A method for intelligent monitoring of deformation and posture correction of a cable-stayed bridge body during construction

Abstract

This invention relates to the field of bridge engineering construction monitoring and structural control technology, specifically a method for intelligent monitoring and attitude correction of deformation throughout the construction process of a cable-stayed bridge. The method includes: dynamically collecting and integrating spatial coordinates, surface strain, environmental conditions, and construction procedure status data; characterizing the total structural deformation based on spatial coordinates and decoupling to obtain elastic transient deformation and plastic permanent deviation; inputting the plastic permanent deviation into a finite element analysis model to obtain stress redistribution state data within the structure; performing multi-objective optimization based on the stress redistribution state data to generate correction strategy instructions and an optimal time window. This invention achieves coordinated control of geometry and stress during bridge construction, reducing the risks of miscorrection, overcorrection, and residual stress locking.

Description

Technical Field

[0001] This invention relates to the field of bridge engineering construction monitoring and structural control technology, specifically a method for intelligent monitoring of deformation and attitude correction throughout the construction process of a cable-stayed bridge. Background Technology

[0002] Cable-stayed bridges are widely used in the construction of long-span bridges. Their construction process is characterized by multiple stages of transition, sensitive structural systems, and significant environmental impact. To ensure the accuracy and safety of the cantilever assembly of the main beam, the phased tensioning of the stay cables, the construction of the steel-concrete composite section, and the closure control, it is usually necessary to continuously monitor the bridge's alignment, attitude, and stress state throughout the entire construction process, and to make timely adjustments to deviations based on the monitoring results in order to avoid the cumulative impact of construction errors on the final bridge condition. However, when monitoring deformation and correcting attitude during the construction of cable-stayed bridges, existing methods generally focus on direct comparison of spatial coordinates or local measuring point data, requiring cable adjustment or jacking based on the measured geometric deviations. Bridge geometric deviation is an important indicator for measuring the construction status, which is directly related to the subsequent closure accuracy and structural safety. To balance construction control requirements and on-site implementation efficiency, the amount of correction is usually determined primarily by linear deviation. The correction determination can be understood as: total structural deformation = elastic transient deformation caused by environmental disturbance + plastic permanent deviation caused by construction procedures. Although this control method can reflect changes in apparent posture, it is difficult to distinguish between recoverable elastic transient deformation caused by environmental factors such as temperature difference, sunlight, and wind load, and plastic permanent deviation caused by segmental assembly errors, uneven cable force adjustment, and temporary constraint release. It also lacks the ability to further map geometric deviations into stress redistribution within the structure and make optimization decisions in conjunction with the timing of operations. This limitation not only easily leads to incorrect or over-correction, but may also cause residual stress to accumulate within the structure, adversely affecting the construction control and the quality of the completed cable-stayed bridge. Summary of the Invention

[0003] To address the aforementioned technical problems, this invention provides a method for intelligent monitoring and attitude correction of deformation throughout the entire construction process of a cable-stayed bridge. Specifically, the technical solution of this invention includes: Step S1 involves dynamically collecting and integrating spatial coordinate data, surface strain data, environmental condition data, and construction procedure status data of the cable-stayed bridge. Step S2: The total structural deformation is obtained based on spatial coordinate data, and the total structural deformation is decoupled based on the integrated data. The decoupling includes calculating the elastic transient deformation and calculating the plastic permanent deviation. Step S3: Based on the pre-built finite element analysis model, finite element analysis calculations are performed using plastic permanent deviation to obtain stress redistribution state data inside the structure. Step S4: Perform multi-objective optimization based on stress redistribution state data, generate correction strategy instructions, and generate the optimal time window by combining environmental prediction data. Step S5: Based on the correction strategy instructions, call the front-end construction control equipment to perform physical correction actions within the optimal time window.

[0004] Preferably, the spatial coordinate data, joint surface strain data, environmental condition data, and construction procedure condition data are preprocessed to obtain preprocessed data. The preprocessing includes missing value imputation, outlier detection, and data standardization. Data standardization uses extreme value normalization or standard deviation standardization methods to uniformly dimensionless process the data to eliminate dimensional differences.

[0005] Preferably, the calculation of elastic transient deformation includes: calculating the structural temperature gradient distribution caused by environmental state data using a thermodynamic heat transfer model based on Fourier's law of thermal conductivity; using a time series analysis network containing a long short-term memory network or a recurrent neural network, taking the structural temperature gradient distribution sequence within a preset historical time window, the wind speed and direction environmental state sequence, and the pre-established and stored historical total structural deformation data as joint inputs of multi-dimensional features, and outputting the expected value of elastic deformation under pure environmental driving as the fitting analysis result; and combining the structural temperature gradient distribution and the fitting analysis result to extract the elastic transient deformation caused by environmental state data.

[0006] Preferably, the calculation of plastic permanent deviation includes: using the total structural deformation represented by spatial coordinate data as a reference input; using the total structural deformation to eliminate elastic transient deformation, separating the plastic permanent deviation caused by the construction process state data, so as to use the plastic permanent deviation for finite element analysis calculation.

[0007] Preferably, a structural mechanical reference state is constructed based on a finite element analysis model; the plastic permanent deviation is input into the finite element analysis model as a boundary displacement condition for real-time calculation; the stress tensor output by the finite element analysis model is extracted to generate stress redistribution state data, and the stress redistribution state data is rendered as stress cloud map data for display by the front-end monitoring equipment.

[0008] Preferably, a multi-objective optimization algorithm is constructed, and the objective function of the multi-objective optimization algorithm is set as minimizing the overall stress variance of the structure and minimizing the stress extrema of the key stress nodes of the bridge; the stress redistribution state data is input into the multi-objective optimization algorithm based on genetic algorithm or particle swarm optimization algorithm for iterative optimization; the correction strategy instruction that satisfies the objective function is output, the correction strategy instruction includes the cable tension adjustment value and the jack pushing displacement value, and the optimal time window is generated in combination with environmental prediction data.

[0009] Preferably, the generation of the optimal time window includes: acquiring meteorological forecast data for a future preset time period to predict future environmental state data; calculating the elastic transient deformation value caused by the future environmental state data; determining the corresponding time period as the optimal time window when the elastic transient deformation value is lower than a preset deformation threshold determined based on the bridge design tolerance error; determining the corresponding time period as a candidate time window when the elastic transient deformation value is equal to the preset deformation threshold; and determining the corresponding time period as an unworkable time window when the elastic transient deformation value is higher than the preset deformation threshold.

[0010] Preferably, the method further includes closed-loop feedback control based on the correction strategy command. The closed-loop feedback control includes: executing physical correction actions within the optimal time window according to the correction strategy command; real-time acquisition of updated spatial coordinate data and updated interface strain data after executing the physical correction actions; and re-execution of dynamic acquisition and integration, structural total deformation decoupling, finite element analysis calculation, and multi-objective optimization using the updated spatial coordinate data and updated interface strain data to conduct a new round of dynamic monitoring and attitude correction.

[0011] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention effectively solves the problems of different sampling frequencies, large dimensional differences, short-term missing data, and abnormal disturbances that can easily affect the reliability of judgments by dynamically collecting, preprocessing, and integrating the spatial coordinate data, surface strain data, environmental condition data, and construction procedure status data of cable-stayed bridges under a unified time reference. 2. This invention first characterizes the total structural deformation, then combines temperature gradient analysis based on a thermodynamic heat transfer model with time series analysis network to fit the historical deformation patterns. This effectively separates recoverable elastic transient deformation caused by sunlight, temperature difference, and wind field from the total deviation, and further separates plastic permanent deviation caused by construction procedures such as segment assembly error, cable force application sequence, unloading redistribution, and temporary constraint release. This avoids misjudging environmental transient response as real construction error. 3. This invention uses the plastic permanent deviation as the boundary displacement condition input to a finite element analysis model that matches the construction stage, extracts the internal stress tensor of the structure and generates stress redistribution state data and stress cloud map data, so that the system can further map the apparent geometric deviation into potential lock-in stress and local stress concentration risk inside the structure. 4. This invention constructs a multi-objective optimization mechanism aimed at minimizing the overall structural stress variance and the stress extrema of key stress nodes. By combining the cable tension adjustment values, jack displacement values, and future weather forecasts to select the optimal time window, it can avoid miscorrection and overcorrection caused by simply pursuing alignment restoration. Ultimately, it achieves coordinated control of the bridge attitude and internal stress state throughout the construction of the cable-stayed bridge. Through closed-loop feedback, it continuously corrects the deviation effect, significantly improving the closure control accuracy, construction safety, and bridge quality. Attached Figure Description

[0012] The present invention will be further explained below with reference to the accompanying drawings and embodiments: Figure 1 This is a flowchart illustrating a method for intelligent deformation monitoring and attitude correction throughout the construction process of a cable-stayed bridge, as provided in an embodiment of this application. Detailed Implementation

[0013] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0014] A method for intelligent monitoring and attitude correction of deformation throughout the construction process of a cable-stayed bridge includes: Step S1 involves dynamically collecting and integrating spatial coordinate data, surface strain data, environmental condition data, and construction procedure status data of the cable-stayed bridge. Step S2: The total structural deformation is obtained based on spatial coordinate data, and the total structural deformation is decoupled based on the integrated data. The decoupling includes calculating the elastic transient deformation and calculating the plastic permanent deviation. Step S3: Based on the pre-built finite element analysis model, finite element analysis calculations are performed using plastic permanent deviation to obtain stress redistribution state data inside the structure. Step S4: Perform multi-objective optimization based on stress redistribution state data, generate correction strategy instructions, and generate the optimal time window by combining environmental prediction data. Step S5: Based on the correction strategy instructions, call the front-end construction control equipment to perform physical correction actions within the optimal time window.

[0015] This embodiment provides a mechanism for intelligent deformation monitoring and attitude correction throughout the construction process of a cable-stayed bridge. Specifically, taking the stage from the cantilever assembly of the main beam of a double-tower steel-concrete composite cable-stayed bridge to the closure stage as a unified engineering scenario, the external geometric attitude and internal stress evolution of the bridge are synchronously identified when the side spans and middle spans are hoisted segment by segment, the stay cables are tensioned in batches, and the steel-concrete composite section enters the key bridge completion control stage. This avoids the accumulation of locking stress caused by forcibly correcting the deviation based solely on the alignment. In detail, step S1 is not simply collecting data from multiple measuring points, but rather establishing a construction status snapshot with temporal consistency and spatial correspondence around the same bridge body; among them, spatial coordinate data reflects the three-dimensional positional changes of control points near the beam end, cable anchor area, tower-beam connection area, and closure joint; combined with surface strain data, it reflects the tensile and compressive transfer of the connection area between the steel beam and the concrete bridge deck and web under the combined action of external forces and temperature differences. Environmental status data reflects solar radiation, air temperature, wind speed and direction, and the temperature difference between upstream and downstream sections; construction process status data reflects the current segment hoisting, temporary support status, cable tensioning sequence, and jack pushing conditions; the system associates this information according to a unified time base. Specifically, it uses the data sequence with the lowest sampling frequency as the base timestamp and employs time window matching or linear interpolation algorithms to resample and align high-frequency sampling data and discretized status data, forming a comprehensive panoramic status representation that reflects the bridge's spatial attitude, environmental conditions, and ongoing process status at the same moment. In step S2, the total deformation of the structure is first characterized using spatial coordinate data, and then the total deformation is divided into recoverable elastic transient parts and plastic permanent parts that need to be corrected. This step aims to identify the source of deviation, rather than treating all deviations as defects. After the bridge is exposed to sunlight during the day, the top plate of the steel box girder usually heats up faster than the bottom plate, and a temperature gradient will be generated in the cross section. The beam ends will arch or deflect laterally. These changes can partially subside after the temperature is uniform at night. They are environmentally driven elastic transient deformations. In contrast, factors such as cumulative errors in segmental assembly, improper order of cable force application, and structural redistribution after crane unloading can cause certain deviations to persist even after environmental recovery. These changes correspond to permanent plastic deviations. The role of the system is to avoid misjudging the former type of transient response as the latter type of construction error. In step S3, the identified plastic permanent deviation is input into a pre-established finite element analysis model. This model corresponds to the structural form of the main beam, tower, stay cables, and steel-concrete composite zone of the actual bridge type, and sets boundary constraints and force transfer relationships consistent with the construction stage. The model calculation does not yield a single displacement result, but rather data on the stress redistribution state of the bridge's internal structure, reflecting which component, which material interface, and which group of stay cables is experiencing unbalanced internal force adjustments. Through this conversion, the system transforms single geometric displacement deviation data into potential consolidation stress data within the structure that would be triggered if not corrected in time. In step S4, the system generates correction strategy instructions based on stress redistribution state data and generates an optimal time window by combining environmental prediction data. This correction process no longer uses simply restoring each control point to the design coordinates as the single optimization objective, but comprehensively considers the overall stress uniformity, the stress safety margin of key nodes, and the feasibility of construction to determine which cable should be adjusted first, by how much, whether it is necessary to use jacks for pushing, and when to implement it to meet the highest safety margin requirements. If it is predicted that the temperature difference is small, the wind speed is low, and the transient disturbance of the structure is weak during a certain period, the system will prioritize recommending to perform correction during that period to reduce additional external force input and reduce the risk of residual stress lock-in. As an anomaly handling mechanism, if a certain type of sensor data is temporarily missing in the current cycle, such as a local strain sensor going offline, a short-term anemometer distorting, or a single coordinate measuring point being obstructed, the system does not directly output a correction command. Instead, it first uses other adjacent measuring points and the trend of previous and subsequent times to perform a status check. If the current structural status still cannot be confirmed after the check, only the monitoring and early warning functions are retained, and the active correction suggestion is suspended. Uncertainty measurement is used to prevent the output of abnormal correction commands. If a sudden change occurs in the construction process, such as the early removal of temporary supports, abnormal shutdown of lifting equipment, or interruption of cable force application, the correction strategy generated in the current cycle will automatically become invalid. The system needs to re-collect the complete status and re-evaluate to prevent the application of outdated decisions to the changed structure. Two days before the mid-span closure of a steel-concrete composite cable-stayed bridge, the system continuously monitored a vertical height difference of several millimeters between the beam ends on both sides of the closure point. At the same time, several control points on both sides of the pylons showed transverse bridge sway in the afternoon. If only the geometric measurement results were used, the site might immediately propose cable adjustment and jacking schemes. However, this embodiment first included the factors of strong sunlight, large bridge deck temperature difference, and southeast gusts in the analysis, and identified that part of the deviation was due to afternoon thermal expansion and wind-induced response, which would naturally subside in the evening. The remaining part that did not subside was then entered into structural stress analysis. Finally, it was determined that what really needed to be corrected were the cumulative attitude error of a group of segments near the closure point and the uneven tension of adjacent cable groups. It was recommended to implement the correction when the temperature difference converged at night. The purpose of this step is to upgrade bridge construction control from simple geometric alignment to coordinated control of geometry and stress, thereby enabling the differentiation and handling of transient environmental disturbances and permanent construction deviations, reducing the risks of miscorrection, overcorrection, and the resulting hidden residual stress.

[0016] Furthermore, before step S2, the method further includes: preprocessing the spatial coordinate data, the surface strain data, the environmental condition data, and the construction procedure condition data to obtain preprocessed data. The preprocessing includes missing value filling, outlier detection, and data standardization. Data standardization includes using extreme value normalization or standard deviation standardization methods to uniformly dimensionless process the data in order to eliminate dimensional differences.

[0017] This embodiment provides a mechanism for preprocessing construction monitoring data. Specifically, in the above-mentioned key construction scenarios before closure, the bridge monitoring data comes from measurement robots, satellite positioning stations, fiber optic strain sensors, temperature acquisition units, wind speed and direction instruments, and construction equipment control terminals. The sampling frequency, physical dimensions, and on-site noise characteristics of various types of data are different. If preprocessing is not performed, subsequent deformation peeling and stress identification may easily mistake sensor failures for structural anomalies. In detail, missing value filling mainly targets the common problem of breakpoints in construction sites where the duration is less than the preset time judgment threshold. The time judgment threshold is set comprehensively based on the average duration of a single construction process of the bridge and the stability cycle of the structural state to distinguish between short-term equipment obstruction and long-term sensor failure. For example, when hoisting equipment passes by, it obstructs the visual measurement point, or a strain channel is briefly interrupted during the vibration of pouring. The system prioritizes using the change trend of the same measurement point in consecutive time periods before and after to fill the missing value. If the duration of the breakpoint is greater than or equal to the preset time judgment threshold, then the constraint is completed by combining the adjacent measuring points, other sensors in the same section, and the synchronous construction status. The physical basis is that the bridge attitude evolution is continuous in the same segment and the same time period, and there will be no non-physical attribute step signal under the condition of sudden change without external excitation. Outlier detection is primarily used to identify measurement results that are inconsistent with the actual engineering mechanism. For example, when the temperature rises slowly, a strain sensor may suddenly show an isolated spike, and there may be no corresponding change at adjacent measuring points. This is more likely due to loose adhesion, optical path disturbance, or electromagnetic interference, rather than extreme strain exceeding the material's allowable limit at that location. Similarly, if the wind speed is stable and the process remains unchanged, and a single coordinate point shows an abnormal displacement that exceeds the allowable range of displacement deviation between adjacent control points, this is more likely due to the measuring reflector being hit or the instrument's line of sight being refracted. This is a non-physical artifact introduced by the measurement system. The system will mark such data that does not match the continuous stress law of the structure as anomalies and remove or reduce their weight before proceeding to subsequent analysis. Data standardization is used to eliminate the interference of dimensional differences between different physical quantities on the comprehensive analysis. Bridge coordinate offsets are typically on the order of millimeters, strain is on the order of micro-strain, temperature is on the order of degrees Celsius, and cable forces and hoisting loads are on the order of kilonewtons. If these quantities are directly mixed, larger physical quantities may mask more sensitive state information during the calculation process. Therefore, this embodiment uses an extreme value normalization method to normalize the data. Specifically, for any raw data item in a certain type of monitoring data sequence Map it to The interval is used to obtain the data after extreme value normalization. The calculation formula is: in, and These represent the minimum and maximum values ​​of this type of monitoring data sequence within the current analysis period, respectively; or, using the standard deviation standardization method, the data is transformed into a mean. Standard deviation is The distribution of the standard deviation is used to obtain the data. ,Right now: in, This represents the sample mean of this type of monitoring data sequence within the current analysis period. The standard deviation is the sample standard deviation. Through the above data standardization process, the system maps various types of data to comparable scales, so that the subsequent decoupling process focuses on the trend of change and relative contribution, rather than being dominated by the unit size. As an anomaly handling mechanism, if a critical measuring point is continuously missing for a duration exceeding the judgment window allowed by the current construction procedure, such as the main control point of the closure section being offline continuously during the entire critical operation, the system will stop providing automatic correction suggestions for this area, only issue a status prompt indicating insufficient data, and suggest that the site be transferred to manual review; if the outlier is not an isolated point, but a cluster of adjacent measuring points in the region simultaneously collects abnormal signals, the system will not directly identify it as an equipment failure, but will also need to check the construction log to determine whether there are synchronous unloading, segmental impact, or strong gust events, to prevent the real emergency conditions from being mistakenly filtered out; On the eve of the mid-span closure, a beam transport device passed through the visual measurement channel above the bridge deck, causing two coordinate points near the closure point to lose three consecutive sampling cycles. At the same time, a single-point spike appeared in a fiber optic strain channel near the steel-concrete joint section, while the other sensors in the same section remained stable. The system used information from adjacent time points and adjacent measurement points to complete the coordinate sequence, identified the strain spike as a suspected measurement interference, and reduced its impact in subsequent analysis. After dimensionless processing, environmental temperature difference, displacement change, and strain response can be judged together under the same state framework. The purpose of this step is to provide a reliable, consistent, and comparable data foundation for subsequent total deformation identification, environmental disturbance stripping, and stress analysis, thereby enabling robust handling of field monitoring noise, temporary obstruction, and heterogeneous data differences.

[0018] Furthermore, the calculation of elastic transient deformation includes: calculating the structural temperature gradient distribution caused by environmental state data using a thermodynamic heat transfer model based on Fourier's law of thermal conductivity; using a time series analysis network containing a long short-term memory network or a recurrent neural network, taking the structural temperature gradient distribution sequence within a preset historical time window, the wind speed and direction environmental state sequence, and the pre-established and stored historical total structural deformation data as joint inputs of multi-dimensional features, and outputting the expected value of elastic deformation under pure environmental driving as the fitting analysis result; combining the structural temperature gradient distribution and the fitting analysis result, extracting the elastic transient deformation caused by environmental state data.

[0019] This embodiment provides a mechanism for calculating elastic transient deformation. Specifically, several days before the bridge closure, the main beam of the bridge is exposed to solar radiation, diurnal temperature differences, and crosswinds for a long time. The heating rates of the outer wall of the steel box girder and the concrete slab are different, and the temperature rise of the sun-facing side and the shaded side of the bridge deck is also inconsistent. These environmental factors will cause reversible deformation of the beam. In order to avoid misjudging this environmentally induced deformation with self-recovery characteristics as assembly error, this embodiment combines thermodynamic heat transfer analysis with historical deformation evolution law to extract the environment-driven elastic transient deformation. In detail, the thermodynamic heat transfer model is used to establish the correspondence between the environment and the temperature gradient of the cross section. Its purpose is that the bridge does not heat up uniformly as a whole, but will form internal temperature differences due to differences in the thermal conductivity of the materials, plate thickness, surface orientation, and air cooling conditions. For example, the top plate heats up faster after being exposed to sunlight, while the bottom plate heats up slower due to different shading and airflow conditions. At this time, the beam may arch or twist. Between the side spans and the middle span of a cable-stayed bridge, the heating rates of different segments vary due to differences in cable surface shielding and ventilation conditions. By analyzing the environmental conditions and the structural temperature field, it is possible to determine which components in the current total deformation are consistent with the heat conduction and convection heat dissipation processes. Time series analysis networks are used to learn the deformation response habits of bridges during historical construction and environmental cycles; their role is not to replace structural mechanisms, but to identify repeatable deformation patterns of bridges under similar temperatures, wind fields and construction rhythms. Specifically, to clarify the state transition mechanism and its engineering physical significance within the time series analysis network, the network's data flow and computation rules are structurally decomposed as follows: The network's input layer receives multi-dimensional time series features, including the cross-sectional temperature distribution sequence, wind speed and direction sequence, and the corresponding historical total structural deformation sequence within a set historical time window; the output layer corresponds to the predicted elastic deformation expectation value under purely environmental driving conditions. Before the network is put into real-time prediction, the network is pre-trained using a bridge monitoring dataset under similar historical environmental conditions. The mean square error between the expected value of elastic deformation predicted by the network and the actual recoverable deformation value after the historical measured deformation data falls back to the baseline is used as the loss function. The weight parameters of the hidden layer state units are updated through the backpropagation algorithm until the loss function converges to the preset accuracy range. During the fitting analysis, the network extracts the influence weights of environmental periodic fluctuations on the total structural deformation through the state units in the hidden layer at each time step and retains the long-term dependency relationship. For example, when the temperature difference fluctuates periodically over time, the network learns the hysteresis response characteristics of the main beam arching and falling back, thereby fitting and outputting a benchmark deformation trajectory that matches the current environmental state. The same segment will gradually arch upwards in the morning for several consecutive days and fall back in the evening. The same cable area will produce regular lateral swaying when the wind direction changes; by fitting the historical total deformation data, the system can identify deformation components that have periodic occurrence and automatic regression to the baseline, thus providing evidence in the time dimension for transient response extraction. When the temperature gradient distribution corroborates the historical fitting results, the system can extract the elastic transient deformation caused by the environmental conditions. If a certain displacement change occurs simultaneously with the increase in cross-sectional temperature difference, and also exhibits similar rebound characteristics in similar environments in the past, then this part is preferentially identified as transient deformation. If a certain displacement occurs during a high-temperature period, but is not accompanied by a reasonable temperature gradient explanation, and has no natural regression trend in history, then it is not easily included in the transient part, but is retained for subsequent permanent deviation identification. As an anomaly handling mechanism, if the environmental condition data is incomplete, such as the temperature measurement point failing locally and the cross-sectional temperature difference cannot be reliably reconstructed, the system can temporarily increase the confidence assessment weight of the historical fitting results, but at the same time reduce the safety level of the automatic judgment; if the bridge is in a clearly non-steady working condition, such as large segment hoisting, concentrated cable force adjustment, temporary support removal, etc., the historical periodic pattern will be interrupted by construction disturbance. At this time, the system will give priority to the current process state constraints and will not rashly attribute the significant deformation of this period to the environmental transient response; Under continuous sunny construction conditions, the system found that the beam segment near the mid-span closure point showed slight upward arching from 13:00 to 16:00 every day, and the temperature of the top plate of the steel box girder was always higher than that of the bottom plate, and the temperature of the web plate on the sun-facing side was higher than that of the web plate on the shaded side. Based on historical data from multiple days, the beam segment returned to a position close to the same baseline after the temperature became uniform at night. Therefore, the system identified the upward arching and some lateral displacement during this period as environmentally driven elastic transient deformation, and did not directly include it in the correction amount. The purpose of this step is to separate the deformations that bridges will naturally undergo in the environment and are capable of rebounding from the total deviation, thereby enabling the identification of transient disturbances such as thermal and wind-induced disturbances and avoiding unnecessary forced cable adjustments and jacking during construction due to misjudgment.

[0020] Furthermore, the calculation of plastic permanent deviation includes: using the total structural deformation represented by spatial coordinate data as the benchmark input; using the total structural deformation to eliminate elastic transient deformation, separating the plastic permanent deviation caused by the construction process state data, so as to use the plastic permanent deviation for finite element analysis calculation.

[0021] This embodiment provides a mechanism for separating permanent plastic deviations. Specifically, based on the identification of the aforementioned elastic transient deformation, this embodiment further extracts the portion of the total deformation that cannot be restored by the environment and disappear on its own, which is used to characterize the actual posture deviation caused by the construction process. This portion is the direct basis for subsequent stress analysis and correction decisions. In detail, the total structural deformation can be understood as the amount of geometric displacement deviation exhibited by the bridge at a certain moment, which is mixed with environmental transient response and construction residual error. After removing the transient part, the remaining deformation is closer to the actual assembly state of the bridge at the current construction stage. The plastic permanent deviation here is not limited to macroscopic plastic yield in the sense of materials science, but also includes residual attitude deviations that are difficult to recover automatically due to segment positioning errors, cable force application sequence, concrete age differences, self-weight redistribution and temporary constraint release. Construction process status data plays an important role in this step; because even though the deviation is retained, the causes may be completely different; for example, after the segment is hoisted, different unloading methods of the lifting equipment will lead to different placement deviations at the beam end; when the stay cables are tensioned in stages, if a group of cables is tensioned prematurely, local segments may form an asymmetrical upward arch. After the steel-concrete composite section is poured, there is a difference between the shrinkage and creep of the concrete and the elastic rebound of the steel components, which will also accumulate long-term residual deformation near the interface. The system reads the construction process records, associates these deviations with the corresponding stages, and thus determines which residual deviations are causally related to the construction process and should be used as the objects of subsequent correction. In terms of microscopic data flow, the total deformation of a certain control section can be understood as the superposition of the environmental rebound portion and the construction residue portion. If the total deformation of a certain section is the first deformation value, and the thermo-elastic transient deformation determined by the system environment analysis is the second deformation value, then the difference between the first deformation value and the second deformation value is determined to be the plastic permanent deviation associated with construction. If the total deformation still retains the difference at night, it further confirms that it does not belong to a reversible environmental response. This shows the separation logic in the data structure, rather than the requirement to use fixed values ​​on site. As an anomaly handling mechanism, if the difference between the total deformation and the elastic transient deformation is close to the sensing noise range, the system does not directly identify a permanent deviation, but marks the result as a low-confidence residual and continues to observe it for several subsequent cycles; if a deviation suddenly reverses before and after different process switching points, it indicates that it may still be affected by temporary supports, unloading paths, or measurement benchmark drift, resulting in geometric consistency anomalies, and the system should suspend inputting it as a stable permanent deviation into the calculation model; if the construction log is missing or the recording sequence is disordered, this step only outputs the geometric residual and does not output a clear process attribution conclusion; After a segment of the main beam was hoisted, the left end of the beam at the closure joint continued to deflect downwards at night, while the additional upward arching that occurred during the day due to thermal expansion had largely disappeared. The system, combining the hoisting and unloading time, the tensioning records of adjacent cable groups, and the time of temporary support removal, determined that the remaining downward deflection was not caused by the environment, but by the residual construction deviation formed by the segment positioning and cable force redistribution. This deviation was used as the actual posture error of the structure and input into the subsequent stress analysis. The purpose of this step is to extract the part that truly requires engineering intervention from the mixed deviations, thereby accurately identifying the target of correction and focusing subsequent analysis and control on the non-negligible residual state caused by construction.

[0022] Further, step S3 specifically includes: constructing a structural mechanical reference state based on the finite element analysis model; inputting the plastic permanent deviation as a boundary displacement condition into the finite element analysis model for real-time calculation; extracting the stress tensor output by the finite element analysis model, generating stress redistribution state data, and rendering the stress redistribution state data as stress cloud map data for display by the front-end monitoring equipment.

[0023] This embodiment provides a mechanism for mapping geometric deviations to internal stress states. Specifically, after plastic permanent deviations have been identified in the bridge, this embodiment no longer stops at the level of identifying apparent geometric deviations, but uses a finite element analysis model to transform the deviations into stress redistribution results within the structure, extending construction control from visible linear problems to internal force problems that cannot be directly observed. In detail, the structural mechanics reference state refers to the theoretical stress reference state corresponding to the current construction stage. This state comprehensively considers the installed segments, tensioned cable groups, temporary support conditions, material partitions, and boundary constraints, and is used to characterize the theoretical stress equilibrium state of the structure under ideal construction conditions and without additional attitude residuals. After establishing this reference, subsequent input plastic permanent deviations can be regarded as disturbances to the ideal state, thereby analyzing the additional stress generated by the actual bridge body due to attitude deviations. Using permanent plastic deviations as boundary displacement conditions input into the model means that the system does not set virtual boundary forces under unconstrained conditions, but rather allows the model to respond according to the geometric facts that have already occurred on site. Its engineering significance is that if a bridge segment is actually assembled in a state of being too high, too low, or torsion, the internal structure will automatically transform this geometric incompatibility into stress redistribution through members, cables, and interface connections. Especially in the steel-concrete composite zone, steel and concrete components have different stiffness and thermal expansion characteristics. Once the geometric deviation is forcibly retained, stress often concentrates near the interface. The model calculation is precisely to make this potential risk explicit in advance. After extracting the stress tensor and generating stress redistribution state data, it can be further rendered into a stress cloud map for display by the front-end monitoring equipment. The high-stress areas in the cloud map often correspond to the locations where the structure is least willing to undergo additional deformation, such as the cable anchor area, near the diaphragm, around the steel-concrete connection components, and the edge of the closure joint. Through color distribution and area range, on-site personnel can intuitively see where a certain correction scheme may cause new stress concentration, thereby avoiding simply pursuing geometric closure while ignoring the safety of local structures. As an anomaly handling mechanism, if the current construction stage is inconsistent with the preset model stage, for example, if the temporary supports on site have not yet been removed but the model is calculated based on the state after removal, the system should prohibit the output of the formal stress cloud map and only indicate that the model working conditions are mismatched; if certain key material parameters are updated due to changes in construction batches, such as the elastic modulus of concrete increasing significantly with age, the model's baseline state needs to be corrected synchronously; before the parameters are updated, only trend risk judgments can be output, and precise correction suggestions should not be given; if the permanent plastic deviation exceeds the applicable range of the model, it indicates that an anomaly exceeding the preset tolerance threshold has occurred on site, and the system should switch to manual review mode; After detecting a continuous downward deviation at the left beam end and a slight torsion at the right beam end near the mid-span closure section, the system inputs the geometric residual into the finite element model corresponding to the construction stage. The calculation results show that if the current posture is maintained and the next closure operation is carried out, there will be significant additional tensile stress concentration in the web connection area of ​​the steel-concrete composite section and near the adjacent cable anchor plates. The stress cloud map formed on the front-end monitoring screen thus visually presents the risks that were originally hidden inside the beam, allowing the chief engineer and monitoring personnel to make simultaneous judgments. The purpose of this step is to transform the explicit displacement deviation into implicit and more critical structural internal force changes, thereby enabling the early identification of lock-in stress, local stress concentration, and subsequent cracking risks.

[0024] Further, step S4 specifically includes: constructing a multi-objective optimization algorithm, setting the objective function of the multi-objective optimization algorithm as minimizing the overall stress variance of the structure and minimizing the stress extrema of the key stress nodes of the bridge; inputting the stress redistribution state data into the multi-objective optimization algorithm based on genetic algorithm or particle swarm optimization algorithm for iterative optimization; and outputting a correction strategy instruction that satisfies the objective function, wherein the correction strategy instruction includes the cable tension adjustment value and the jack pushing displacement value.

[0025] This embodiment provides a deviation correction decision mechanism oriented towards stress equilibrium. Specifically, after obtaining the stress redistribution state data, this embodiment does not adopt the single-objective correction approach of using the node with the largest geometric deviation as the single control target for reverse displacement compensation. Instead, it simultaneously focuses on the stress dispersion degree of the entire bridge and the peak stress of key nodes to generate deviation correction strategy instructions that take into account both safety and construction feasibility. In detail, focusing solely on minimizing geometric deviations for correction has significant drawbacks in cable-stayed bridge construction. The rapid return of a control point's coordinates to its design position does not necessarily indicate a more rational stress distribution across the entire bridge. Conversely, excessive adjustments to a particular set of cables in pursuit of local geometric alignment may lead to higher stress peaks in the cable anchorage zone, crossbeam connection zone, or steel-concrete interface. Therefore, this embodiment considers both the overall stress variance and the extreme stress values ​​at key nodes as decision-making objectives. The overall stress variance reflects the uniformity of stress distribution across the entire bridge, while the extreme stress values ​​at key nodes reflect the safety of locally vulnerable areas. Only when both are considered simultaneously can the correction result more closely approximate the optimal state required for the completed bridge. During strategy generation, the system limits executable means to controllable quantities that are actually operable on-site, such as cable tension adjustment values ​​and jack displacement values. The resulting solution is not an abstract optimization conclusion, but a construction plan that can be directly converted into work instructions. To clearly demonstrate the integration of algorithm selection and business logic, a structured decomposition of multi-objective iterative optimization is performed using a genetic algorithm as an example. The coding structure of individuals in the algorithm population is defined, with each individual represented by a real-number multidimensional vector. The elements of the vector correspond one-to-one with the tension adjustment values ​​of each group of cables and the jacking displacement values ​​of each group of jacks on site. The multi-objective function is transformed into a fitness evaluation mechanism. For any individual representing a hypothetical candidate correction scheme, it is input into the finite element analysis model and combined with stress redistribution state data for evolution calculation. The overall structural stress variance at each measuring point of the entire bridge and the stress extrema of the key stress nodes of the bridge are obtained. A weighted summation method is used to assign preset weight coefficients to the overall structural stress variance and the stress extrema of the key stress nodes, respectively, to construct a comprehensive fitness evaluation function. If the stress extrema exceed the design safety limit, a large fitness penalty is imposed on the individual. By selecting, crossing, and mutating, a new generation of strategy combinations is generated within the allowable physical adjustment range of the equipment. Through multiple generations of iteration, non-convergent strategy schemes that cause local stress concentration or the overall stress variance to exceed the control threshold are continuously eliminated. Through the above mechanism, for different bridge types, the system can simultaneously evaluate multiple adjustment combinations. For example, it can prioritize fine-tuning adjacent cable groups, synchronously release local temporary constraints, or use small-amplitude multiple jacking instead of a single large displacement jacking to reduce the impact on local structures. From the perspective of data flow, the system output can be understood as the selection of several candidate solutions; for example, solution A can rapidly reduce the height difference at the closure point, but the stress extreme value in the cable anchor area increases; solution B is slightly slower in improving the geometry, but can significantly alleviate the additional tensile stress in the steel-concrete joint section; solution C only achieves alignment through jacking, but will cause uneven stress in the tower-beam connection area; the system retains the solution that can both improve the attitude and reduce the uneven stress of the entire bridge through multi-objective optimization, and outputs the corresponding cable adjustment amount and jacking amount as the final command; As an anomaly handling mechanism, if several candidate schemes are similar in stress effect, but one of them exceeds the jack stroke, the allowable adjustment range of the cable, or the implementation capability of the on-site team, the system will automatically eliminate the scheme; if all candidate schemes will cause the stress at a certain key node to exceed the design control value, the system will not output a correction command, but will instead issue a prompt that the construction steps need to be adjusted or the work postponed; if key input data is updated during the optimization process, such as when another set of cables has just been tensioned on site, the original scheme will automatically become invalid and be recalculated. During the attitude check before the closure, the system found that although simply increasing the tension of one side span cable could quickly eliminate the height difference at the beam end, it would cause the additional tensile stress on one side of the steel-concrete composite section to continue to increase. If the adjustment was changed to a small coordinated adjustment of the two adjacent cable groups, and limited jacking was carried out with the jacks at the closure point, the stress distribution of the entire bridge would be more uniform and the local stress extreme value would be lower. Therefore, the system output a correction strategy of fine-tuning the cable force in groups and jacking with small displacements to the site, rather than a simple large-scale single cable adjustment scheme. The purpose of this step is to transform the correction decision from a single linear correction to a comprehensive control that takes into account both the overall and local stresses, thereby achieving a construction posture adjustment scheme that is more in line with the structural safety mechanism.

[0026] Further, step S4 generates the optimal time window, including: acquiring meteorological forecast data for a future preset time period and predicting future environmental state data; calculating the elastic transient deformation value caused by the future environmental state data; when the elastic transient deformation value is lower than a preset deformation threshold determined based on the bridge design tolerance error, determining the corresponding time period as the optimal time window; when the elastic transient deformation value is equal to the preset deformation threshold, determining the corresponding time period as a candidate time window; when the elastic transient deformation value is higher than the preset deformation threshold, determining the corresponding time period as an unworkable time window.

[0027] This embodiment provides a mechanism for selecting the time window for correction and closure operations. Specifically, after the correction scheme has been generated, this embodiment further addresses the question of when to execute it most appropriately. Since the elastic transient deformation amplitude of the main girder of a cable-stayed bridge varies significantly under different weather conditions, if the same set of cable adjustment and jacking operations are carried out under high temperature and strong sunlight, the elastic transient deformation may be mistakenly identified as a permanent deviation and locked into the structure. If the operations are carried out under uniform temperature and stable wind conditions, it is easier to approach the designed state without additional stress. In detail, the system acquires meteorological forecast data for a preset time period in the future, focusing on temperature changes, solar radiation intensity, cloud cover, wind speed and direction, and diurnal temperature range trends, and predicts future environmental conditions accordingly. Combining the existing thermal response characteristics of the bridge, it assesses the level of elastic transient deformation that may be caused in each time period. The preset deformation threshold here is not arbitrarily set, but is based on the bridge design allowable error and construction stage control standards, and is used to determine whether environmental disturbances in a certain time period are sufficient to affect the authenticity of the correction action. When the predicted elastic transient deformation value is lower than the preset deformation threshold, it indicates that the environmental disturbance is relatively weak during this period, and the bridge geometry is closer to its true residual attitude. At this time, it is less likely to introduce additional errors when performing correction. Therefore, the corresponding time period is determined to be the optimal time window. When the predicted elastic transient deformation value is equal to the preset deformation threshold, it indicates that the environmental disturbance is in a critical state. Work can still be carried out, but more stringent monitoring is required. Therefore, the corresponding time period is determined to be the alternative time window. When the predicted elastic transient deformation value is higher than the preset deformation threshold, it indicates that the environment is significantly amplifying the apparent deviation. At this time, correction is prone to misjudgment and over-adjustment. Therefore, the corresponding time period is determined to be the non-operational time window. This timing window selection logic is particularly applicable to the sensitive stages before and after the closure; because once the closure operation is completed, some internal forces will solidify with the transformation of the structural system. If it is carried out in an unfavorable thermal environment, even if the geometric alignment seems ideal at the time, new internal force imbalances and linear changes may be exposed after the environment recovers. Therefore, this embodiment regards the timing of the operation as a control variable that is as important as the cable adjustment scheme. As an anomaly handling mechanism, if the confidence level of future weather forecasts is low, or if there are drastic short-term weather changes, such as sudden rainfall, strong gust warnings, or rapid changes in cloud cover, the system can downgrade the original optimal time window to an alternative time window and require on-site measurement confirmation again before operation. If multiple time periods do not meet the conditions for operation, the system will not force the recommendation of the nearest time, but will prompt for postponement of operation. If the predicted elastic transient deformation value hovers near the preset deformation threshold for a long period of time, the system allows on-site manual verification, temporary sunshade, or local wind protection measures before deciding whether to activate the alternative time window. The day before the mid-span closure operation, the system, based on the next day's weather forecast, found that the bridge deck radiation would be strong in the afternoon and the cross-bridge wind speed would be high. However, from 11 PM that day to 3 AM the next day, the temperature change would slow down, cloud cover would increase, and the wind speed would decrease. After assessment, the elastic transient deformation value that might be caused during the nighttime period was significantly lower than the preset deformation threshold, and therefore it was determined to be the optimal time window. The hour before and after sunrise in the early morning was within the critical range and was listed as an alternative. The afternoon period was determined to be unsuitable for operation. Ultimately, the cable adjustment and jacking operations were scheduled to be completed at night. The purpose of this step is to incorporate the timing of construction into the attitude control closed loop, thereby achieving the arrangement of correction operations under the condition of minimizing environmental disturbances, and reducing the risk of miscorrection and residual stress lock-in caused by improper timing selection.

[0028] Furthermore, the method also includes closed-loop feedback control based on the correction strategy command, wherein the closed-loop feedback control includes: executing physical correction actions within the optimal time window according to the correction strategy command; collecting updated spatial coordinate data and updated interface strain data in real time after executing the physical correction actions; and using the updated spatial coordinate data and updated interface strain data, combined with the current environmental status data and construction procedure status data, re-executing steps S1 to S4 to perform a new round of dynamic monitoring and attitude correction.

[0029] This embodiment provides a mechanism for closed-loop feedback control of construction correction. Specifically, after the above scheme has given the correction instructions and the optimal time window, this embodiment further solves the problem of whether the expected result is achieved after execution and how to continue to correct it if it is not achieved. Since the bridge is a large-scale flexible structure, there is a coupling relationship between cable adjustment, jacking, and unloading. The on-site execution results often differ from the static prediction. Therefore, it is necessary to remeasure, reanalyze, and re-decide after the action is implemented to form a closed-loop control. In detail, when performing physical correction actions within the optimal time window, the operation can be carried out according to the cable adjustment sequence and jack displacement control rhythm output by the system. During the execution, spatial coordinate data is used to observe the changes in the main beam alignment, the relative position of the closure joint, and the attitude changes of key points of the tower and beam. Combined with surface strain data, it is used to monitor whether abnormal tension and compression concentrations occur in the steel-concrete interface and local stress-sensitive areas due to the correction actions. After both are updated synchronously, the system re-enters the data integration, transient stripping, stress analysis, and strategy generation process to evaluate the effect of this round of correction. The significance of closed-loop feedback lies in the fact that bridge correction is not a one-time action, but a process of gradually approaching the target state; a round of cable adjustment may improve the height difference at the closure point, but it may cause a slight increase in local interface strain; another round of small-scale jacking may further eliminate residual torsion, but at the same time trigger the redistribution of stress on adjacent cable groups; through round-by-round updates, the system can continuously compare whether the current round is closer to the target of stress equilibrium and reasonable attitude than the previous round, thereby avoiding the irreversible risks caused by a one-time large-scale action; In terms of micro-data flow, it can be understood that after each round of operation, the system treats the new measurement results as the new reality of the bridge, and then judges which of them are still environmentally reversible items and which have become new residual deviations, and adjusts the control quantities for the next round accordingly. If the height difference at the closure section decreases and is lower than the first preset threshold after the first round, but the strain of the steel-concrete joint section continues to be greater than the second preset strain threshold, then the next round will prioritize fine-tuning the cable force with a smaller adjustment step size rather than continuing to increase the jacking. If the stress distribution has become significantly balanced after the first round and the remaining deviation is close to the noise or allowable error, then the system can actively terminate the subsequent correction to avoid excessive intervention. As an anomaly handling mechanism, if there is a contradiction between the updated coordinate data and the updated strain data after the correction is performed, such as geometrically approaching alignment but with abnormally amplified local strain, the system should prioritize suspending subsequent operations according to structural safety principles and prompt manual inspection of anchorages, temporary connections, and sensor status. If a sudden weather change occurs within the optimal time window, causing a rapid increase in environmental disturbance, the current closed-loop cycle can be suspended and resumed after the environment recovers. If the attitude and stress control conditions cannot be simultaneously met after multiple iterations, it indicates that the existing cable adjustment and jacking methods are insufficient to solve the problem. In this case, the system should suggest adjusting the construction sequence, verifying model parameters, or adding temporary auxiliary measures. During the nighttime recommended window, the two sets of stay cables were first slightly tensioned according to the system's suggestions, and a limited jacking was completed in conjunction with the closure jacks. After the action was completed, the system immediately read the new beam end coordinates and the strain of the steel-concrete joint section, and found that the geometric deviation had been significantly reduced, but the tensile strain at one interface was still higher than that of the adjacent area. Therefore, the system re-analyzed and determined that it was not advisable to continue unidirectional jacking at this point, but rather to use the adjacent cable groups for coordinated fine-tuning in the next round. After two rounds of closed-loop correction, the closure attitude met the control requirements, and the stress distribution of the key stress nodes was restored to equilibrium. The system then ended the active correction of this stage. The purpose of this step is to integrate analysis, decision-making, execution, and re-analysis into a continuous control chain, thereby achieving dynamic correction of the bridge's attitude and internal force state throughout the construction process and improving the consistency between the correction results and the actual structural response.

[0030] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for intelligent monitoring of deformation and attitude correction throughout the construction process of a cable-stayed bridge, characterized in that, include: Step S1 involves dynamically collecting and integrating spatial coordinate data, surface strain data, environmental condition data, and construction procedure status data of the cable-stayed bridge. Step S2: Based on the spatial coordinate data, the total structural deformation is characterized, and the total structural deformation is decoupled based on the integrated data. The decoupling includes calculating the elastic transient deformation and calculating the plastic permanent deviation. Step S3: Based on the pre-constructed finite element analysis model, perform finite element analysis calculations using the plastic permanent deviation to obtain stress redistribution state data inside the structure; Step S4: Perform multi-objective optimization based on the stress redistribution state data, generate correction strategy instructions, and generate the optimal time window by combining environmental prediction data. Step S5: According to the correction strategy instruction, the front-end construction control equipment is invoked to perform physical correction actions within the optimal time window.

2. The intelligent monitoring and attitude correction method for deformation of the entire construction process of a cable-stayed bridge as described in claim 1, characterized in that, Before step S2, the following is also included: The spatial coordinate data, the bonding surface strain data, the environmental state data, and the construction procedure state data are preprocessed to obtain preprocessed data. The preprocessing includes missing value imputation, outlier detection, and data standardization. The data standardization involves using extreme value normalization or standard deviation standardization to uniformly render the data dimensionless, thereby eliminating dimensional differences.

3. The intelligent monitoring and attitude correction method for deformation of the entire construction process of a cable-stayed bridge as described in claim 1, characterized in that, Calculation of elastic transient deformation includes: The structural temperature gradient distribution caused by the environmental state data was calculated using a thermodynamic heat transfer model based on Fourier's law of thermal conductivity. By using a time series analysis network that includes a long short-term memory network or a recurrent neural network, the structural temperature gradient distribution sequence, wind speed and direction environmental state sequence, and pre-established and stored historical structural total deformation data within a preset historical time window are used as multi-dimensional feature joint inputs, and the expected value of elastic deformation under pure environmental driving is output as the fitting analysis result. Based on the temperature gradient distribution of the structure and the fitting analysis results, the elastic transient deformation caused by the environmental state data is extracted.

4. The intelligent monitoring and attitude correction method for deformation of the entire construction process of a cable-stayed bridge as described in claim 3, is characterized in that, Calculation of permanent plastic deviation includes: The total structural deformation represented by the spatial coordinate data is used as the reference input; The elastic transient deformation is eliminated by using the total structural deformation, and the plastic permanent deviation caused by the construction process state data is separated, so that the plastic permanent deviation can be used for finite element analysis calculation.

5. The intelligent monitoring and attitude correction method for deformation of the entire construction process of a cable-stayed bridge as described in claim 1, characterized in that, Step S3 specifically includes: The structural mechanical reference state is constructed based on the finite element analysis model. The plastic permanent deviation is input as a boundary displacement condition into the finite element analysis model for real-time calculation; The stress tensor output by the finite element analysis model is extracted to generate the stress redistribution state data, and the stress redistribution state data is rendered as stress cloud map data for display by the front-end monitoring equipment.

6. The intelligent monitoring and attitude correction method for deformation of the entire construction process of a cable-stayed bridge as described in claim 1, characterized in that, Step S4 specifically includes: A multi-objective optimization algorithm is constructed, and the objective function of the multi-objective optimization algorithm is set as minimizing the overall stress variance of the structure and minimizing the stress extrema of the key stress nodes of the bridge. The stress redistribution state data is input into the multi-objective optimization algorithm based on genetic algorithm or particle swarm optimization algorithm for iterative optimization; Output the correction strategy instruction that satisfies the objective function, wherein the correction strategy instruction includes the cable tension adjustment value and the jack pushing displacement value.

7. The intelligent monitoring and attitude correction method for deformation of the entire construction process of a cable-stayed bridge as described in claim 6, characterized in that, The step S4 of generating the optimal time window includes: Acquire meteorological forecast data for a preset time period in the future to predict future environmental conditions; Calculate the elastic transient deformation value caused by the future environmental state data; When the value of the elastic transient deformation is lower than the preset deformation threshold determined based on the allowable error of bridge design, the corresponding time period is determined to be the optimal time window; When the value of the elastic transient deformation is equal to the preset deformation threshold, the corresponding time period is determined to be the candidate time window. When the value of the elastic transient deformation is higher than the preset deformation threshold, the corresponding time period is determined to be an unworkable time window.

8. The intelligent monitoring and attitude correction method for deformation of the entire construction process of a cable-stayed bridge as described in claim 7, characterized in that, The method further includes closed-loop feedback control based on the correction strategy command, wherein the closed-loop feedback control includes: According to the correction strategy instructions, physical correction actions are performed within the optimal time window; Real-time acquisition of updated spatial coordinate data and updated interface strain data after the physical correction action is performed; Using the updated spatial coordinate data and the updated interface strain data, combined with the current environmental status data and construction procedure status data, steps S1 to S4 are executed again to perform a new round of dynamic monitoring and attitude correction.

Images