Prestress tension parameter optimization method and system for prefabricated section of steel-concrete composite beam bridge

Abstract

The invention relates to the technical field of parameter optimization, and discloses a steel-concrete composite beam bridge prefabricated segment prestress tension parameter optimization method and system, and the method comprises the steps: obtaining a design BIM model, classifying prefabricated segments, constructing an initial digital twinborn model, and obtaining monitoring data; processing the monitoring data to obtain corrected data; comparing the corrected data with a model simulation result, and calibrating the model to generate a precise digital twinborn model; evaluating a multi-dimensional robustness index to construct a feature vector; generating a self-adaptive tensioning strategy sequence through a graph neural network in combination with the feature vector and the precise model; according to the method, intelligent optimization of the tensioning parameters is achieved, and the construction precision and the structural safety of the steel-concrete composite beam bridge are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of parameter optimization, and more particularly to a method and system for optimizing prestressed tensioning parameters of prefabricated segments of a steel-concrete composite beam bridge. Background Art

[0002] During the prestressing and tensioning construction process of prefabricated segments of steel-concrete composite girder bridges, the optimization of prestressing and tensioning parameters is crucial to ensuring the accuracy of bridge construction and structural safety. Chinese patent application with publication number CN114201796A discloses a method for accurately calculating prestress loss that takes into account the influence of tensioning sequence. This method calculates the actual tensioning stress value after the loss of prestressed tendons by constructing a tensioning stress matrix and a prestressing loss calculation matrix. Chinese patent application with publication number CN102941623A provides a method for controlling the tensioning parameters of a box girder prestressing and tensioning system. Through equipment such as a controller, a pressure sensor, an oil pump truck, an oil-pressure through-hole jack, and a displacement sensor, real-time monitoring and adjustment of the tensioning process are achieved.

[0003] However, existing technologies have deficiencies in constructing temporal dependencies and spatial coupling association graphs between tensioning control units. Traditional methods find it difficult to identify implicit stress influences across segments through intelligent models, such as the indirect effect of tensioning of Class A segments on the preload of Class C nodes, which results in tensioning strategy generation being limited to a single segment or fixed process and unable to adapt to complex stress scenarios with dynamic coupling between segments. When Class A segments are tensioned, their stress transfer may cause changes in the preload of Class C nodes. If an effective association graph is not established, these changes cannot be accurately identified and quantified, which will form a blind spot for tensioning strategy optimization. In addition, the decision-making model and the digital twin model in the existing technology are not well coordinated, making it difficult to fully utilize the simulation capabilities of the precise digital twin model and the autonomous learning capabilities of the graph neural network-reinforcement learning, resulting in difficulty in optimizing tensioning parameters based on real-time robustness status and dynamic adjustment. This causes deviations between strategy execution and the actual stress state of the structure, affecting construction accuracy and structural safety. Summary of the Invention

[0004] To overcome the aforementioned shortcomings of the prior art, the present invention provides a method and system for optimizing the prestressing parameters of precast segments of steel-concrete composite beam bridges. This system constructs a complete process integrating segment classification, data correction, model calibration, and strategy generation. By constructing a temporal dependency and spatial coupling association diagram between tensioning control units, and combining a precise digital twin model with graph neural network-reinforcement learning technology, dynamic optimization of tensioning parameters is achieved. This method can accurately identify the hidden force influences across segments, significantly improving the accuracy and efficiency of prestressing, effectively ensuring the construction accuracy and structural safety of steel-concrete composite beam bridges, and achieving intelligent and precise full-process optimization from data acquisition to strategy execution.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] The optimization method for prestressing parameters of precast segments of steel-concrete composite beam bridges includes:

[0007] Obtain the design BIM model, classify the prefabricated segments based on the design BIM model, build the initial digital twin model of the prefabricated segments based on the classification results, and deploy monitoring points to obtain the original monitoring data stream;

[0008] Screening, verifying and correcting the original monitoring data stream to obtain a corrected original monitoring data stream;

[0009] Compare the corrected original monitoring data stream with the simulation results of the initial digital twin model of the prefabricated segment to determine whether a calibration signal is triggered. If so, perform dynamic calibration of the initial digital twin model to generate an accurate digital twin model of the prefabricated segment.

[0010] Based on the modified original monitoring data stream, a multi-dimensional robustness index evaluation is performed to construct a robustness index feature vector;

[0011] Based on the robustness index eigenvector, a tensioning control unit is constructed, and a control unit association diagram is established. Combined with the precise digital twin model of the prefabricated segment, a graph neural network tensioning decision model is constructed and trained to generate an adaptive tensioning strategy sequence.

[0012] Furthermore, the classification of prefabricated segments based on the design BIM model includes:

[0013] The structural parameters of the steel-concrete composite beam bridge are extracted from the designed BIM model; according to the structural parameters, the precast segments are divided into three types of segments, including Class A segments, i.e., negative bending moment zone segments, Class B segments, i.e., mid-span segments, and Class C segments, i.e., connection node segments.

[0014] Furthermore, the construction of the initial digital twin model of the prefabricated segment includes:

[0015] Based on the classification results of the three types of segments, the detailed structural parameters of each type of segment are extracted to generate segment classification identification data; based on the segment classification identification data, the initial digital twin model of the prefabricated segment is constructed on the basis of the designed BIM model.

[0016] Furthermore, the method for obtaining the corrected original monitoring data stream includes:

[0017] Perform preliminary screening on the original monitoring data stream, remove abnormal data, and obtain the filtered original monitoring data stream;

[0018] Perform adjacent comparison and secondary verification on the filtered original monitoring data stream to obtain the reliability strain data and reliability displacement data of Class A segments, the reliability strain data and reliability displacement data of Class B segments, and the reliability temperature data and reliability preload data of Class C segments;

[0019] According to the reliability temperature data of the Class C segment, the reliability strain data and reliability displacement data of the Class A segment, the reliability strain data and reliability displacement data of the Class B segment, and the reliability preload data of the Class C segment are corrected for temperature effects to obtain a corrected original monitoring data stream; the corrected original monitoring data stream includes the corrected strain data and displacement data of the Class A segment, the corrected strain data and displacement data of the Class B segment, and the corrected preload data of the Class C segment.

[0020] Furthermore, the method for determining whether to trigger a calibration signal includes:

[0021] The corrected original monitoring data stream is compared with the initial digital twin model simulation results of the prefabricated segment to determine whether the Class A segment strain calibration signal, Class A segment displacement calibration signal, Class B segment strain calibration signal, Class B segment displacement calibration signal and Class C segment connection calibration signal are triggered. If any of the Class A segment strain calibration signal, Class A segment displacement calibration signal, Class B segment strain calibration signal, Class B segment displacement calibration signal and Class C segment connection calibration signal is triggered, the calibration signal is determined to be triggered.

[0022] Furthermore, the method for constructing a robustness indicator feature vector includes:

[0023] Based on the corrected strain and displacement data of the Class A segments in the corrected original monitoring data stream, the crack risk assessment and deformation coordination assessment in the negative bending moment area are performed to generate a comprehensive risk status record table for the Class A segments.

[0024] Based on the corrected strain and displacement data of the Class B segment in the corrected original monitoring data stream, the stress level and linear status of the mid-span area are evaluated, and a comprehensive status record table of the Class B segment is generated;

[0025] Based on the corrected preload data of the Class C segment and the reliability temperature data of the Class C segment in the corrected original monitoring data stream, the connection reliability evaluation and temperature stability evaluation are performed to generate a comprehensive status record table of the Class C segment;

[0026] Based on the comprehensive risk status record table of Class A segments, the comprehensive status record table of Class B segments, and the comprehensive status record table of Class C segments, a robustness index feature vector is constructed.

[0027] Furthermore, the method for constructing the tensioning control unit includes:

[0028] Extract prestressed steel tendon parameters from the design BIM model; assign unique segment numbers to each Class A segment, Class B segment, and Class C segment based on segment classification identification data;

[0029] According to the prestressed steel tendon parameters, each prestressed steel tendon is spatially matched with the segment number sequence to determine the segment type and segment number to which each prestressed steel tendon belongs;

[0030] Based on the preset tensioning levels, the prestressed steel strands in each segment are grouped according to the tensioning levels to form a tensioning control unit consisting of segment number and tensioning level.

[0031] Furthermore, the method for establishing a control unit association diagram includes:

[0032] Analyze the temporal dependencies of different tensioning-level control units within the same segment to form an irreversible sequence;

[0033] Analyze the spatial coupling relationship between adjacent segments and identify the structural correlation between different segments based on the three-category segment classification results;

[0034] Taking the tension control unit as the node, the temporal dependency relationship forms a directed edge, and the spatial coupling relationship forms an undirected edge to construct a control unit association graph.

[0035] Furthermore, the method for generating an adaptive tensioning strategy sequence includes:

[0036] The robustness indicator feature vector in the current state is expanded using the state of the tensioned control unit and the initial state of the untensioned control unit to obtain an expanded robustness indicator feature vector; the expanded robustness indicator feature vector is input into the trained graph neural network tensioning decision model, and the tensioning decision is output, where the tensioning decision is the next control unit to be tensioned and its tensioning force value;

[0037] A precise digital twin model is used to predict the structural response after the tensioning decision is executed, and verify whether the tensioning constraints set when training the graph neural network tensioning decision model are met; if so, the tensioning decision is retained; otherwise, the tensioning decision is regenerated, and the above process is iterated until a complete tensioning strategy sequence covering all tensioning control units is generated as an adaptive tensioning strategy sequence.

[0038] A prestressed tensioning parameter optimization system for precast segments of steel-concrete composite beam bridges is used to implement the above-mentioned prestressed tensioning parameter optimization method for precast segments of steel-concrete composite beam bridges. The system includes:

[0039] Segment classification module: used to obtain the design BIM model and classify prefabricated segments based on the design BIM model;

[0040] Data acquisition module: Based on the classification results, the initial digital twin model of the prefabricated segment is constructed and monitoring points are arranged to obtain the original monitoring data stream;

[0041] Data processing module; used to screen, verify and correct the original monitoring data stream to obtain the corrected original monitoring data stream;

[0042] Model calibration module: Compares the corrected original monitoring data stream with the simulation results of the initial digital twin model of the prefabricated segment to determine whether a calibration signal is triggered. If so, the initial digital twin model is dynamically calibrated to generate an accurate digital twin model of the prefabricated segment.

[0043] Robustness indicator evaluation module: performs multi-dimensional robustness indicator evaluation based on the modified original monitoring data stream and constructs the robustness indicator feature vector;

[0044] Strategy generation module: Based on the robustness index eigenvector, a tensioning control unit is constructed, and a control unit association diagram is established. Combined with the precise digital twin model of the prefabricated segment, a graph neural network tensioning decision model is constructed and trained to generate an adaptive tensioning strategy sequence.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] The present invention obtains the design BIM model to classify precast segments and construct an initial digital twin model, and combines the complete process of monitoring data processing, model calibration, robustness assessment and adaptive tensioning strategy generation to achieve accurate optimization of prestressed tensioning parameters of precast segments of steel-concrete composite beam bridges. Among them, the screening, verification and correction of the original monitoring data stream effectively eliminates abnormal interference, ensures that the data truly reflects the stress state of the structure, and provides a reliable basis for subsequent analysis; through the comparison and calibration of the corrected data and the initial model simulation results, the model parameters are dynamically optimized so that the generated accurate digital twin model can accurately map the actual characteristics of the structure and improve the reliability of the simulation; based on the multi-dimensional robustness index evaluation and characteristic vector construction of the corrected data, the risk status of different segments and the structural performance are quantitatively integrated to provide a clear risk orientation and quantitative basis for the formulation of tensioning strategies, avoiding the one-sidedness and subjectivity of traditional evaluations; and based on A relationship diagram between tensioning control units and control units is constructed based on the characteristic vector, and an adaptive tensioning strategy sequence is generated through the graph neural network-reinforcement learning method in combination with the precise digital twin model. This can fully consider the spatial correlation and temporal dependence of each segment, so that the tensioning operation can focus on high-risk areas while taking into account the overall force coordination of the structure. The strategy can be dynamically adjusted according to the real-time status, overcoming the limitation of the traditional fixed tensioning process that is difficult to adapt to the dynamic changes of the structure. Ultimately, it significantly improves the accuracy and efficiency of prestressed tensioning, ensures the construction accuracy and structural safety of steel-concrete composite beam bridges, and realizes the intelligence and precision of the entire process from data acquisition to strategy execution. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0048] Figure 1 This is a flow chart of the method for optimizing the prestressing parameters of precast segments of steel-concrete composite beam bridges in the present invention;

[0049] Figure 2 A flow chart showing the principles of generating an accurate digital twin model of a prefabricated segment for the present invention;

[0050] Figure 3 A flowchart of the method for performing multi-dimensional robustness index evaluation and constructing robustness index feature vectors for the present invention;

[0051] Figure 4 It is a principle schematic diagram of the control unit association diagram of the present invention;

[0052] Figure 5This is a functional module diagram of the prestressed tensioning parameter optimization system for precast segments of steel-concrete composite beam bridges in the present invention. DETAILED DESCRIPTION

[0053] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0054] Example 1

[0055] See also Figure 1 As shown, this embodiment provides a method for optimizing prestressing parameters of precast segments of steel-concrete composite beam bridges, including:

[0056] Step S10: Obtain the design BIM model, classify the prefabricated segments based on the design BIM model, build an initial digital twin model of the prefabricated segments based on the classification results, arrange monitoring points, and obtain the original monitoring data stream;

[0057] Furthermore, step S10 includes:

[0058] Step S11, obtaining a design BIM model, and extracting structural parameters of the steel-concrete composite beam bridge from the design BIM model;

[0059] Step S12: Classify the precast segments into three types of segments according to the construction parameters, including type A segments, i.e., negative moment area segments, type B segments, i.e., mid-span segments, and type C segments, i.e., connection node segments;

[0060] Step S13, based on the three-category segment type classification results, extract the detailed structural parameters of each type of segment and generate segment classification identification data; the segment classification identification data includes the detailed structural parameters of Class A segments, the detailed structural parameters of Class B segments and the detailed structural parameters of Class C segments.

[0061] Step S14: constructing an initial digital twin model of the prefabricated segment based on the segment classification identification data and the design BIM model;

[0062] Furthermore, step S14 includes:

[0063] Step S141: establishing a three-dimensional spatial model of shear stud arrangement based on the detailed structural parameters of the Class A segment in the segment classification identification data, and calculating the initial combined section stiffness parameter EIA of the Class A segment;

[0064] Step S142: establishing a coordinated force relationship model between the minor main beam and the corrugated steel plate bridge deck based on the detailed structural parameters of the Class B segment in the segment classification identification data;

[0065] Step S143: Based on the detailed structural parameters of the C-type segment in the segment classification identification data, a tooth-slot connection stiffness calculation model is established to determine the tooth-slot connection stiffness K C .

[0066] Step S144: The initial combined section stiffness parameter EIA of the A-type segment, the coordinated force relationship model of the B-type segment, and the tooth-slot connection stiffness K of the C-type segment are converted into C The integration is performed to form an initial digital twin model of the prefabricated segment.

[0067] In step S15, based on the initial digital twin model of the prefabricated segment, monitoring points are arranged at the key stress-bearing parts of different segment types to obtain the original monitoring data stream. The original monitoring data stream includes the strain data εA(t) and displacement data δA(t) of Class A segments, the strain data εB(t) and displacement data δB(t) of Class B segments, and the temperature data TC(t) and preload data FC(t) of Class C segments, where t represents time.

[0068] Specifically, a design BIM model containing the three-dimensional structural information of the entire steel-concrete composite beam bridge is obtained. The model covers data such as geometric dimensions, material properties, and component connection relationships. The structural parameters of the steel-concrete composite beam bridge are extracted through the secondary development interface of the BIM software. The structural parameters specifically include the bridge span length, the three-dimensional coordinate information of each precast segment, the cross-sectional dimensions, the material mechanical properties parameters, and the connection structure type within the precast segment. The cross-sectional dimensions include, for example, the beam height, web thickness, and flange width. The material mechanical properties parameters include, for example, the elastic modulus of steel and the compressive strength grade of concrete. The connection structure type within the precast segment includes, for example, whether it includes tooth grooves and bolt assemblies. According to the above structural parameters, the precast segments are divided into Class A, Class B, and Class C, which are negative bending moment zone segments, mid-span segments, and connection node segments, respectively. The judgment basis for Class A segments is that the distance between the center point of the segment and the pier support is less than the preset negative bending moment influence distance threshold L neg , L neg It is determined by multiplying the span length by KA. KA is a coefficient used to define the influence range of Class A segments. Specifically, it is the ratio of the distance where the negative bending moment of the segment in the negative bending moment zone is significantly affected to the span length of the bridge. It is determined based on the negative bending moment distribution characteristics of simply supported beams in structural mechanics. Negative bending moment has a significant impact near the piers, and its influence range is usually 0.15 to 0.25 times the span. Such segments require key monitoring of the bridge deck tensile stress. The judgment basis for Class B segments is that the distance from the segment center point to the mid-span position is less than the preset mid-span influence distance threshold L. mid , L midThe KB coefficient is the product of the span length and KB, a factor used to define the impact range of Category B segments. Specifically, it is the ratio of the distance at which the midspan segment is significantly affected by vertical deflection to the bridge span length. KB is determined based on the variation in deflection in the midspan region. Vertical deflection changes are most pronounced within the range of 0.2 to 0.3 times the span near the midspan, necessitating particular attention to vertical deflection. Category C segments are identified based on the inclusion of tooth-and-groove connections and bolt assemblies within the segment, requiring monitoring of the connection status. Based on the above classification results, detailed structural parameters of each type of segment are extracted to generate segment classification identification data. For Class A segments, the ratio of the moment of inertia of the steel main beam section around the neutral axis in the negative bending moment zone to the farthest fiber distance, namely the section modulus parameter WA of the steel main beam in the negative bending moment zone, and the ratio of the total cross-sectional area of ​​the longitudinal reinforcement of the bridge deck to the effective cross-sectional area of ​​the bridge deck, namely the bridge deck reinforcement ratio parameter ρA, are extracted as the detailed structural parameters of Class A segments. For Class B segments, the product of the elastic modulus of the minor beam material and the section moment of inertia, namely the minor beam bending stiffness parameter EIB, and the cross-sectional area moment of the combination of the corrugated steel sheet and concrete, namely the corrugated steel sheet bridge deck combined section parameter SB, are extracted as the detailed structural parameters of Class B segments. For Class C segments, the tooth groove geometric dimension parameters GC, including tooth groove depth, width, and slope, and the bolt preload design value parameter determined according to the bolt diameter, material strength grade, and connection load requirements, are extracted as the detailed structural parameters of Class C segments.

[0069] According to the segment classification identification data, an initial digital twin model is constructed on the basis of the designed BIM model. Specifically, for Class A segments, according to their detailed structural parameters, a three-dimensional model of shear studs is arranged according to the designed spacing and quantity in the three-dimensional modeling software. By establishing the contact relationship between the steel main beam and the concrete bridge deck, the superposition method is used to calculate the initial combined section stiffness parameter EIA, that is, the steel main beam section stiffness and the concrete bridge deck section stiffness are calculated separately, and the connection stiffness reduction coefficient determined by the ratio of the shear bearing capacity obtained based on the shear stud test to the design value is considered, and the two are superimposed to obtain EIA; for Class B segments, according to their detailed structural parameters, by defining the friction coefficient and bonding strength parameters between the minor main beam and the corrugated steel plate bridge deck determined based on material tests, a collaborative force relationship model between the two is established to reflect the common deformation characteristics; for Class C segments, according to their detailed structural parameters, based on the tooth groove contact area and bolt preload, a spring-damping model is used to simulate the tooth groove connection stiffness. , the calculation formula is ,in, is the tooth-groove contact area, is the elastic modulus of the tooth groove concrete, is the design value of bolt preload, and the tooth groove connection stiffness Contact area with tooth groove , elastic modulus of tooth groove concrete and bolt preload design value There is a positive correlation. is the friction coefficient of the tooth-groove contact surface, which is determined based on the experimental properties of the tooth-groove contact surface material and surface treatment status. is the tooth groove contact depth, is the number of bolts, is the stiffness coefficient of a single bolt, which is determined by the bolt mechanical properties test. This formula comprehensively considers the contact stiffness of the tooth groove concrete and the bolt connection stiffness. Follow and The model parameters of the above-mentioned A, B, and C segments are integrated to form an initial digital twin model containing the prefabricated segments of the entire bridge.

[0070] Based on the initial digital twin model of the precast segment, monitoring points are arranged at the key stress-bearing parts of different segment types. For Class A segments, strain monitoring points are arranged at equal intervals on the top surface of the bridge deck to obtain the strain data εA(t) of the Class A segments, and vertical displacement monitoring points are arranged at both ends of the segment to obtain the displacement data δA(t) of the Class A segments. For Class B segments, strain monitoring points are arranged at equal intervals on the lower flange of the minor main beam to obtain the strain data εB(t) of the Class B segments, and vertical displacement monitoring points are arranged at the bottom of the mid-span to obtain the displacement data δB(t) of the Class B segments. For Class C segments, temperature sensors are arranged on both sides of the tooth groove to obtain the temperature data TC(t) of the Class C segments, and preload monitoring plates are installed on the bolt heads to obtain the preload data FC(t) of the Class C segments. All monitoring data are uploaded to the digital twin system to form the original monitoring data stream.

[0071] Compared to the traditional method of roughly classifying by location, classification based on specific structural parameters makes the definition of Class A, B, and C segments more consistent with the structural stress characteristics. For example, Class A segments are divided based on a specific distance threshold from the bridge pier, ensuring accurate identification of negative bending moment zones and making the placement of monitoring points more targeted. The extraction of detailed structural parameters from the segment classification identification data provides a clear direction for the subsequent construction of the digital twin model. The WA and ρA of Class A segments are directly related to their crack resistance monitoring requirements. The EIB and SB of Class B segments directly affect the accuracy of vertical deflection calculations. The tooth groove geometry parameters GC and bolt preload design value parameters of Class C segments ensure the effectiveness of connection status monitoring. This correspondence ensures a high degree of matching of the initial digital twin model parameters with the monitoring targets, improving the model's simulation accuracy. During the construction of the initial digital twin model, differentiated modeling methods were adopted for different segment types. The combined section stiffness calculation of Class A segments took into account the actual effect of shear studs, the synergistic force model of Class B segments reflected the overall deformation characteristics of the structure, and the connection stiffness model of Class C segments took into account the combined effect of the tooth grooves and bolts. The integration of these three methods enables the model to reflect both the independent characteristics of each segment and the overall force correlation of the entire bridge, making up for the defect of traditional models that cannot adequately depict the characteristics of different parts. The layout of monitoring points is based on the key stress-bearing parts of the segment type, including bridge deck strain and end displacement monitoring of Class A segments, main beam strain and mid-span deflection monitoring of Class B segments, and temperature and preload monitoring of Class C segments, forming a targeted monitoring network. Compared with the method of evenly arranging monitoring points throughout the entire bridge, the layout of sensors in the present invention directly points to the risk points of each segment. This targeting enables the collected data to directly reflect the stress state of key parts of the structure, avoiding interference of non-critical data in the analysis process, and at the same time focusing more on areas that require key monitoring in resource allocation, which not only reduces the overall cost and data transmission load of the monitoring system, but also ensures that key parameters that play a decisive role in structural safety can be accurately captured. The accuracy of segment classification provides a clear direction for the extraction of detailed structural parameters, and the detailed parameters lay the foundation for the precise construction of the digital twin model. The accuracy of the model guides the optimal layout of monitoring points, and the targeted monitoring data can feed back into the subsequent calibration of the model. This closed-loop relationship makes the entire process, from data acquisition to model construction to monitoring implementation, an organic whole. It not only realizes the effective monitoring of key parts of prefabricated segments of steel-concrete composite beam bridges, but also provides an accurate data foundation for the subsequent optimization of prestressed tensioning parameters. This is because the monitoring data of different segments can directly reflect their sensitivity to tensioning, providing a basis for the formulation of differentiated tensioning strategies.

[0072] Step S20, screening, verifying and correcting the original monitoring data stream to obtain a corrected original monitoring data stream;

[0073] Furthermore, step S20 includes:

[0074] Step S21, preliminarily screening the original monitoring data stream to remove abnormal data to obtain a screened original monitoring data stream; the screened original monitoring data stream includes strain data and displacement data after screening of Class A segments, strain data and displacement data after screening of Class B segments, and temperature data and preload data after screening of Class C segments;

[0075] Step S22: performing adjacent comparison and secondary verification on the filtered original monitoring data stream to obtain reliability strain data and reliability displacement data of the Class A segment, reliability strain data and reliability displacement data of the Class B segment, and reliability temperature data and reliability preload data of the Class C segment;

[0076] Furthermore, step S22 includes:

[0077] Step S221: perform strain verification and displacement verification on the strain data and displacement data of the Class A segments after screening, and obtain the reliability strain data ε of the Class A segments. reli A(t) and reliability displacement data δε reli A(t);

[0078] Step S222: perform strain verification and displacement verification on the strain data and displacement data of the Class B segment after screening, and obtain the reliability strain data ε of the Class B segment. reli B(t) and reliability displacement data δε reli B(t);

[0079] Step S223: perform temperature verification and preload verification on the filtered temperature data and preload data of the Class C segments, respectively, to obtain the reliability temperature data T of the Class C segments. reli C(t) and reliability preload data F reli C(t);

[0080] Step S23, based on the reliability temperature data of the Class C segment, the reliability strain data and reliability displacement data of the Class A segment, the reliability strain data and reliability displacement data of the Class B segment, and the reliability preload data of the Class C segment are corrected for temperature effects to obtain a corrected original monitoring data stream; the corrected original monitoring data stream includes the corrected strain data and displacement data of the Class A segment, the corrected strain data and displacement data of the Class B segment, and the corrected preload data of the Class C segment.

[0081] Specifically, the raw monitoring data stream is first screened to remove anomalous data. Anomalous data refers to monitoring values ​​that significantly deviate from the normal stress range or physical laws of the structure. Examples include sudden changes due to loose sensor cables, high-frequency oscillations caused by electromagnetic interference, or saturation values ​​exceeding the sensor's range. Anomalous data identification utilizes a statistically-based approach. The mean and standard deviation of the data from the same monitoring point over multiple consecutive acquisition cycles are calculated. Values ​​that deviate from the mean by more than three standard deviations are considered anomalous. Step-like sudden changes in the time series, such as when the difference between two consecutive acquisition cycles is greater than twice the historical maximum fluctuation value for that monitoring point, are also considered anomalous. After preliminary screening, a filtered raw monitoring data stream is obtained, containing strain and displacement data for Class A segments, strain and displacement data for Class B segments, and temperature and preload data for Class C segments. Based on this, adjacent comparisons and secondary verification are performed on the filtered raw monitoring data streams. Strain verification and displacement verification are performed on the filtered strain and displacement data for Class A segments, respectively. The strain difference Δε is calculated for the two sides of the same section on the top surface of the bridge deck. sec A, strain difference Δε sec A is the difference between the strain sensor monitoring values ​​at the left and right edges of the same cross section on the top surface of the bridge deck; if Δε sec A is greater than the strain difference threshold ε thre A, will be used to calculate the strain difference Δε sec A set of strain data is marked as needing verification; the displacement difference Δδ at both ends of the segment is calculated. ends A, Δδ ends A is the difference between the displacement sensor monitoring values ​​at both ends of the A-type segment; Δδ ends If A is greater than the displacement difference threshold δ thre A, then it is marked as requiring verification; the strain data and displacement data of the Class A segments that are not marked as requiring verification are marked as the reliability strain data ε of the Class A segments. reli A(t) and reliability displacement data δε reli A(t); strain difference threshold ε thre A calculates the maximum asymmetric strain of this type of segment under the design load through finite element simulation, and takes 90% of it as the strain difference threshold. Displacement difference threshold δ thre A is determined based on the segment length and the maximum allowable torsion angle. The maximum allowable torsion angle is obtained from the material crack resistance test and is converted into the displacement difference threshold by multiplying the segment length and the maximum torsion angle.

[0082] The strain data and displacement data after screening of the Class B segment are respectively verified by strain verification and displacement verification: the strain gradient ε of adjacent measuring points is calculated gradB, the strain gradient is the ratio of the strain difference between two adjacent points to the distance between the two points. If the strain gradient mutation is greater than the first gradient threshold, it is marked as requiring verification; the first gradient threshold is determined by analyzing the strain distribution law of this type of segment under normal stress state, that is, statistics are performed on multiple groups of strain data under normal working conditions, and the 95% percentile value of the strain gradient is taken as the threshold; the ratio of the mid-span displacement to the theoretical value of the mid-span displacement simulated by the initial digital twin model is calculated. If the ratio is greater than the set ratio range [R min ,R max ], it is marked as needing verification, where R min is the lower limit of the ratio range, R max is the upper limit of the ratio range; the theoretical value of mid-span displacement is calculated by inputting the load parameters and material parameters of the segment into the initial digital twin model, R min and R max The error range of the initial digital twin model simulation is determined, for example, [0.85, 1.15]. The strain data and displacement data of the Class B segments that are not marked as requiring verification are marked as the reliability strain data ε of the Class B segments. reli B(t) and reliability displacement data δε reli B(t).

[0083] The temperature data and preload data after screening of the C-type segment are respectively verified by temperature verification and preload verification: the difference between the temperature sensor monitoring values ​​on both sides of the tooth groove is calculated, that is, the temperature difference ΔT slot C, if the temperature difference ΔT slot C duration T dura Greater than the temperature difference threshold T dura thre, the temperature gradient is marked as abnormal; the temperature difference threshold is determined according to the temperature stress limit of the tooth groove concrete material. The concrete tensile stress caused by the temperature gradient is calculated, and the temperature difference when the tensile stress is close to the tensile strength of the material is used as the temperature difference threshold; the preload difference ΔF of the bolts in the same group is calculated. bolt group, if the preload difference ΔF bolt group is greater than the preload difference threshold F diff thre, the preload abnormality is marked; the same group of bolts refers to symmetrically distributed bolts connecting the same tooth groove structure. The bolt preload difference is the maximum difference between the preload monitoring values ​​of the bolts in the same group. The preload difference threshold is determined based on the force uniformity requirement of the bolt connection. For example, 10% of the bolt preload design value is taken as the allowable difference range; the temperature data and preload data of the Class C segments that are not marked with temperature gradient abnormality and preload abnormality are marked as the reliability temperature data T of the Class C segments. reli C(t) and reliability preload data F reli C(t).

[0084] Step S23 corrects other data for temperature effects based on the reliability temperature data of the Class C segments. The principle of temperature correction is that the strain caused by temperature changes in a material is positively correlated with the temperature change and the material's linear expansion coefficient. For the strain data of Class A and Class B segments, the correction formula is: monitored strain value - (temperature change × material linear expansion coefficient), where the temperature change is the difference between the current monitored temperature and the initial temperature, the initial temperature being the temperature at the time of segment installation, and the material linear expansion coefficient is determined based on the material properties of steel or concrete. The monitored strain values ​​include the monitored strain values ​​for Class A segments and the monitored strain values ​​for Class B segments, and are derived from the reliability strain data for Class A segments and the reliability strain data for Class B segments. For displacement data, the temperature-induced displacement is calculated by multiplying the temperature change, segment length, and the linear expansion coefficient. The corrected displacement data is obtained by deducting the temperature-induced displacement from the monitored displacement values. The monitored displacement values ​​for Class A segments and Class B segments are derived from the reliability displacement data for Class A segments and the reliability displacement data for Class B segments. For the preload data of the Class C segment, temperature changes cause the bolts to expand and contract. The correction method is to calculate the temperature correction for the preload based on the linear expansion coefficient of the bolt material and the temperature change. This correction is then added to the monitored preload value to obtain the corrected preload data. This process results in a corrected raw monitoring data stream containing the corrected strain and displacement data for the Class A and Class B segments, as well as the corrected preload data for the Class C segment.

[0085] The preliminary screening in step S20 identifies and eliminates abnormal data through statistical characteristics, reducing the impact of sensor failure or environmental interference on subsequent analysis, and increasing the proportion of valid signals in the data. The secondary verification adopts differentiated judgment standards for the stress characteristics of different segments. The strain and displacement difference verification of Class A segments uses structural symmetry to ensure data reliability. The strain gradient and theoretical ratio verification of Class B segments combines the local change rate with the overall simulation deviation. The temperature difference and preload difference verification of Class C segments focuses on the environment and stress uniformity of the connection parts. This classification verification method makes the data reliability judgment more in line with the physical nature of each segment. Compared with the unified verification standard, the data misjudgment rate is reduced. The temperature effect correction uses the temperature data of the Class C segment to calibrate the strain and displacement of the Class A and Class B segments, eliminating the non-stress deformation interference caused by temperature changes, making the corrected strain data closer to the actual stress state of the structure, and the displacement data more accurately reflecting the deformation caused by the load. The preliminary screening reduces the amount of invalid data processing for the secondary verification, and the secondary verification provides reliable basic data for temperature correction. The temperature correction further eliminates the systematic errors of environmental factors. The data processing chain formed by the three makes the final corrected data stream have the characteristics of low noise, high reliability and high fidelity. The temperature data of the Class C segment is not only used for its own verification, but also becomes the correction benchmark for the Class A and Class B segment data, realizing the correlation optimization of cross-segment data; through temperature correction, the monitoring data at different ambient temperatures are comparable, providing a consistent benchmark for long-term performance evaluation. Step S20 helps to improve the accuracy of subsequent digital twin model calibration, as well as the reliability of data-based risk assessment and tensioning strategy formulation.

[0086] Step S30: Compare the corrected original monitoring data stream with the simulation results of the initial digital twin model of the prefabricated segment to determine whether a calibration signal is triggered. If the calibration signal is triggered, perform dynamic calibration of the initial digital twin model to generate an accurate digital twin model of the prefabricated segment.

[0087] See also Figure 2 As shown, further, step S30 includes:

[0088] Step S31: Compare the corrected original monitoring data stream with the initial digital twin model simulation results of the prefabricated segment to determine whether the Class A segment strain calibration signal, Class A segment displacement calibration signal, Class B segment strain calibration signal, Class B segment displacement calibration signal, and Class C segment connection calibration signal are triggered;

[0089] In step S32, if any of the Class A segment strain calibration signal, Class A segment displacement calibration signal, Class B segment strain calibration signal, Class B segment displacement calibration signal, and Class C segment connection calibration signal is triggered, the trigger calibration signal is determined, and the dynamic calibration of the initial digital twin model is performed to obtain an accurate digital twin model of the prefabricated segment; otherwise, the initial digital twin model is kept unchanged.

[0090] Specifically, the specific process of comparing the corrected original monitoring data stream with the simulation results of the initial digital twin model of the prefabricated segment to determine the deviation and generate an accurate digital twin model includes: first, comparing the corrected original monitoring data stream with the simulation results of the initial digital twin model. The simulation results of the initial digital twin model refer to the initial digital twin model constructed based on step S14, and the theoretical values ​​calculated by inputting the design load, boundary conditions and material property parameters, including the strain simulation data and displacement simulation data of Class A segments, the strain simulation data and displacement simulation data of Class B segments, and the preload simulation data of Class C segments.

[0091] During the comparison process, the deviation values ​​for each segment are calculated: For Class A segments, the strain deviation is calculated by subtracting the simulated strain data from the corrected strain data; the displacement deviation is calculated by subtracting the simulated displacement data from the corrected displacement data. The strain deviation reflects the degree of deviation between the actual force and the theoretical calculation, while the displacement deviation reflects the difference between the actual deformation and the theoretical deformation. When the absolute value of the strain deviation of a Class A segment exceeds the first strain deviation threshold, a strain calibration signal for the Class A segment is triggered; when the absolute value of the displacement deviation exceeds the first displacement deviation threshold, a displacement calibration signal for the Class A segment is triggered. The determination of the first strain deviation threshold is based on the allowable force error range of the material. The ultimate tensile strain of the bridge deck material of this type of segment is obtained through material mechanics tests, and 5% of it is taken as the threshold. For example, the ultimate tensile strain of a C50 concrete bridge deck is 200με, so the first strain deviation threshold is set to 10με; the first displacement deviation threshold is determined according to the maximum deformation value allowed by the design, which is the product of the segment length and the allowable deflection limit. The allowable deflection limit is converted from the span ratio specified in the specification. The specified span ratio can be, for example, 1:500.

[0092] For Class B segments, the strain deviation is calculated by subtracting the simulated strain data from the corrected strain data. If the absolute value of the strain deviation is greater than the preset second strain deviation threshold, a strain calibration signal is triggered. The displacement deviation is calculated by subtracting the simulated displacement data from the corrected displacement data. If the absolute value of the displacement deviation is greater than the preset second displacement deviation threshold, a displacement calibration signal is triggered. The second strain deviation threshold is determined based on the fatigue strain limit of the main beam material. The allowable strain fluctuation range of the main beam of this type of segment within its design life, as determined through fatigue testing, is set at 10% as the threshold. The second displacement deviation threshold is 8% of the mid-span design deflection, which is determined through structural calculations based on the span length and load level. For Class C segments, the preload deviation is calculated by subtracting the simulated preload data from the corrected preload data. If the absolute value of the preload deviation exceeds the preset preload deviation threshold, a calibration signal for the Class C segment connection is triggered. The preload deviation threshold is determined by the bolt connection's safety factor, which is determined by the ratio of the bolt material's tensile strength to the design preload. The allowable preload deviation range is inferred from the safety factor. For example, if the design preload is 100kN, the preload deviation threshold is set to 10kN.

[0093] When any signal is triggered, a calibration signal is triggered, and dynamic calibration of the initial digital twin model is performed. If a Class A segment strain calibration signal is triggered, the initial composite section stiffness parameter EIA is adjusted. Based on the linear relationship between the strain deviation and EIA of the Class A segment, and the inverse relationship between strain and stiffness, the required EIA correction is calculated. Through iterative calculation, the deviation between the simulated strain corresponding to the corrected EIA and the reliability strain data is kept within a threshold. If a Class A segment displacement calibration signal is triggered, the number or diameter of shear studs in the three-dimensional spatial model of the shear stud arrangement is adjusted to reduce displacement deviation by increasing the connection stiffness. If a Class B segment strain calibration signal is triggered, the minor girder bending stiffness parameter EIB is corrected. Based on the inverse relationship between the strain deviation and EIB of the Class B segment, an increase in EIB results in a decrease in strain. The EIB is gradually adjusted until the simulated strain meets the requirements. If a Class B segment displacement calibration signal is triggered, the composite section parameter SB of the corrugated steel plate bridge deck is optimized. By increasing the moment of inertia of the composite section and adjusting the corrugated steel plate thickness, the mid-span deflection is reduced. For the C-type segment, if the C-type segment connection calibration signal is triggered, adjust the tooth-slot connection stiffness K C , according to the preload deviation of the C-type segment and K C The positive correlation between K C By increasing the preload force transmission efficiency, the bolt preload force design value parameter or the tooth groove geometry parameter GC can be modified to make K CThe deviation between the corresponding simulated preload force and the reliability preload force data is within the threshold. After multiple rounds of iterative calibration, until all deviations are within the corresponding threshold, an accurate digital twin model is obtained. When the calibration signal is not triggered, the initial digital twin model remains unchanged.

[0094] Step S40, performing a multi-dimensional robustness index evaluation based on the corrected original monitoring data stream to construct a robustness index feature vector;

[0095] See also Figure 3 As shown, further, step S40 includes:

[0096] Step S41, performing a crack risk assessment and deformation coordination assessment in the negative bending moment zone based on the corrected strain data and displacement data of the Class A segment in the corrected original monitoring data stream, and generating a comprehensive risk status record table for the Class A segment;

[0097] Specifically, crack risk assessment requires establishing a crack risk assessment system based on the corrected strain data for Class A segments, including low-risk and high-risk strain thresholds. The low-risk strain threshold is 50% of the ultimate tensile strain of the bridge deck concrete material, obtained through axial tensile testing of concrete by multiplying the average ultimate tensile strain of concrete specimens of the same strength grade by 50%. The high-risk strain threshold is 80% of the ultimate tensile strain, similarly calculated from test data. When the strain value corresponding to the corrected strain data for a Class A segment is less than the low-risk strain threshold, the crack risk is classified as low. When the strain value corresponding to the corrected strain data for a Class A segment is greater than or equal to the low-risk strain threshold and less than the high-risk strain threshold, the crack risk is classified as low. When the strain value corresponding to the corrected strain data for a Class A segment is greater than or equal to the high-risk strain threshold, the crack risk is classified as high. The deformation compatibility assessment is based on the corrected displacement data of Class A segments. The relative rotation angle at each end of the segment is calculated. The relative rotation angle is the ratio of the displacement difference between the two ends of the segment to the segment length. The displacement difference is the difference between the displacement sensor values ​​at both ends, and the segment length is extracted from the precise digital twin model. The rotation angle threshold is determined based on the maximum allowable torsion angle for this type of segment in the structural design code. This value is calculated through finite element simulation of the precise digital twin model to calculate the maximum allowable torsion angle of the segment under the design load. When the relative rotation angle exceeds the rotation angle threshold, the segment is marked as deformed incompatible. The crack risk level and deformation compatibility status are integrated to generate a comprehensive risk status record table for Class A segments, including the crack risk level and deformation compatibility status.

[0098] Step S42: Based on the corrected strain data and displacement data of the Class B segment in the corrected original monitoring data stream, a stress level assessment and a linear state assessment are performed in the mid-span area to generate a comprehensive state record table for the Class B segment.

[0099] Stress level assessment requires calculating the ratio of actual stress to design allowable stress. The design allowable stress is determined based on the yield strength and safety factor of the main beam material. The safety factor is selected according to bridge design specifications. For example, for Q355 steel, the yield strength is 355 MPa, the safety factor is 1.2, and the design allowable stress is 296 MPa. When the ratio of the actual stress corresponding to the corrected strain data of a Class B segment to the design allowable stress exceeds the warning ratio, a stress warning state is indicated. The warning ratio is determined based on material fatigue testing and can be set to 90%. When it exceeds 90%, the material fatigue life is significantly reduced. Alignment assessment requires calculating the alignment deviation. The alignment deviation is the difference between the corrected displacement data of a Class B segment and the design deflection value. The design deflection value is extracted from the simulation results of a precise digital twin model. The alignment deviation threshold is determined by comparing data correlating alignment deviation with structural safety across multiple sets of real bridges. It is typically set to 10% of the design deflection value. When the absolute value of the alignment deviation exceeds the alignment deviation threshold, the alignment state is indicated as requiring adjustment. The stress level assessment results and the linear state assessment results are integrated to form a Class B segment comprehensive state record table including the stress level state and linear state.

[0100] Step S43, performing connection reliability assessment and temperature stability assessment based on the corrected preload force data of the Class C segment and the reliability temperature data of the Class C segment in the corrected original monitoring data stream, and generating a Class C segment comprehensive status record table;

[0101] Connection reliability assessment requires calculation of the preload ratio. This ratio is the ratio of the preload value in the corrected preload data to the bolt preload design value, which is extracted from the precise digital twin model. The reliability threshold is determined based on bolt connection testing and is generally set to 0.9. When the preload ratio falls below the reliability threshold, the connection is marked as unreliable. Temperature stability assessment requires analysis of the spatiotemporal distribution of the temperature gradient. The temperature gradient is the ratio of the temperature difference across the tooth groove to the tooth groove contact depth. The second gradient threshold is determined by calculating the concrete temperature stress. The gradient value at which the concrete tensile stress caused by the temperature gradient approaches its tensile strength is the second gradient threshold. The duration is determined by calculating the shortest time it takes for the structure to crack after the temperature gradient exceeds the second gradient threshold. For example, in one project, microcracks will appear if the duration exceeds two hours. Therefore, the preset duration is two hours. When the temperature gradient exceeds the second gradient threshold and the duration exceeds the preset duration, the connection is marked as thermally unstable. The connection reliability and temperature stability assessment results are integrated to form a comprehensive status record for Class C segments, which includes both the connection reliability and temperature stability status.

[0102] Step S44 : constructing a robustness index feature vector based on the Class A segment comprehensive risk status record table, the Class B segment comprehensive status record table, and the Class C segment comprehensive status record table.

[0103] Specifically, the crack risk level and deformation compatibility status in the Class A segment comprehensive risk status record table are vectorized and coded. Low-risk crack risk is coded as 00, medium risk is coded as 01, and high risk is coded as 10. The deformation compatibility status is coded as 0 for compatibility and 1 for incompatibility. These two are combined to form the Class A segment characteristic vector φA. For example, low-risk and compatibility is coded as 000. The stress level and alignment status in the Class B segment comprehensive status record table are vectorized and coded. Normal stress level is coded as 0, warning stress level is coded as 1, and normal alignment status is coded as 0, adjustment required is coded as 1. These two are combined to form the Class B segment characteristic vector φB. For example, stress warning and alignment required adjustment is coded as 11. The connection reliability and temperature stability status in the Class C segment comprehensive status record table are vectorized and coded. Reliable connection is coded as 0, unreliable connection is coded as 1, and stable temperature is coded as 0, unstable temperature is coded as 1. These two are combined to form the Class C segment characteristic vector φC. For example, reliable connection and unstable temperature is coded as 01. The segment feature vector φA of category A, the segment feature vector φB of category B and the segment feature vector φC of category C are concatenated to construct the robustness index feature vector Φ=[φA,φB,φC].

[0104] In step S40, the evaluation indicators of each type of segment form a precise correspondence with the segment characteristics. The crack risk assessment of the Class A segment is directly related to the crack resistance requirements of the bridge deck in the negative bending moment area, and the deformation coordination assessment reflects the overall stress balance of the segment. The combination of the two allows the risk status of the Class A segment to be fully presented; the stress level assessment of the Class B segment focuses on the bearing capacity of the mid-span main beam, and the linear state assessment focuses on the overall linear accuracy of the structure. The two work together to achieve dual control of the stress and deformation in the mid-span area; the connection reliability assessment of the Class C segment focuses on the bolt preload, the core connection parameter, and the temperature stability assessment considers the impact of environmental factors on connection performance. The combination of the two ensures a comprehensive judgment of the node connection status. Step S40 is coordinated with the original monitoring data stream corrected in step S20. The corrected data eliminates temperature interference and the influence of outliers, making the evaluation results closer to the actual state of the structure; combined with the precise digital twin model of step S30, the design parameters provided by the precise digital twin model provide a benchmark for the evaluation, improving the accuracy of the evaluation. The construction of the robustness index characteristic vector converts various types of state information into quantifiable vector form, providing structured input data for the subsequent tensioning strategy generation in step S50, so that the adaptive tensioning strategy generation process can be directly associated with the actual risk status of the structure. Without this step, the subsequent tensioning strategy will lose its clear risk orientation, which may lead to a disconnect between the strategy and the actual needs of the structure. For example, targeted tensioning measures are not taken in high-risk crack areas, or segments with inconsistent deformation are not adjusted. Overall, step S40 achieves accurate mapping from monitoring data to structural status through multi-dimensional evaluation and feature quantification. Its collaboration with the previous data processing and model building steps forms a closed loop for the entire prestressed optimization process, which not only improves the reliability of risk assessment, but also lays a data foundation for subsequent strategy generation. At the same time, the encoding method of the characteristic vector enables different types of state information to be processed uniformly, providing the possibility for comprehensive analysis across segments.

[0105] In step S50, a tensioning control unit is constructed based on the robustness index feature vector, and a control unit association diagram is established. Combined with the precise digital twin model of the prefabricated segment, a graph neural network tensioning decision model is constructed and trained to generate an adaptive tensioning strategy sequence.

[0106] Furthermore, step S50 includes:

[0107] Step S51, extracting prestressed steel tendon parameters from the design BIM model and constructing a tensioning control unit in combination with segment classification identification data;

[0108] Specifically, the prestressed steel strand parameters are extracted from the design BIM model, including the spatial position coordinates of the prestressed steel strand, the strand number, the strand specification, and the design tension value. The spatial position coordinates of the prestressed steel strand are obtained through the three-dimensional coordinate system of the BIM model, accurate to the millimeter level. The strand number is assigned according to the numbering rules of the design drawings, such as web strands starting with W and top strands starting with T. The strand specifications include diameter and material strength grade, such as φ15.2mm steel strand corresponding to 1860MPa tensile strength. The design tension value is determined according to the strand force calculation and is specified in the design documents. For example, the design tension value of a web strand is 1500kN. Based on the segment classification identification data, a unique segment number is assigned to each Class A segment, Class B segment, and Class C segment. The numbering rule is the first letter of the segment type plus a serial number, such as A01, B03, and C02, to form a segment number sequence; according to the prestressed steel tendon parameters, each prestressed steel tendon is spatially matched with the segment number sequence, and the collision detection function of the designed BIM model is used to determine the segment range where the steel tendon is located, and the segment type and segment number to which each steel tendon belongs are determined; based on the preset tensioning levels, such as initial tensioning, intermediate tensioning, and final tensioning, the steel tendons in each segment are grouped according to the tensioning levels. Initial tensioning is for steel tendons with temporary fixation requirements, intermediate tensioning is used to adjust the structural line shape, and final tensioning reaches the designed stress state. Each group forms a tensioning control unit consisting of a segment number and tensioning level, such as A01-initial and B03-final.

[0109] Step S52: establishing a control unit association graph based on the tensioning control units and the robustness index characteristic vector Φ;

[0110] Specifically, the temporal dependencies between different tensioning control units within the same segment were analyzed. Initial tensioning must be completed before intermediate tensioning, which in turn must be completed before final tensioning, forming an irreversible sequence. The spatial coupling relationships between adjacent segments were analyzed. The tensioning of steel strands in adjacent segments can affect each other. For example, tensioning of segment A01 can cause additional stress in segment A02. Based on the results of the three-segment classification, the structural correlations between different segments were identified. Segments of the same type exhibited strong correlations due to similar stress characteristics. For example, segments A01 and A02, both located in the negative moment zone, exhibited significant tensioning interactions. Segments of different types, such as A01 and B03, exhibited weak correlations due to their different stress mechanisms. A control unit association graph was constructed with tension control units as nodes. Temporal dependencies form directed edges, with arrows pointing to subsequent tensioning stages, such as A01-initial → A01-intermediate. Spatial coupling relationships form undirected edges, connecting tension control units in adjacent segments, such as A01-final → B01-final. Edge weights are determined by both the spatial distance between segments and the structural correlation. The closer the distance and the stronger the correlation, the greater the weight. For example, if A01 and A02 are 5 meters apart and of the same type, the weight is set to 0.8; if A01 and B03 are 20 meters apart and of different types, the weight is set to 0.3. The robustness index eigenvector Φ is used as the global state feature of the control unit association graph, and each node is assigned a current risk status attribute, such as the high crack risk attribute associated with the A01-final node.

[0111] For example, Figure 4 As shown, the circles represent tension control units, including A01-initial, A01-middle, A01-final, A02-initial, B01-initial, and A02-final;

[0112] Table 1 Figure 4 The directed edges in are described in detail.

[0113] Table 1 Figure 4 Detailed description of the directed edges of the control unit association graph

[0114]

[0115] Table 2 Figure 4 The undirected edges in are explained in detail.

[0116] The tensile control unit forms a control unit association graph through directed edges representing temporal dependencies and undirected edges representing edge weights.

[0117] Table 2 Figure 4 Detailed description of the undirected edges of the control unit association graph

[0118]

[0119] Step S53: Based on the control unit association graph and the robustness index feature vector, combined with the precise digital twin model, a graph neural network tensioning decision model is constructed and trained;

[0120] The graph neural network tensioning decision model uses the control unit association graph as the input structure and the robustness index feature vector as the node feature. The tensioning decision objective function is defined, including crack risk minimization, linear deviation minimization, and connection reliability maximization. In the crack risk minimization objective function, the segments with higher crack risk levels have a greater corresponding weight, the linear deviation minimization uses the sum of the squares of the difference between the actual linear shape and the designed linear shape as the indicator, and the connection reliability maximization uses the proximity of the preload ratio to the reliability threshold as the indicator; tensioning constraints are set, the single tensioning force does not exceed the designed tensioning force value to avoid over-tensioning fracture of the steel strand, and the displacement difference between adjacent segments does not exceed the ratio range [R min ,R max ] to prevent connection damage caused by excessive displacement differences. Using a precise digital twin model as the simulation environment, the current tensioning state parameters are input to output structural responses, such as strain, displacement, and preload changes. The parameters of the graph neural network tensioning decision model are trained using a reinforcement learning algorithm: After initializing the model parameters, the model outputs the tensioning decision in each round of training, and the digital twin model returns the structural response and objective function value under this decision. If the objective function value improves, the parameters are adjusted to enhance the selection probability of the decision, and vice versa. Training is iterated until the objective function converges, for example, the objective function value fluctuation is less than 5% after 100 consecutive rounds of training.

[0121] Step S54: obtaining the robustness index feature vector in the current state, and using the trained graph neural network tensioning decision model to generate an adaptive tensioning strategy sequence;

[0122] The robustness index characteristic vector for the current state is obtained and expanded using the state of the tensioned control unit and the initial state of the untensioned control unit to obtain the expanded robustness index characteristic vector. The state of the tensioned control unit, such as A01 (initial tensioning), corresponds to a tensioning force of 1000 kN; the initial state of the untensioned control unit, such as A01 (final untensioning), corresponds to an initial tensioning force of 0 kN. The robustness index characteristic vector is essentially a quantitative description of the current risk state of the structure, such as the crack risk of a Class A segment, the linear deviation of a Class B segment, and the connection reliability of a Class C segment, but it does not include dynamic information about the tensioning process. The generation of tensioning strategies depends on both the structural risk status and the tensioning implementation progress. The completion status of tensioned units affects the stress environment of untensioned units. Therefore, two types of information must be supplemented to the robustness index eigenvector: the status of the tensioned control unit, including its segment number, tensioning level, actual tensioning force, and the corresponding structural response changes, such as the reduction in crack risk after tensioning; and the initial state of the untensioned control unit, including its preset tensioning level, designed tensioning force, and the risk status of the currently associated segment, indicating whether the corresponding segment is in a high-risk state. The robustness index eigenvector in the current state incorporates the status of the tensioned control unit and the initial state information of the untensioned control unit to form a complete eigenvector in the current state, namely the expanded robustness index eigenvector. The expanded robustness index feature vector still uses φA, φB, and φC as the core skeleton, and new information, namely the state of the tensioned control unit and the initial state of the untensioned control unit, is embedded as an additional dimension. For example, the identifier of the tensioned unit of the corresponding segment is added to φA to ensure consistency with the input structure when training the graph neural network.

[0123] The expanded robustness indicator feature vector is input into the trained graph neural network tensioning decision model. The graph neural network tensioning decision model extracts node-related features through the graph neural network layer and outputs a tensioning decision, namely the next control unit to be tensioned and its tensioning force value. A precise digital twin model is used to predict the structural response after the tensioning decision is executed, verifying whether the tensioning constraints set during the training of the graph neural network tensioning decision model are met. If so, the tensioning decision is retained; otherwise, a new tensioning decision is generated. This process is iterated until a complete tensioning strategy sequence covering all tensioning control units is generated, which serves as the adaptive tensioning strategy sequence.

[0124] Step S55: Closed-loop execution is performed according to the adaptive tensioning strategy sequence, and dynamic correction is performed through real-time feedback data.

[0125] Tensioning operations are performed sequentially according to the adaptive tensioning strategy sequence. After each tensioning control unit is tensioned, real-time strain, displacement, and preload data are acquired through the monitoring system, and the robustness indicator feature vector is updated. For example, after the final tensioning of a Class A segment is completed, its crack risk level changes from high to medium. Based on the updated robustness indicator feature vector, it is determined whether the subsequent tensioning strategy needs to be adjusted. If the deviation between the real-time risk status of a particular segment and the prediction of the graph neural network tensioning decision model exceeds a preset threshold—for example, if the model predicts low risk but the actual monitoring shows medium risk—the updated robustness indicator feature vector is re-input into the trained graph neural network tensioning decision model to generate a new subsequent tensioning strategy. After the tensioning operations of all tensioning control units are completed, the tensioning effect is evaluated based on the full-process monitoring data, including the reduction in crack risk level and the degree of linear deviation control. The parameters of the graph neural network tensioning decision model are optimized based on the evaluation results, such as adjusting the weights of indicators such as crack risk, linear deviation, and connection reliability in the objective function.

[0126] The integration of prestressed steel strand parameters and segment classification in step S50 enables the tensioning control unit to be clearly structure-specific, avoiding the localized stress imbalances caused by traditional uniform tensioning sequences. The control unit association diagram incorporates both temporal dependencies and spatial coupling, enabling decisions to balance construction sequence constraints with overall structural stress dependencies. This reduces the generation of secondary stresses across segments compared to tensioning strategies that only consider a single segment. The synergy between a graph neural network (GNN) and reinforcement learning (RL) approaches, coupled with a precise digital twin model, enables a dynamic mapping from structural state to tensioning decisions. The GNN's ability to handle complex relationships enables the model to identify implicit cross-segment influences, such as the indirect effect of tensioning on the preload of a Class A segment on a Class C node. Reinforcement learning continuously optimizes the strategy through simulation feedback from the digital twin model, addressing the limitations of traditional empirical tensioning strategies in adapting to dynamic structural changes. The closed-loop execution mechanism of the adaptive tensioning strategy sequence enables the strategy to adjust based on real-time conditions, preventing strategy failures caused by construction errors or material property fluctuations. For example, if the actual stiffness of a segment is less than the model's prediction, the strategy automatically reduces the tensioning force to maintain a safe strain range. Step S50 works in conjunction with the robustness indicator eigenvector of step S40. The robustness indicator eigenvector provides a precise risk-based guide for tensioning strategy generation, prioritizing tensioning operations in high-risk areas, such as Class A segments with high crack risk. Working in conjunction with the precise digital twin model of step S30, the accurate simulation results provided by the precise digital twin model ensure the effectiveness of reinforcement learning training, making the generated tensioning strategy feasible in actual implementation. Without this step, prestressing would rely on a fixed process and be unable to adjust according to the real-time state of the structure. This could result in high-risk areas not being promptly controlled or excessive tensioning causing new damage. The present invention demonstrates significant effectiveness in the collaborative optimization of different segment types. For example, by adjusting the tensioning sequence of Class B segments, the deformation coordination of Class A segments can be indirectly improved. This cross-segment correlation and control mechanism is difficult to implement in traditional tensioning strategies. Through mutual feedback and dynamic adaptation of the stress states of multiple segment types, it significantly improves the overall robustness of the structure.

[0127] Example 2

[0128] This embodiment provides a prestressing parameter optimization system for precast segments of steel-concrete composite beam bridges based on embodiment 1, such as Figure 5 Shown, including:

[0129] Segment classification module: used to obtain the design BIM model and classify prefabricated segments based on the design BIM model;

[0130] Data acquisition module: Based on the classification results, the initial digital twin model of the prefabricated segment is constructed and monitoring points are arranged to obtain the original monitoring data stream;

[0131] Data processing module; used to screen, verify and correct the original monitoring data stream to obtain the corrected original monitoring data stream;

[0132] Model calibration module: Compares the corrected original monitoring data stream with the simulation results of the initial digital twin model of the prefabricated segment to determine whether a calibration signal is triggered. If so, the initial digital twin model is dynamically calibrated to generate an accurate digital twin model of the prefabricated segment.

[0133] Robustness indicator evaluation module: performs multi-dimensional robustness indicator evaluation based on the modified original monitoring data stream and constructs the robustness indicator feature vector;

[0134] Strategy generation module: Based on the robustness index eigenvector, a tensioning control unit is constructed, and a control unit association diagram is established. Combined with the precise digital twin model of the prefabricated segment, a graph neural network tensioning decision model is constructed and trained to generate an adaptive tensioning strategy sequence.

[0135] In the robustness index evaluation module, the specific process of constructing the robustness index feature vector includes:

[0136] Step S41, performing a crack risk assessment and deformation coordination assessment in the negative bending moment zone based on the corrected strain data and displacement data of the Class A segment in the corrected original monitoring data stream, and generating a comprehensive risk status record table for the Class A segment;

[0137] Step S42: Based on the corrected strain data and displacement data of the Class B segment in the corrected original monitoring data stream, a stress level assessment and a linear state assessment are performed in the mid-span area to generate a comprehensive state record table for the Class B segment.

[0138] Step S43, performing connection reliability assessment and temperature stability assessment based on the corrected preload force data of the Class C segment and the reliability temperature data of the Class C segment in the corrected original monitoring data stream, and generating a Class C segment comprehensive status record table;

[0139] Step S44 : constructing a robustness index feature vector based on the Class A segment comprehensive risk status record table, the Class B segment comprehensive status record table, and the Class C segment comprehensive status record table.

[0140] In the strategy generation module, a tensioning control unit is constructed, and a control unit association diagram is established. Combined with the precise digital twin model of the prefabricated segment, a graph neural network tensioning decision model is constructed and trained. The specific process of generating an adaptive tensioning strategy sequence includes:

[0141] Step S51, extracting prestressed steel tendon parameters from the design BIM model and constructing a tensioning control unit in combination with segment classification identification data;

[0142] Step S52: establishing a control unit association graph based on the tensioning control units and the robustness index characteristic vector Φ;

[0143] Step S53: Based on the control unit association graph and the robustness index feature vector, combined with the precise digital twin model, a graph neural network tensioning decision model is constructed and trained;

[0144] Step S54: obtaining the robustness index feature vector in the current state, and using the trained graph neural network tensioning decision model to generate an adaptive tensioning strategy sequence;

[0145] Step S55: Closed-loop execution is performed according to the adaptive tensioning strategy sequence, and dynamic correction is performed through real-time feedback data.

[0146] The methods and systems of the present application may be implemented in many ways. For example, the methods and systems of the present application may be implemented using software, hardware, firmware, or any combination of software, hardware, and firmware. The above order of steps used in the method is for illustration only, and the steps of the method of the present application are not limited to the order specifically described above unless otherwise specified.

[0147] In addition, the parts of the above technical solutions provided in the embodiments of the present application that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail to avoid excessive redundancy.

[0148] The above-described specific embodiments further illustrate the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above description is merely a specific embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for optimizing prestressing parameters of precast segments of steel-concrete composite beam bridges, characterized in that: The method comprises: Obtain the design BIM model, classify the prefabricated segments based on the design BIM model, build the initial digital twin model of the prefabricated segments based on the classification results, and deploy monitoring points to obtain the original monitoring data stream; Screening, verifying and correcting the original monitoring data stream to obtain a corrected original monitoring data stream; Compare the corrected original monitoring data stream with the simulation results of the initial digital twin model of the prefabricated segment to determine whether a calibration signal is triggered. If so, perform dynamic calibration of the initial digital twin model to generate an accurate digital twin model of the prefabricated segment. Based on the modified original monitoring data stream, a multi-dimensional robustness index evaluation is performed to construct a robustness index feature vector; Based on the robustness index eigenvector, a tensioning control unit is constructed, and a control unit association diagram is established. Combined with the precise digital twin model of the prefabricated segment, a graph neural network tensioning decision model is constructed and trained to generate an adaptive tensioning strategy sequence.

2. The method for optimizing prestressing parameters of precast segments of steel-concrete composite beam bridges according to claim 1 is characterized in that: The classification of prefabricated segments based on the design BIM model includes: The structural parameters of the steel-concrete composite beam bridge are extracted from the designed BIM model; according to the structural parameters, the precast segments are divided into three types of segments, including Class A segments, i.e., negative bending moment zone segments, Class B segments, i.e., mid-span segments, and Class C segments, i.e., connection node segments.

3. The method for optimizing prestressing parameters of precast segments of steel-concrete composite beam bridges according to claim 2 is characterized in that: The construction of the initial digital twin model of the prefabricated segment includes: Based on the classification results of the three types of segments, the detailed structural parameters of each type of segment are extracted to generate segment classification identification data; based on the segment classification identification data, the initial digital twin model of the prefabricated segment is constructed on the basis of the designed BIM model.

4. The method for optimizing prestressing parameters of precast segments of steel-concrete composite beam bridges according to claim 3 is characterized in that: The method for obtaining the corrected original monitoring data stream includes: Perform preliminary screening on the original monitoring data stream, remove abnormal data, and obtain the filtered original monitoring data stream; Perform adjacent comparison and secondary verification on the filtered original monitoring data stream to obtain the reliability strain data and reliability displacement data of Class A segments, the reliability strain data and reliability displacement data of Class B segments, and the reliability temperature data and reliability preload data of Class C segments; According to the reliability temperature data of the Class C segment, the reliability strain data and reliability displacement data of the Class A segment, the reliability strain data and reliability displacement data of the Class B segment, and the reliability preload data of the Class C segment are corrected for temperature effects to obtain a corrected original monitoring data stream; the corrected original monitoring data stream includes the corrected strain data and displacement data of the Class A segment, the corrected strain data and displacement data of the Class B segment, and the corrected preload data of the Class C segment.

5. The method for optimizing prestressing parameters of precast segments of steel-concrete composite beam bridges according to claim 4 is characterized in that: The method for determining whether to trigger a calibration signal includes: The corrected original monitoring data stream is compared with the initial digital twin model simulation results of the prefabricated segment to determine whether the Class A segment strain calibration signal, Class A segment displacement calibration signal, Class B segment strain calibration signal, Class B segment displacement calibration signal and Class C segment connection calibration signal are triggered. If any of the Class A segment strain calibration signal, Class A segment displacement calibration signal, Class B segment strain calibration signal, Class B segment displacement calibration signal and Class C segment connection calibration signal is triggered, the calibration signal is determined to be triggered.

6. The method for optimizing prestressing parameters of precast segments of steel-concrete composite beam bridges according to claim 5 is characterized in that: The method for constructing a robustness index characteristic vector includes: Based on the corrected strain and displacement data of the Class A segments in the corrected original monitoring data stream, the crack risk assessment and deformation coordination assessment in the negative bending moment area are performed to generate a comprehensive risk status record table for the Class A segments. Based on the corrected strain and displacement data of the Class B segment in the corrected original monitoring data stream, the stress level and linear status of the mid-span area are evaluated, and a comprehensive status record table of the Class B segment is generated; Based on the corrected preload data of the Class C segment and the reliability temperature data of the Class C segment in the corrected original monitoring data stream, the connection reliability evaluation and temperature stability evaluation are performed to generate a comprehensive status record table of the Class C segment; Based on the comprehensive risk status record table of Class A segments, the comprehensive status record table of Class B segments, and the comprehensive status record table of Class C segments, a robustness index feature vector is constructed.

7. The method for optimizing prestressing parameters of precast segments of steel-concrete composite beam bridges according to claim 6 is characterized in that: The method for constructing the tensioning control unit comprises: Extract prestressed steel tendon parameters from the design BIM model; assign unique segment numbers to each Class A segment, Class B segment, and Class C segment based on segment classification identification data; According to the prestressed steel tendon parameters, each prestressed steel tendon is spatially matched with the segment number sequence to determine the segment type and segment number to which each prestressed steel tendon belongs; Based on the preset tensioning levels, the prestressed steel strands in each segment are grouped according to the tensioning levels to form a tensioning control unit consisting of segment number and tensioning level.

8. The method for optimizing prestressing parameters of precast segments of steel-concrete composite beam bridges according to claim 7 is characterized in that: The method for establishing a control unit association diagram comprises: Analyze the temporal dependencies of different tensioning-level control units within the same segment to form an irreversible sequence; Analyze the spatial coupling relationship between adjacent segments and identify the structural correlation between different segments based on the three-category segment classification results; Taking the tension control unit as the node, the temporal dependency relationship forms a directed edge, and the spatial coupling relationship forms an undirected edge to construct a control unit association graph.

9. The method for optimizing prestressing parameters of precast segments of steel-concrete composite beam bridges according to claim 8 is characterized in that: The method for generating an adaptive tensioning strategy sequence includes: The robustness indicator feature vector in the current state is expanded using the state of the tensioned control unit and the initial state of the untensioned control unit to obtain an expanded robustness indicator feature vector; the expanded robustness indicator feature vector is input into the trained graph neural network tensioning decision model, and the tensioning decision is output, where the tensioning decision is the next control unit to be tensioned and its tensioning force value; A precise digital twin model is used to predict the structural response after the tensioning decision is executed, and verify whether the tensioning constraints set when training the graph neural network tensioning decision model are met; if so, the tensioning decision is retained; otherwise, the tensioning decision is regenerated, and the above process is iterated until a complete tensioning strategy sequence covering all tensioning control units is generated as an adaptive tensioning strategy sequence.

10. A system for optimizing prestressed tensioning parameters of precast segments of steel-concrete composite beam bridges, which is used to implement the method for optimizing prestressed tensioning parameters of precast segments of steel-concrete composite beam bridges according to any one of claims 1 to 9, characterized in that: The system comprises: Segment classification module: used to obtain the design BIM model and classify prefabricated segments based on the design BIM model; Data acquisition module: Based on the classification results, the initial digital twin model of the prefabricated segment is constructed and monitoring points are arranged to obtain the original monitoring data stream; Data processing module; used to screen, verify and correct the original monitoring data stream to obtain the corrected original monitoring data stream; Model calibration module: Compares the corrected original monitoring data stream with the simulation results of the initial digital twin model of the prefabricated segment to determine whether a calibration signal is triggered. If so, the initial digital twin model is dynamically calibrated to generate an accurate digital twin model of the prefabricated segment. Robustness indicator evaluation module: performs multi-dimensional robustness indicator evaluation based on the modified original monitoring data stream and constructs the robustness indicator feature vector; Strategy generation module: Based on the robustness index eigenvector, a tensioning control unit is constructed, and a control unit association diagram is established. Combined with the precise digital twin model of the prefabricated segment, a graph neural network tensioning decision model is constructed and trained to generate an adaptive tensioning strategy sequence.

Images