in step S2, the rapid prediction model is first used for preliminary screening, and then the construction mechanical behavior prediction model is used for fine optimization.

[0049] The rapid prediction model based on simplified analytical theory is an analytical model that is established and calibrated based on stochastic medium theory or energy variational method and can quickly estimate the surface settlement trough caused by tunnel excavation. Its technical functional boundary is a rapid coarse screening of a wide range of candidate parameters, which are derived from analytical theory derivation and field data calibration.

[0050] Specifically, the rapid prediction model is based on stochastic medium theory or energy variational method, and after calibration, it can quickly estimate the surface settlement trough caused by tunnel excavation. In the rolling control process, step S2 is refined into two steps: first, a rapid analytical model is used to quickly screen a large number of candidate parameter combinations, eliminating parameter groups that will obviously lead to excessive settlement; then, a fine three-dimensional finite element model is used to accurately simulate and compare the remaining few preferred parameter groups to determine the final recommended parameters. This two-level model system of coarse screening and fine evaluation significantly reduces the computational load of optimization search without significantly sacrificing the accuracy of the final decision, resulting in a shorter rolling control cycle and a more agile system response. In emergency situations such as sudden water inrush or equipment failure, traditional methods often force work stoppages. The two-level model system of this invention enables the system to complete the evaluation of new working conditions and parameter re-optimization in a short time, providing strong technical support for rapid on-site decision-making. This agility in dealing with uncertainty is not available in traditional static design methods.

[0051] Furthermore, the construction safety rolling prediction and collaborative optimization control method also includes: automatically converting the optimized construction control parameters into control commands for construction equipment and issuing them; monitoring the deviation between the equipment execution parameters and the optimized values ​​in real time, and issuing an early warning if the deviation exceeds the limit.

[0052] The automatic conversion of optimized construction control parameters into control commands for construction equipment and their subsequent distribution is an automated execution process whereby the system automatically converts parameter values ​​into a set of control commands recognizable by the tunnel boring machine and distributes them to the main controller via the industrial network. The issuance of early warnings for exceeding deviation limits is a safety monitoring mechanism that automatically issues tiered alarms when the system reads data from equipment sensors in real time and compares it with optimized values.

[0053] Specifically, the system adds an instruction execution and monitoring module. After the prediction and optimization module outputs a set of recommended construction parameters, this module automatically converts these parameter values ​​into a control instruction set recognizable by the tunnel boring machine (TBM) and sends them to the TBM's main controller via the industrial network. Simultaneously, it reads real-time data on the actual soil pressure and propulsion speed from the TBM's sensors. The system interface continuously displays a curve comparing the optimized and actual values. If the actual value of a parameter deviates continuously from the optimized value and exceeds a certain limit, the system automatically issues a warning. If the deviation further widens, a higher-level alarm is issued, and a suspension of tunneling is recommended. All deviation data is recorded and fed back to the model management and update module as one of the inputs for model updates. Through this series of automated links, the potential for errors and delays caused by manual parameter setting is greatly reduced, ensuring that the optimization strategy is faithfully and promptly executed. This truly realizes the theoretical control closed loop and improves the execution accuracy and process stability of the entire construction safety rolling prediction and collaborative optimization control method system.

[0054] Furthermore, the present invention also proposes a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the aforementioned construction safety rolling prediction and collaborative optimization control method.

[0055] Furthermore, this invention also proposes a rolling prediction and collaborative optimization control system for subway construction safety, comprising: The data acquisition and fusion module is used to acquire hydrological and engineering geological data, physical model test data, numerical simulation data, and real-time monitoring data at the construction site. The data acquisition and fusion module is responsible for accessing various data sources and performing spatiotemporal alignment and formatting. Its technical function boundary is the aggregation and cleaning of multi-source heterogeneous data, and its dynamic feature is the continuous reception of monitoring stream data as the construction progresses.

[0056] The model management and update module is used to build, calibrate, and continuously update the construction mechanics behavior prediction model based on real-time monitoring data. The model management and update module maintains the life cycle of the construction mechanics behavior prediction model and provides functions for model creation, calibration, version saving, and parameter inversion and update. The technical function boundary is the evolution and version control of the numerical model.

[0057] The prediction and optimization module is used to predict the environmental impact of the next construction step based on the updated model, and to generate a set of construction control parameters using a multi-objective control optimization algorithm. The prediction and optimization module calls the model to perform simulation and runs the optimization algorithm to find the best parameters. The technical functional boundary is to find the optimal parameter combination based on the current model state.

[0058] The instruction execution and monitoring module is used to convert the construction control parameter set into equipment control instructions, monitor the execution process, and feed back execution deviation information to the model management and update module. Specifically, the instruction execution and monitoring module is responsible for communicating with the construction equipment and monitoring its execution status; its technical functional boundary is closed-loop monitoring of instruction issuance and execution deviation.

[0059] Specifically, the data acquisition and fusion module is responsible for accessing various data sources, including geological database interfaces, experimental data files, and monitoring sensor network data streams, and performing spatiotemporal alignment and formatting. The model management and update module maintains the lifecycle of the construction mechanics behavior prediction model, providing functions for model creation, calibration, version saving, and parameter inversion and updating. The prediction and optimization module calls the model for simulation and runs optimization algorithms to find the optimal parameters. The instruction execution and monitoring module is responsible for communicating with construction equipment and monitoring its execution status. These four modules are tightly coupled through an internal data bus, working collaboratively to form the hardware and software foundation for the described construction safety rolling prediction and collaborative optimization control method. This system can be deployed at the project site command center as the core of the intelligent construction safety management and control platform.

[0060] 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 rolling prediction and collaborative optimization control of construction safety, characterized in that, include: S1. Obtain hydrogeological data of the construction area and establish an initial construction mechanical behavior prediction model based on it. The initial construction mechanical behavior prediction model is a numerical model established using finite element software based on the hydrogeological data to simulate construction mechanical behavior. Its technical functional boundary is to predict the redistribution of ground stress and deformation of surrounding structures caused by the entire construction process. The establishment of the initial construction mechanical behavior prediction model also includes calibrating the initial model by integrating physical model test data and numerical simulation analysis data. That is, comparing the settlement curves of key points measured by centrifuge test and the results of fine numerical simulation with the prediction results of the initial model, and adjusting the soil constitutive relation parameters in the model to make the predicted trend and magnitude match the test and verification simulation results. S2. Before the start of the current construction sequence, based on the construction mechanical behavior prediction model, predict the impact of the construction sequence on the surrounding environment, and optimize and generate a set of construction control parameters with the goal of controlling the impact within a safe threshold. The optimization and generation of a set of construction control parameters refers to the coordinated optimization of at least two of the shield tunneling parameters, grouting parameters, and foundation pit support parameters to simultaneously meet the control requirements of multiple monitoring targets. The construction safety rolling prediction and collaborative optimization control method further includes: establishing a rapid prediction model based on simplified analytical theory; in S2, the rapid prediction model is first used for preliminary screening, and then the construction mechanical behavior prediction model is used for fine optimization; wherein, the rapid prediction model based on simplified analytical theory is an analytical model that is established based on stochastic medium theory or energy variational method and calibrated to quickly estimate the surface settlement trough caused by tunnel excavation, and the technical functional boundary is a rapid coarse screening of a large range of candidate parameters, which is derived from analytical theory derivation and field data calibration; S3. Execute the current construction sequence according to the optimized construction control parameters, and simultaneously monitor and acquire environmental response data caused by construction in real time. The environmental response data is a physical quantity that reflects the impact of construction on the surrounding environment. The construction safety rolling prediction and collaborative optimization control method further includes: automatically converting the optimized construction control parameters into control commands for construction equipment and issuing them; monitoring the deviation between the equipment execution parameters and the optimized values ​​in real time, and issuing an early warning if the deviation exceeds the limit. S4. Using the real-time monitoring data obtained in step S3, update the construction mechanical behavior prediction model, and use the updated model for prediction and optimization of the next construction step. Return to step S2. The update of the construction mechanical behavior prediction model specifically includes: comparing the real-time monitoring data with the model prediction value. If the deviation exceeds the preset tolerance, the soil mechanical parameters or boundary conditions in the model are corrected through back analysis. The preset tolerance is a relative error ratio threshold between the predicted value and the measured value set according to the engineering risk level. The back analysis is an iterative process of adjusting the corresponding mechanical parameters in the model in reverse with the measured settlement data as the target so that the model output value approaches the measured value.

2. The construction safety rolling prediction and collaborative optimization control method according to any one of claims 1, characterized in that, The working conditions of the construction area are that the shield tunnel passes under or side through operating railway facilities, roads, underground culverts or building foundations; the construction mechanical behavior prediction model is a three-dimensional finite element model that considers the interaction between the shield, soil and structure; the construction control parameters include soil pressure, advance speed, grouting pressure and grouting volume.

3. The construction safety rolling prediction and collaborative optimization control method according to any one of claims 1, characterized in that, The working conditions of the construction area are deep foundation pits or pit-within-pit excavation with multi-layer internal supports; the construction mechanical behavior prediction model is a three-dimensional finite element model that can simulate the sequence of earthwork excavation and support erection; the construction control parameters include the earthwork excavation dimensions of each layer and the pre-applied axial force of the supports.

4. A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the construction safety rolling prediction and collaborative optimization control method as described in any one of claims 1 to 3.

Images