Long-span rigid frame bridge long-term deformation prediction method and system

Abstract

This invention proposes a method and system for predicting the long-term deformation of long-span rigid frame bridges, belonging to the field of bridge deformation prediction technology. Addressing the shortcomings of existing prediction methods in considering the coupling effect of geometric nonlinearity and concrete shrinkage and creep, and lacking dynamic response capabilities to damage events during operation, leading to significant deviations between predicted and actual results, this invention establishes a spatial finite element model and constructs structural equilibrium equations that simultaneously consider geometric nonlinearity and shrinkage and creep effects, achieving synchronous coupling analysis of these two effects. Through benchmark simulation and damage event identification based on a probability distribution model, combined with monitoring data, the model is dynamically corrected for damage and the simulation is restarted, constructing a closed-loop prediction mechanism of identification, correction, and restart. This effectively improves the accuracy and precision of long-term deformation and internal force prediction, and enhances adaptability in complex scenarios.

Description

Technical Field

[0001] This invention relates to the field of bridge deformation prediction technology, and in particular to a method and system for predicting the long-term deformation of long-span rigid frame bridges. Background Technology

[0002] Long-span rigid frame bridges have been widely used in modern bridge engineering due to their outstanding advantages such as good structural integrity, superior seismic performance, wide visibility under the bridge, and low maintenance costs. With advancements in materials science and construction technology, the span records for rigid frame bridges are constantly being broken, demonstrating unique advantages when crossing complex terrains such as deep valleys and straits, and showing broad development prospects. Especially in the construction of highway networks and urban expressways, long-span rigid frame bridges have become an indispensable bridge type choice.

[0003] However, existing analytical methods still have significant shortcomings in predicting the long-term service performance of long-span rigid frame bridges. Traditional deformation prediction methods often calculate geometric nonlinearity and concrete shrinkage and creep effects separately and then linearly superimpose them, failing to adequately consider the impact of their coupled effect on the long-term performance of the structure. More importantly, existing methods generally lack the ability to dynamically respond to damage events during operation, such as vehicle collisions and earthquakes. They cannot automatically correct the calculation model and re-predict after identifying structural damage. This one-off prediction mode leads to a large deviation between the calculation results and the actual service condition, making it difficult to provide a reliable basis for preventive maintenance and precise treatment of bridges. Especially after unexpected damage, traditional methods cannot achieve dynamic adjustment of the prediction path and adaptive updating of the model, seriously affecting the accuracy of long-term deformation prediction and its universality in complex scenarios.

[0004] Therefore, there is an urgent need for a more accurate and universally applicable method and system for predicting the long-term deformation of long-span rigid frame bridges. Summary of the Invention

[0005] To address this, the present invention provides a method and system for predicting the long-term deformation of long-span rigid frame bridges. This method overcomes the problems in existing technologies, which fail to adequately consider the impact of the coupling effect of geometric nonlinearity and concrete shrinkage and creep on the long-term performance of the structure, lack dynamic response capabilities to damage events during operation, and result in large deviations between the calculation results and the actual service conditions. This makes it difficult to provide a reliable basis for the preventive maintenance and precise treatment of bridges.

[0006] To achieve the above objectives, the present invention provides a method for predicting the long-term deformation of a long-span rigid frame bridge, comprising:

[0007] S1. Based on the bridge's design drawings, material properties, and boundary conditions, a spatial finite element model is established.

[0008] S2, Based on the aforementioned spatial finite element model, construct a structural equilibrium equation that simultaneously considers geometric nonlinearity and concrete shrinkage and creep effects;

[0009] S3. Based on the structural equilibrium equation, a benchmark simulation is performed on the spatial finite element model to determine the evolution of the benchmark mechanical state from the start of construction to the target predicted lifespan.

[0010] S4 identifies damage events based on a pre-established baseline probability distribution model;

[0011] S5. When a damage event is detected, the baseline simulation is paused, and damage correction is performed on the spatial finite element model based on the monitoring data to obtain a corrected finite element model characterizing the state after the damage event. Based on the corrected finite element model, a corrected structural equilibrium equation is constructed.

[0012] S6. Based on the modified finite element model and the modified structural equilibrium equation, perform a baseline simulation again until no damage event is detected.

[0013] S7 outputs the predicted results of long-term deformation and internal forces of the bridge at the target age;

[0014] The damage events include at least vehicle collisions and earthquakes.

[0015] Furthermore, the benchmark simulation performed on the spatial finite element model includes:

[0016] S31, starting from the construction start point, proceeds step by step in chronological order to the target prediction year, dividing the total time into continuous time steps;

[0017] S32, execute the baseline simulation loop within each current time step;

[0018] S33, advance the time to the next time step, and repeat S32 until the simulation time reaches the target prediction year.

[0019] Furthermore, the execution of the benchmark simulation loop includes:

[0020] S32a, applies the concrete shrinkage and creep effects up to the previous time step as equivalent nodal forces to the current structural configuration;

[0021] S32b, Solve the structural equilibrium equations, and update the structural displacements and end forces of the members through iterative calculations until the solution converges;

[0022] S32c, based on the updated rod end force, calculates the shrinkage and creep equivalent nodal force for the next time step by fitting a recursive formula for the creep coefficient based on an exponential function.

[0023] Furthermore, the establishment of the baseline probability distribution model includes:

[0024] Collect historical monitoring data of the bridge under normal operating conditions to form a training sample set;

[0025] Based on the training sample set, a baseline probability distribution model for each mechanical state monitoring parameter is established and its statistical characteristic values ​​are determined.

[0026] Based on hypothesis testing methods, upper and lower control limits are set for the abnormal index used to identify damage events.

[0027] Furthermore, the identification of damage events based on a pre-established baseline probability distribution model includes:

[0028] S41, acquire monitoring data that has not been involved in identification and judgment within the actual time interval corresponding to the current time step, and use it as new monitoring data;

[0029] S42, Calculate the comprehensive anomaly index of the new monitoring data based on the benchmark probability distribution model;

[0030] S43, using the sliding time window technique, analyze the changing trend of the comprehensive anomaly index within continuous time steps;

[0031] S44, when the comprehensive abnormality index is identified to exceed the control upper limit for k consecutive time steps, or to be below the control lower limit for k consecutive time steps, a damage event is determined to have occurred;

[0032] Where k is a preset sensitivity parameter.

[0033] Furthermore, the calculation of the anomaly index of the new monitoring data based on the baseline probability distribution model includes:

[0034] S42a, For each new monitoring data point, calculate its standardized residual relative to the baseline probability distribution model;

[0035] S42b, Based on the standardized residuals, the local anomaly index of each monitoring data is obtained by calculating Mahalanobis distance;

[0036] S42c, based on the spatial distribution characteristics of the monitoring data, weights and fuses each local anomaly index to form a comprehensive anomaly index of the overall structure.

[0037] Furthermore, the damage correction based on monitoring data on the spatial finite element model includes:

[0038] S51, Based on the monitoring data that triggers damage event identification and the spatial distribution of its abnormal index, identify the damaged area in the structure;

[0039] S52, within the damaged area, the residual between the dynamic characteristics or static response of the spatial finite element model and the measured data is minimized by weighted least squares method, thereby identifying and correcting the structural parameters of the damaged area.

[0040] S53. Based on the structural parameters of the corrected damaged area, the coordinates of all nodes of the spatial finite element model are corrected to obtain the corrected finite element model.

[0041] Furthermore, the monitoring data includes data collected by at least one of the following sensors: strain data collected by a fiber optic grating sensor, rotational angle data collected by an inclinometer, and vibration data collected by an accelerometer.

[0042] Furthermore, the long-term deformation and internal force prediction results of the bridge at the target age include: the time-varying deformation envelope diagram of the entire bridge, the time history of internal forces at key sections, and the long-term stress distribution cloud map.

[0043] This invention also provides a long-term deformation prediction system for long-span rigid frame bridges, wherein the system employs any of the methods described above for predicting long-term deformation of long-span rigid frame bridges, specifically including:

[0044] The model building module is used to build a spatial finite element model based on the bridge's design drawings, material properties, and boundary conditions.

[0045] The equation construction module, connected to the model building module, is used to construct structural equilibrium equations that simultaneously consider geometric nonlinearity and concrete shrinkage and creep effects based on the spatial finite element model.

[0046] The cyclic simulation module, connected to the equation construction module, is used to perform a benchmark simulation on the spatial finite element model based on the structural equilibrium equations to determine the evolution of the benchmark mechanical state from the start of construction to the target predicted lifespan.

[0047] An event recognition module, connected to the cyclic simulation module, is used to identify damage events based on a pre-established baseline probability distribution model.

[0048] The event correction module, connected to the event identification module, is used to pause the baseline simulation when a damage event is identified, and perform damage correction on the spatial finite element model based on the monitoring data to obtain a corrected finite element model characterizing the state after the damage event, and construct a corrected structural equilibrium equation based on the corrected finite element model.

[0049] The correction simulation module, connected to the event correction module, is used to re-perform the baseline simulation based on the corrected finite element model and the corrected structural equilibrium equation;

[0050] The result prediction module, connected to the correction simulation module, is used to output the long-term deformation and internal force prediction results of the bridge at the target age.

[0051] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0052] Firstly, this invention achieves simultaneous coupled analysis of geometric nonlinearity and concrete shrinkage and creep effects using the initial strain method and equivalent nodal force conversion technology, avoiding the error accumulation caused by linear superposition. Simultaneously, it constructs a dynamic closed loop of damage event identification, model damage correction, and simulation restart. Based on improved DBSCAN spatial clustering and weighted least squares methods, it accurately locates the damaged area and corrects structural parameters and geometric configuration, ensuring that the calculation model always closely matches the actual service state of the bridge. This solves the problem of large deviations between traditional one-time predictions and reality, improving the accuracy and precision of long-term deformation and internal force prediction results.

[0053] Secondly, this invention improves the simulation efficiency during long-term operation by using non-uniform time step division and recursive calculation techniques, avoiding the pressure of storing massive amounts of historical data. It constructs a comprehensive anomaly index using standardized residuals, Mahalanobis distance, and weighted fusion techniques, combined with sliding time window trend analysis, to reduce the false alarm and false negative rates of damage event identification, providing high-quality data input for model correction. Through spatial interpolation, mesh optimization, and visual modeling, it achieves refined presentation and verification of the entire bridge deformation, internal forces at key sections, and stress distribution. The multi-technology collaboration throughout the entire process from data acquisition and model calculation to result output further reduces the deviation between predicted and actual values, achieving higher precision and accuracy in prediction.

[0054] Thirdly, this invention addresses the pain point of traditional methods lacking dynamic response capabilities to damage events by establishing a complete response mechanism from event identification and simulation adjustment to model correction. First, by combining a baseline probability distribution model and sliding time window technology with standardized residuals and Mahalanobis distance weighted fusion calculations, it achieves accurate identification of damage events such as vehicle collisions and earthquakes, effectively avoiding misjudgments of instantaneous fluctuations and missed event reports. Second, after identifying a damage event, it can immediately pause the baseline simulation and save the current mechanical state, ensuring the continuity of subsequent corrections and simulations, solving the problem of inability to reconnect after interruption in traditional predictions. Finally, by using improved DBSCAN spatial clustering to locate the damage area, weighted least squares method to correct structural parameters, and radial basis function interpolation to update node coordinates, it achieves adaptive adjustment of the model, enabling the corrected model to truly reflect the structural morphology and mechanical properties after damage. Then, the simulation is restarted based on the corrected model, ensuring that the long-term deformation and internal force predictions after the damage event still conform to the actual service state, completely changing the limitation of traditional one-time predictions being unable to cope with sudden damage, and significantly improving the universality and reliability of prediction results in complex service scenarios. Attached Figure Description

[0055] Figure 1 This is a flowchart of a method for predicting the long-term deformation of a long-span rigid frame bridge according to an embodiment of the present invention.

[0056] Figure 2 This is a structural block diagram of a long-span rigid frame bridge long-term deformation prediction system according to an embodiment of the present invention. Detailed Implementation

[0057] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0058] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0059] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0060] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0061] Example 1

[0062] like Figure 1 As shown, this invention proposes a method for predicting the long-term deformation of long-span rigid frame bridges, including:

[0063] S1. Based on the bridge's design drawings, material properties, and boundary conditions, a spatial finite element model is established.

[0064] In one possible implementation, key structural nodes are generated using a quadratic interpolation method based on the axial coordinates of the main beam and the piers. The node spacing is controlled by the ratio of the spacing between adjacent piers to 20. The concrete members are simulated using Timoshenko beam elements that take into account shear deformation. The actual centroid position of the members is accurately reproduced by setting the section offset.

[0065] The concrete material used is C55 concrete with an elastic modulus of 3.55 × 10⁻⁶. 4 MPa, Poisson's ratio 0.2, density 26 kN / m³, the shrinkage and creep model adopted is the CEB-FIP2010 model, considering parameters such as ambient humidity 70% and theoretical thickness of the component; the prestressed steel bars adopt a three-segment elastoplastic model, specifically, the elastic modulus is 1.95 × 10⁻⁶. 5 MPa, yield strength is 1860MPa.

[0066] For the pier base consolidation, all six degrees of freedom are constrained; for the fixed support, the three translational degrees of freedom are constrained; for the sliding support, the longitudinal displacement of the beam is released; and spring elements are used to simulate the compression and shear stiffness of the support.

[0067] The mesh density was verified using the h-convergence method to ensure that the internal force error of the critical section was less than 2%.

[0068] S2, Based on the aforementioned spatial finite element model, construct a structural equilibrium equation that simultaneously considers geometric nonlinearity and concrete shrinkage and creep effects;

[0069] In one possible implementation, S21, the fundamental geometric nonlinear equations are established. Specifically, the beam element equilibrium equations are established in the element local coordinate system using the rotating coordinate method.

[0070] Considering the large displacement effect, the incremental form of the virtual work equation is established using the UL formulation; thus forming the element tangent stiffness matrix that includes the geometric stiffness matrix.

[0071] S22 introduces initial shrinkage and creep strain. Specifically, the shrinkage strain and creep strain of concrete are treated as initial strains; the shrinkage strain is calculated according to the standard formula, taking into account time and development law; the creep strain is fitted with an exponential function.

[0072] S23, calculate the equivalent nodal forces. Specifically, the initial strain of shrinkage and creep is converted into equivalent nodal forces through integration; a recursive formula is established to avoid storing historical data; and the equivalent nodal forces for the next time step are calculated using the recursive results.

[0073] S24, coordinate system transformation. Specifically, the equilibrium equations in the local coordinate system are transformed to the overall structural coordinate system through a transformation matrix; the contribution of the geometric stiffness matrix is ​​considered to form the overall structural tangent stiffness matrix.

[0074] S25. Establish the coupled equilibrium equations. Specifically, incorporate the shrinkage and creep equivalent nodal forces as additional load terms into the equilibrium equations to form the final incremental equilibrium equations.

[0075] S26. Establish a solution strategy. Specifically, use the Newton-Raphson iteration method to solve the nonlinear equations. Set the convergence criterion as the ratio of the unbalanced force norm to the reference load norm being less than the tolerance. Perform iterative calculations in each time step until convergence.

[0076] This invention solves the error accumulation problem of traditional separate calculations of geometric nonlinearity and shrinkage / creep effects by using the initial strain method and equivalent nodal force conversion technology, achieving simultaneous coupled analysis of the two complex effects and improving the accuracy and reliability of long-term deformation prediction. Through exponential function fitting and recursive calculation techniques, it overcomes the technical bottleneck of large historical data storage in creep analysis, achieving efficient data management for massive time-step calculations and reducing computational resource consumption and storage requirements. By combining the rotational coordinate method with the UL formulation, it solves the accuracy problem of establishing equilibrium equations under large displacement conditions, achieving accurate description of geometric nonlinear effects and improving the calculation accuracy of structural response under complex stress states. By setting a Newton iteration method solution strategy, it solves the convergence and stability problem of nonlinear equation systems, achieving efficient numerical solutions for complex coupled systems and improving the stability and convergence efficiency of the calculation process.

[0077] Alternative solutions for geometric nonlinearity processing include the Lagrange formula method, the corotation coordinate method, and the finite displacement theory method. Specifically, the Lagrange formula method uses a full Lagrange formulation to establish equilibrium equations based on the initial configuration, making it suitable for large displacement and small strain cases. The corotation coordinate method decomposes rigid body motion and deformation at the element level, simplifying nonlinear calculations. The finite displacement theory method uses higher-order strain-displacement relationships to improve calculation accuracy under large rotational conditions. The Lagrange formula method is theoretically rigorous but computationally intensive, making it suitable for theoretical research; the corotation coordinate method has high computational efficiency, making it suitable for engineering applications; and the finite displacement theory method has high accuracy but is complex to implement.

[0078] Alternative approaches to calculating creep shrinkage effects include the age-adjusted effective modulus method, the integral constitutive relation method, and the internal variable theory method. Specifically, the age-adjusted effective modulus method transforms the creep effect into an adjustment of the elastic modulus, simplifying the calculation process; the integral constitutive relation directly uses integral equations to describe creep development and is solved using numerical integration methods; the internal variable theory method introduces internal state variables to establish a more accurate creep prediction model. The age-adjusted effective modulus method is simple to calculate but has limited accuracy; the integral constitutive relation method has high accuracy but is computationally complex; and the internal variable theory method has clear physical meaning but is difficult to determine parameters.

[0079] S3. Based on the structural equilibrium equation, a benchmark simulation is performed on the spatial finite element model to determine the evolution of the benchmark mechanical state from the start of construction to the target predicted lifespan.

[0080] A benchmark simulation is performed on the aforementioned spatial finite element model, including:

[0081] S31, starting from the construction start point, proceeds step by step in chronological order to the target prediction year, dividing the total time into continuous time steps;

[0082] S32, execute the baseline simulation loop within each current time step;

[0083] S33, advance the time to the next time step, and repeat S32 until the simulation time reaches the target prediction year.

[0084] Execute the benchmark simulation loop, including:

[0085] S32a, applies the concrete shrinkage and creep effects up to the previous time step as equivalent nodal forces to the current structural configuration;

[0086] S32b, Solve the structural equilibrium equations, and update the structural displacements and end forces of the members through iterative calculations until the solution converges;

[0087] S32c, based on the updated rod end force, calculates the shrinkage and creep equivalent nodal force for the next time step by fitting a recursive formula for the creep coefficient based on an exponential function.

[0088] In one possible implementation, the time series is established by taking the first concrete component pouring day of the bridge as the starting point and using a non-uniform time step, such as a step of 15 days for the construction period of 0-2 years, a step of 30 days for the initial operation period of 2-5 years, and a step of 90 days for the long-term operation period of 5-50 years.

[0089] Within each time step, the equivalent nodal forces of shrinkage and creep from the previous time step are read from the stored historical data and applied as external loads to the current structural configuration. The structural tangent stiffness matrix is ​​assembled under the current configuration and solved using the Newton-Raphson iteration method. The convergence criterion is that the ratio of the unbalanced force norm to the total load norm is less than 0.0001. The converged displacements and end forces are output. Based on the converged end forces, the creep effect for the next time step is calculated using a recursive formula. The shrinkage strain increment is calculated and converted into equivalent nodal forces. The creep effect and shrinkage strain increment are stored for use in the next time step.

[0090] If the time represented by the current time step is greater than the target number of years, terminate the loop; otherwise, save the current state and return to S32 to execute the calculation of the next time step.

[0091] In the process of dividing the time step, alternative solutions include the adaptive time step method and the logarithmic time coordinate method. The adaptive time step method dynamically adjusts the time step based on the convergence speed and the rate of change of the response. It automatically reduces the step size when the response is drastic and increases the step size when it is stable. Its advantage is that it balances computational efficiency and accuracy. Its disadvantage is that it is complex to implement and requires setting adjustment criteria. The logarithmic time coordinate method divides the time step using logarithmic coordinates on the time axis. The step size is smaller in the early stage and gradually increases in the later stage. Its advantage is that it conforms to the development law of contraction and creep. Its disadvantage is that it may miss important changes in the later stage.

[0092] In the calculation of time-varying effects, alternative solutions include the age-adjusted effective modulus method and the genetic integral method. The age-adjusted effective modulus method transforms the creep effect into the time-dependent change of the elastic modulus, using the effective elastic modulus instead of the instantaneous elastic modulus. Its advantages are avoiding the storage of historical data and simple calculation, but its disadvantages are relatively low accuracy and approximation. The genetic integral method uses numerical integration to directly calculate the creep integral, using the trapezoidal rule or Simpson's rule for integral approximation. Its advantages are rigorous theory and high accuracy, but its disadvantages are large computational load and the need to store complete stress history.

[0093] In the process of solving the equations, alternative solutions include the quasi-Newton method and the explicit time integration method. The quasi-Newton method uses the rank-2 update to approximate the Hess matrix, reducing the number of times the stiffness matrix is ​​reformatted. Its advantage is improved efficiency in nonlinear solution, but its convergence is slightly worse than that of the Newton method. The explicit time integration method uses the central difference method for time integration, which does not require iterative solution of the nonlinear equations. Its advantage is avoiding iterative convergence problems, but its disadvantage is conditional stability and limited time step.

[0094] This invention addresses the shortcomings of traditional fixed-step methods, such as insufficient accuracy during construction and low efficiency during operation, through a non-uniform time-step partitioning technique. It optimizes the allocation of computing resources along the time axis, improving the computational efficiency and accuracy of life-cycle simulation. Furthermore, by employing a recursive calculation technique based on shrinkage and creep effects, it overcomes the challenge of massive historical data storage in long-term simulations, enabling efficient data management for massive time-step calculations, reducing computational and storage requirements, and increasing the upper limit of prediction time. Finally, by utilizing Newton's iteration method and convergence control techniques, it solves the convergence problem in coupled analysis of geometric and material nonlinearities, achieving efficient and stable solutions for complex nonlinear systems and improving the reliability and robustness of the computation process. Finally, by employing a phased cyclic simulation architecture, it addresses the technical connection between continuous simulation during construction and operation, enabling the tracking of the complete mechanical state evolution from the start of construction to the target lifespan and establishing a complete database of structural life-cycle performance development.

[0095] S4, based on a pre-established baseline probability distribution model, identifies damage events, including:

[0096] The establishment of the baseline probability distribution model includes:

[0097] Collect historical monitoring data of the bridge under normal operating conditions to form a training sample set;

[0098] Based on the training sample set, a baseline probability distribution model for each mechanical state monitoring parameter is established and its statistical characteristic values ​​are determined.

[0099] Based on hypothesis testing methods, upper and lower limits for the control of anomaly indices used to identify injury events are set.

[0100] In one possible implementation, a damage event is defined as an event or state change that occurs during bridge operation and significantly impacts structural safety or performance. Such events cause an abnormal deviation in the mechanical state of the bridge structure, exceeding the predictable range under normal operating conditions; examples include vehicle collisions and earthquakes. Specifically, historical monitoring data for at least two consecutive years during the bridge's normal operating period is collected. This data includes main girder deflection, tower displacement, structural frequency, and strain at key sections. The historical monitoring data undergoes data cleaning, alignment, environmental factor correction, and normalization. The dataset is divided by quarter to account for seasonal variations, and a sample database is established, including fields such as timestamps, monitoring values, and environmental parameters.

[0101] For each monitoring parameter, a distribution fit test is performed. Commonly used distribution types are normal distribution, log-normal distribution, and Weibull distribution. The KS test is used to determine the optimal distribution type. Marginal probability distribution models for each parameter are established. The mean and standard deviation of each parameter distribution are calculated. The quantile values ​​of the 95% and 99% confidence intervals are determined. Shape parameters such as skewness and kurtosis are calculated. A correlation coefficient matrix between parameters is established.

[0102] The initial control limits are set using the three-standard-deviation principle. The upper control limit is the sum of the mean and three standard deviations, and the lower control limit is the difference between the mean and three standard deviations. The control limits are adjusted based on the actual false alarm rate. The control limits are verified and fine-tuned through historical data to ensure that the target false alarm rate is controlled within 0.5%.

[0103] The model performance was tested using a reserved validation dataset. The false positive rate and false negative rate of the model were calculated, and the control limit settings were optimized based on the false positive rate and false negative rate to further balance sensitivity and specificity.

[0104] S41, acquire monitoring data that has not been involved in identification and judgment within the actual time interval corresponding to the current time step, and use it as new monitoring data;

[0105] S42, Calculate the comprehensive anomaly index of the new monitoring data based on the benchmark probability distribution model;

[0106] S43, using the sliding time window technique, analyze the changing trend of the comprehensive anomaly index within continuous time steps;

[0107] S44, when the comprehensive abnormality index is identified to exceed the control upper limit for k consecutive time steps, or to be below the control lower limit for k consecutive time steps, a damage event is determined to have occurred;

[0108] Where k is a preset sensitivity parameter.

[0109] Based on the aforementioned baseline probability distribution model, the anomaly index of the new monitoring data is calculated, including:

[0110] S42a, For each new monitoring data point, calculate its standardized residual relative to the baseline probability distribution model;

[0111] S42b, Based on the standardized residuals, the local anomaly index of each monitoring data is obtained by calculating Mahalanobis distance;

[0112] S42c, based on the spatial distribution characteristics of the monitoring data, weights and fuses each local anomaly index to form a comprehensive anomaly index of the overall structure.

[0113] In one possible implementation, a data interface is established with the bridge health monitoring system. After each simulated time step is calculated, monitoring data within the corresponding actual time interval is automatically collected. The collected raw data is then processed as follows:

[0114] Data validity verification: Remove invalid data generated by sensor malfunctions;

[0115] Timestamp alignment: unifying the time base of data from all monitoring points;

[0116] Data Identification: Add a "Not Participated in Identification and Judgment" status identifier to new monitoring data;

[0117] Data storage: Store valid data in the monitoring database at the current time step.

[0118] For each new monitoring data point, perform the following calculations:

[0119] Call the corresponding baseline probability distribution model to obtain the mean μ and standard deviation σ of the monitoring parameter;

[0120] Calculate the standardized residual Z = (x - μ) / σ, where x is the monitored value;

[0121] Outlier screening was performed on the standardized residuals to remove obvious outliers with |Z|>10.

[0122] Based on standardized residuals, the local anomaly index is calculated using the following method:

[0123] Multiple relevant parameters from the same monitoring point are combined to form a feature vector X: [Z1, Z2, ..., Z...]. p ];

[0124] Obtain the covariance matrix Σ from the baseline probability distribution model;

[0125] Calculate Mahalanobis distance In this context, the superscript T denotes the transpose of a vector, and the superscript -1 denotes the inversion of a matrix.

[0126] The Mahalanobis distance D 2 This serves as a local anomaly index for the monitoring point.

[0127] Based on the spatial distribution characteristics of the monitoring data, a weighted fusion process is performed, specifically including:

[0128] Determine the weighting coefficient W for each monitoring point i , where i takes values ​​in the range [1,2,3,...,i], and the factors considered include the structural importance of the measuring point location, the criticality of the monitoring parameters, and the reliability of historical data;

[0129] The weighted composite anomaly index I is calculated using the following formula: ;

[0130] The comprehensive anomaly index is normalized to conform to a standard normal distribution.

[0131] Trend analysis using the sliding time window technique includes:

[0132] Set the window length to m time steps, and determine the specific value based on the monitoring frequency;

[0133] Calculate the moving average, slope of change, and fluctuation range of the comprehensive anomaly index within the window;

[0134] Analyze the persistence and development trend of abnormal indices.

[0135] Optionally, the sensitivity parameter k can be implemented in the range of [3,8], and preferably, the preferred embodiment of the sensitivity parameter k is 4.

[0136] This invention addresses the issues of data reuse and insufficient timeliness in traditional methods through real-time data acquisition and status identification technology, achieving precise synchronization between monitoring data and simulated time steps, thus improving the freshness and accuracy of event identification. Through a multi-level anomaly index calculation system, it solves the problem of low reliability in single-parameter anomaly judgment, realizing a systematic anomaly assessment from the parameter level, measurement point level to the structure level, improving the comprehensiveness and accuracy of anomaly identification. By fusing standardized residuals and Mahalanobis distance, it achieves standardized quantification of multi-parameter anomaly severity, reducing false alarm and false negative rates. Through sliding time window trend analysis, it solves the problem of misjudging instantaneous fluctuations as structural damage, enabling continuous tracking of anomaly development trends and improving the reliability and stability of event identification. Finally, through a multi-condition joint judgment mechanism, it solves the problem of poor adaptability of single threshold judgment, realizing intelligent event identification based on statistical significance, and improving the scientific rigor and accuracy of damage event judgment.

[0137] S5. When a damage event is detected, the baseline simulation is paused, and damage correction is performed on the spatial finite element model based on the monitoring data to obtain a corrected finite element model characterizing the state after the damage event. Based on the corrected finite element model, a corrected structural equilibrium equation is constructed.

[0138] Damage correction is performed on the spatial finite element model based on monitoring data, including:

[0139] S51, Based on the monitoring data that triggers damage event identification and the spatial distribution of its abnormal index, identify the damaged area in the structure;

[0140] S52, within the damaged area, the residual between the dynamic characteristics or static response of the spatial finite element model and the measured data is minimized by weighted least squares method, thereby identifying and correcting the structural parameters of the damaged area.

[0141] S53. Based on the structural parameters of the corrected damaged area, the coordinates of all nodes of the spatial finite element model are corrected to obtain the corrected finite element model.

[0142] The monitoring data includes data collected by at least one of the following sensors: strain data collected by a fiber optic grating sensor, rotational angle data collected by an inclinometer, and vibration data collected by an accelerometer.

[0143] In one possible implementation, when a damage event is detected, the baseline simulation process is immediately paused, all computational states of the current simulation time step are saved, the timestamp of the damage event, the peak value of the anomaly index and its spatial distribution are recorded, the complete state data of the current spatial finite element model is backed up, and then a dedicated thread for model correction is started, while the main computation thread remains in a paused state.

[0144] Extract characteristic parameters of the injury event, including: the maximum abnormal index value and its location, the duration of the abnormal index exceeding the threshold, and the spatial distribution range parameters, and generate a preliminary assessment report of the injury event, including possible injury types and severity.

[0145] An improved DBSCAN spatial clustering algorithm was used for spatial clustering analysis of abnormal data, generating a heatmap of the spatial distribution of anomaly indices. The resolution was set to 0.5m × 0.5m. The optimal number of clusters was determined using the silhouette coefficient method to ensure that the clustering quality was greater than 0.7. The search radius of the improved DBSCAN spatial clustering algorithm was adaptively determined based on the average spacing between measurement points, and the minimum number of points was set according to the monitoring network density.

[0146] A damage region identification matrix is ​​established, the boundary of the damage region is determined by the convex hull algorithm, and the geometric features of each damage region are calculated, such as the region centroid coordinates, equivalent damage radius, and region area / volume.

[0147] S52a, determine the parameters to be corrected and the target response, specifically including:

[0148] Select the parameter set to be corrected θ=[E,γ,ɑ,β], where E is the elastic modulus, γ is Poisson's ratio, α is the stiffness reduction factor, and β is the mass density correction factor;

[0149] After a damage event, the measured response data of the damaged area and surrounding key measuring points are collected, mainly including static response, such as displacement and strain, and dynamic characteristics, such as vibration frequency and mode shape.

[0150] S52b, constructing the error minimization objective function, specifically including:

[0151] For each measured data point at each monitoring point, calculate the difference (i.e., residual) between it and the corresponding value calculated by the finite element model under the current parameters.

[0152] Different weights are assigned to different types of monitoring data based on their reliability and importance. For example, displacement data is considered more reliable than strain data, so displacement data is given a higher weight.

[0153] The weighted squared residuals of all monitoring points are summed; this sum is the objective function value that needs to be minimized. The smaller this value, the better the model matches the actual situation.

[0154] The S52c uses an iterative optimization algorithm to solve for the optimal parameters, specifically including:

[0155] A sequential quadratic programming optimization algorithm is employed. By solving an approximate quadratic programming subproblem in each iteration, the algorithm rapidly approximates the optimal solution. The calculation begins with a set of initial parameter guesses, which are either design values ​​or values ​​identified under healthy conditions. Based on the current parameters, the objective function value and its gradient are calculated. A search direction and step size that decreases the objective function are then determined, generating a new and better set of parameter values. The finite element model is updated using these new parameters, and the model response and objective function values ​​are recalculated. This iterative process is repeated until one of the following convergence criteria is met:

[0156] The rate of change of the objective function is small enough; for example, the change is less than 1% in two consecutive iterations, indicating that the optimization has reached a stable state.

[0157] If the change in the parameter itself is small enough, for example, less than 0.5%, it indicates that the parameter value has been basically determined.

[0158] Reach the preset maximum number of iterations, such as 50, to prevent infinite loops.

[0159] S52d output and verification, specifically including:

[0160] Once the iteration converges, the final set of optimal structure parameters is output.

[0161] Substitute the final optimal set of structural parameters into the finite element model and run a complete analysis to verify whether the model's calculated response, such as displacement and frequency, is in high agreement with the measured data, thus ensuring the effectiveness of the correction.

[0162] Due to the limited number of monitoring points, the displacement of each model node cannot be directly obtained. Therefore, it is necessary to calculate the continuous and smooth displacement field of the entire structure using mathematical methods based on the sparse displacement data of the monitoring points. Specifically, this includes:

[0163] Collect displacement data of all monitoring points after the damage event occurs, such as linear displacement and rotation angle;

[0164] The radial basis function interpolation method is used to process irregularly distributed scattered data. Each monitoring point is regarded as an influence source, and its displacement value spreads to the surrounding space. The closer the distance, the greater the influence. By superimposing the influence of all monitoring points, the displacement value of any node in the model can be calculated, thereby constructing a continuous three-dimensional displacement field covering the entire structure.

[0165] Once the complete displacement field is obtained, the geometry of the original model can be updated, specifically including:

[0166] The three directional displacement components of each node calculated in the first step are superimposed onto the original coordinate values ​​of that node. The new coordinates are then used to replace the node coordinates in the original finite element model in batches. At this point, the mesh topology of the model remains unchanged, but the geometry of the entire structure has changed. For example, the piers may have tilted, and the main beam may have experienced new deflection.

[0167] In regions of severe deformation, directly updating coordinates can lead to mesh distortion, affecting computational accuracy. Therefore, mesh optimization is necessary, specifically including:

[0168] Check the corrected model for any abnormally distorted elements, such as overly elongated triangular or quadrilateral elements.

[0169] If twisted elements are found, the Laplace smoothing algorithm is used for optimization. This algorithm makes the mesh distribution more uniform and reasonable by slightly moving the nodes towards the geometric center of their adjacent nodes, while maintaining the main deformation trend determined by the monitoring data.

[0170] After completing all the above operations, a corrected finite element model with a geometric configuration consistent with the actual shape of the structure after the damage event is obtained. This model provides an accurate physical basis for reconstructing the equilibrium equations and continuing the simulation. Through this systematic geometric correction process, it is ensured that the computational model is no longer an idealized initial design shape, but a digital model that can truly reflect the geometric characteristics of the structure under the current damage state.

[0171] Based on the modified finite element model, a modified structural equilibrium equation that simultaneously considers geometric nonlinearity and concrete shrinkage and creep effects is reconstructed.

[0172] This invention addresses the issues of ambiguous damage region location and inaccurate boundary definition by improving DBSCAN spatial clustering and heatmap visualization techniques, thereby enhancing the accuracy and spatial resolution of damage region identification. It also solves the problems of low efficiency in multi-source monitoring data fusion and slow optimization convergence during damage parameter identification by employing weighted least squares and sequential quadratic programming optimization algorithms, thus improving the accuracy and computational efficiency of structural parameter correction. Furthermore, it addresses the difficulty of supporting full-structure geometric correction with sparse monitoring data through radial basis function interpolation and node coordinate update techniques, improving the consistency between the finite element model's geometric configuration and the actual damage state. Finally, it resolves the issue of decreased computational accuracy due to mesh distortion after damage region geometric correction through Laplace smoothing mesh optimization, reducing numerical calculation errors and improving the reliability of subsequent simulations. Finally, it addresses the resource waste and insufficient accuracy issues caused by traditional overall model correction through precise damage region location and local parameter correction techniques, reducing computational complexity and improving the targeting and efficiency of model correction. Finally, it utilizes multi-type monitoring data fusion correction techniques to address the low reliability and weak anti-interference ability of correction from a single data source, improving the robustness and comprehensiveness of damage correction.

[0173] S6. Based on the modified finite element model and the modified structural equilibrium equation, perform a baseline simulation again until no damage event is detected.

[0174] In one possible implementation, the complete data of the modified finite element model in S5 is called, including the modified node coordinates, structural parameters of the damaged area, such as elastic modulus E, Poisson's ratio γ, stiffness reduction factor α, mesh topology, and optimized element shape; the modified structural equilibrium equations based on the modified model, which consider both geometric nonlinearity and concrete shrinkage and creep effects, are loaded simultaneously to ensure that the tangent stiffness matrix and equivalent nodal force terms in the equations match the parameters of the modified model.

[0175] Restore the simulation state when S5 was paused, including the current time step number, the duration of the completed simulation, the equivalent nodal force data of concrete shrinkage and creep in the previous time step, and the historical records of structural displacement and end force. This ensures the continuity of the simulation process and avoids the break in the time axis and mechanical state caused by model correction.

[0176] Starting from the paused time step in S5, proceed sequentially according to the preset time step length, and record the actual time interval corresponding to each current time step. For example, if the pause is in the 10th time step of the 3rd year of operation, the corresponding actual time is the 270th to 300th day of the 3rd year, then the next time step is the 300th to 330th day of the 3rd year.

[0177] If a new damage event is identified, the current correction simulation is immediately paused, and the damage correction process of S5 is repeated: new damage areas are identified based on new monitoring data and the spatial distribution of the anomaly index, the structural parameters of the area are corrected by weighted least squares, the model node coordinates are updated and the mesh is optimized, and the corrected structural equilibrium equations are reconstructed; after the correction is completed, return to this step and restart the simulation from the current paused time step.

[0178] If no damage event is identified, archive all data from the entire correction simulation phase, including the structural mechanical state at each time step, the history of equivalent nodal forces for shrinkage and creep, the time of damage event occurrence and damage correction records, monitoring data, and anomaly index change curves.

[0179] S7 outputs the long-term deformation and internal force prediction results of the bridge at the target age, including: the time-varying deformation envelope diagram of the whole bridge, the time history of internal forces at key sections, and the long-term stress distribution cloud map.

[0180] In one possible implementation, complete time series data from the start of construction to the target prediction period, such as 50 years, is extracted, including the simulation timestamp of each time step and the corresponding structural mechanical state calculation results; displacement data of all nodes of the entire bridge, internal force data of key sections, and stress data of all elements of the entire bridge are extracted. At the same time, abnormal data caused by non-convergence of iterations during the simulation process are removed, and linear interpolation is used to fill in missing time steps to ensure data continuity, and all data units are unified and standardized.

[0181] Envelope graph modeling and visualization, specifically including:

[0182] Based on the axis projection of the three-dimensional finite element model of the bridge, the coordinates of all bridge nodes are imported into CAD or BIM software to establish a simplified model consistent with the actual bridge geometry. For each time interval, the maximum / minimum displacement values ​​of the nodes in the corresponding interval are superimposed on the simplified model. The displacement extreme points of adjacent nodes are connected by cubic spline interpolation to form the deformation envelope of the interval. The displacement extreme values ​​and corresponding times of key nodes are marked in the figure, and the displacement unit and time interval division legend are marked.

[0183] Time history generation of internal forces at key sections, specifically including:

[0184] Using Origin software, with the operational years as the horizontal axis and internal force values ​​as the vertical axis, the range is set according to the type of internal force, such as the bending moment range of -5000~5000kN. m, plot three time history curves for each key section, namely axial force, bending moment and shear force, and use different colors to distinguish different types of internal forces, such as red for bending moment, blue for axial force and green for shear force;

[0185] Mark feature points on the curve, including:

[0186] Peak internal forces, such as the maximum bending moment during the closure phase of construction, and the corresponding time;

[0187] Internal force mutation points, such as nodes where internal force suddenly increases / decreases after a damage event, and event descriptions;

[0188] Long-term stable value, that is, the range of values ​​within which the internal force tends to stabilize in the later stages of operation;

[0189] Plot the internal force time history curves of similar key sections on the same graph, compare and analyze the differences in internal forces at different locations, add standard limit lines to the graph, and intuitively judge whether the internal forces at each time node meet the target requirements.

[0190] The generation of long-term stress distribution cloud maps includes:

[0191] The stress data of all elements of the entire bridge are grouped according to time nodes;

[0192] The radial basis function interpolation method is used to interpolate discrete element stress data onto a continuous spatial grid of the entire bridge, generating full-bridge stress field data for each key time node;

[0193] In finite element analysis software such as ANSYS or ABAQUS, import the corrected full-bridge finite element model and map the interpolated stress field data onto the corresponding elements and nodes of the model.

[0194] A blue-green-yellow-red gradient color mapping scheme is adopted to set the correspondence between stress values ​​and colors;

[0195] Mark stress concentration areas in the cloud map, such as the bottom of the pier, the connection between the main beam and the support, and the parts where the stress value exceeds 30% of the surrounding area. Mark them with red circles and number them.

[0196] Add text descriptions below the cloud map, including:

[0197] Stress distribution characteristics at key time points, such as the main beam's mid-span bottom compressive stress stabilizing at -6 to -8 MPa after 10 years of operation;

[0198] Analysis of the causes of stress concentration areas, such as stress concentration at the bottom of bridge piers caused by uneven foundation settlement;

[0199] Maintenance recommendations include conducting non-destructive testing on stress concentration area number 1 every two years.

[0200] This invention uses a time-varying deformation envelope diagram of the entire bridge to visually present the maximum or minimum displacement extremes and deformation trends of each node of the entire bridge at different stages from the construction period to the target lifespan. This solves the problem of scattered deformation data in traditional text reports, making it difficult for engineers to quickly grasp the overall deformation pattern. It significantly improves the efficiency of visually understanding the spatial distribution characteristics and temporal evolution of bridge deformation over long term, providing direct evidence for judging whether the bridge has risks such as excessive deflection or lateral displacement, and helping maintenance personnel to formulate targeted deformation control measures in advance. The time history of internal forces at key sections clearly shows the dynamic changes of axial force, bending moment, and shear force in the core stress-bearing parts throughout the entire life cycle, accurately marking the peak points, abrupt change points, and long-term stable values ​​of internal forces. This solves the problem that traditional methods cannot quantify the fluctuation law of internal forces at key sections over time, allowing engineers to quickly identify abnormal changes in internal forces. At the same time, by comparing with the standard limit line, it can directly determine whether the internal forces at each stage meet the design requirements, providing quantitative data support for the selection of reinforcement timing for key sections and the assessment of material fatigue life, thus improving the scientificity and accuracy of maintenance decisions.

[0201] Example 2

[0202] like Figure 2 As shown, the present invention also provides a long-term deformation prediction system for long-span rigid frame bridges. The system uses any of the long-term deformation prediction methods for long-span rigid frame bridges described in Example 1, specifically including:

[0203] The model building module is used to build a spatial finite element model based on the bridge's design drawings, material properties, and boundary conditions.

[0204] The equation construction module, connected to the model building module, is used to construct structural equilibrium equations that simultaneously consider geometric nonlinearity and concrete shrinkage and creep effects based on the spatial finite element model.

[0205] The cyclic simulation module, connected to the equation construction module, is used to perform a benchmark simulation on the spatial finite element model based on the structural equilibrium equations to determine the evolution of the benchmark mechanical state from the start of construction to the target predicted lifespan.

[0206] An event recognition module, connected to the cyclic simulation module, is used to identify damage events based on a pre-established baseline probability distribution model.

[0207] The event correction module, connected to the event identification module, is used to pause the baseline simulation when a damage event is identified, and perform damage correction on the spatial finite element model based on the monitoring data to obtain a corrected finite element model characterizing the state after the damage event, and construct a corrected structural equilibrium equation based on the corrected finite element model.

[0208] The correction simulation module, connected to the event correction module, is used to re-perform the baseline simulation based on the corrected finite element model and the corrected structural equilibrium equation;

[0209] The result prediction module, connected to the correction simulation module, is used to output the long-term deformation and internal force prediction results of the bridge at the target age.

[0210] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A method for predicting the long-term deformation of a long-span rigid frame bridge, characterized in that, include: S1. Based on the bridge's design drawings, material properties, and boundary conditions, a spatial finite element model is established. S2, Based on the aforementioned spatial finite element model, construct a structural equilibrium equation that simultaneously considers geometric nonlinearity and concrete shrinkage and creep effects; S3. Based on the structural equilibrium equation, a benchmark simulation is performed on the spatial finite element model to determine the evolution of the benchmark mechanical state from the start of construction to the target predicted lifespan. S4 identifies damage events based on a pre-established baseline probability distribution model; S5. When a damage event is detected, the baseline simulation is paused, and damage correction is performed on the spatial finite element model based on the monitoring data to obtain a corrected finite element model characterizing the state after the damage event. Based on the corrected finite element model, a corrected structural equilibrium equation is constructed. S6. Based on the modified finite element model and the modified structural equilibrium equation, perform a baseline simulation again until no damage event is detected. S7 outputs the predicted results of long-term deformation and internal forces of the bridge at the target age; The damage events include at least vehicle collisions and earthquakes.

2. The method for predicting long-term deformation of long-span rigid frame bridges according to claim 1, characterized in that, The benchmark simulation on the spatial finite element model includes: S31, starting from the construction start point, proceeds step by step in chronological order to the target prediction year, dividing the total time into continuous time steps; S32, execute the baseline simulation loop within each current time step; S33, advance the time to the next time step, and repeat S32 until the simulation time reaches the target prediction year.

3. The method for predicting long-term deformation of long-span rigid frame bridges according to claim 2, characterized in that, The execution benchmark simulation loop includes: S32a, applies the concrete shrinkage and creep effects up to the previous time step as equivalent nodal forces to the current structural configuration; S32b, Solve the structural equilibrium equations, and update the structural displacements and end forces of the members through iterative calculations until the solution converges; S32c, based on the updated rod end force, calculates the shrinkage and creep equivalent nodal force for the next time step by fitting a recursive formula for the creep coefficient based on an exponential function.

4. The method for predicting long-term deformation of long-span rigid frame bridges according to claim 1, characterized in that, The establishment of the baseline probability distribution model includes: Collect historical monitoring data of the bridge under normal operating conditions to form a training sample set; Based on the training sample set, a baseline probability distribution model for each mechanical state monitoring parameter is established and its statistical characteristic values ​​are determined. Based on hypothesis testing methods, upper and lower control limits are set for the abnormal index used to identify damage events.

5. The method for predicting long-term deformation of long-span rigid frame bridges according to claim 1, characterized in that, The identification of damage events based on a pre-established baseline probability distribution model includes: S41, acquire monitoring data that has not been involved in identification and judgment within the actual time interval corresponding to the current time step, and use it as new monitoring data; S42, Calculate the comprehensive anomaly index of the new monitoring data based on the benchmark probability distribution model; S43, using the sliding time window technique, analyze the changing trend of the comprehensive anomaly index within continuous time steps; S44, when the comprehensive abnormality index is identified to exceed the control upper limit for k consecutive time steps, or to be below the control lower limit for k consecutive time steps, a damage event is determined to have occurred; Where k is a preset sensitivity parameter.

6. The method for predicting long-term deformation of long-span rigid frame bridges according to claim 5, characterized in that, The calculation of the anomaly index of the new monitoring data based on the baseline probability distribution model includes: S42a, For each new monitoring data point, calculate its standardized residual relative to the baseline probability distribution model; S42b, Based on the standardized residuals, the local anomaly index of each monitoring data is obtained by calculating Mahalanobis distance; S42c, based on the spatial distribution characteristics of the monitoring data, weights and fuses each local anomaly index to form a comprehensive anomaly index of the overall structure.

7. The method for predicting long-term deformation of long-span rigid frame bridges according to claim 1, characterized in that, The damage correction based on monitoring data on the spatial finite element model includes: S51, Based on the monitoring data that triggers damage event identification and the spatial distribution of its abnormal index, identify the damaged area in the structure; S52, within the damaged area, the residual between the dynamic characteristics or static response of the spatial finite element model and the measured data is minimized by weighted least squares method, thereby identifying and correcting the structural parameters of the damaged area. S53. Based on the structural parameters of the corrected damaged area, the coordinates of all nodes of the spatial finite element model are corrected to obtain the corrected finite element model.

8. The method for predicting long-term deformation of long-span rigid frame bridges according to claim 1, characterized in that, The monitoring data includes data collected by at least one of the following sensors: strain data collected by a fiber optic grating sensor, rotational angle data collected by an inclinometer, and vibration data collected by an accelerometer.

9. The method for predicting long-term deformation of long-span rigid frame bridges according to claim 1, characterized in that, The predicted results of the bridge's long-term deformation and internal forces over the target life include: time-varying deformation envelope diagram of the entire bridge, time history of internal forces at key sections, and long-term stress distribution cloud map.

10. A long-term deformation prediction system for long-span rigid frame bridges, characterized in that, The system employs the long-term deformation prediction method for long-span rigid frame bridges as described in any one of claims 1 to 9, specifically including: The model building module is used to build a spatial finite element model based on the bridge's design drawings, material properties, and boundary conditions. The equation construction module, connected to the model building module, is used to construct structural equilibrium equations that simultaneously consider geometric nonlinearity and concrete shrinkage and creep effects based on the spatial finite element model. The cyclic simulation module, connected to the equation construction module, is used to perform a benchmark simulation on the spatial finite element model based on the structural equilibrium equations to determine the evolution of the benchmark mechanical state from the start of construction to the target predicted lifespan. An event recognition module, connected to the cyclic simulation module, is used to identify damage events based on a pre-established baseline probability distribution model. The event correction module, connected to the event identification module, is used to pause the baseline simulation when a damage event is identified, and perform damage correction on the spatial finite element model based on the monitoring data to obtain a corrected finite element model characterizing the state after the damage event, and construct a corrected structural equilibrium equation based on the corrected finite element model. The correction simulation module, connected to the event correction module, is used to re-perform the baseline simulation based on the correction finite element model and the correction structural equilibrium equation until no damage event is detected. The result prediction module, connected to the correction simulation module, is used to output the long-term deformation and internal force prediction results of the bridge at the target age.

Images