CN-122018395-A - Digital twinning-based short-line cantilever construction bridge linear prediction and regulation method

CN122018395ACN 122018395 ACN122018395 ACN 122018395ACN-122018395-A

Abstract

The invention provides a digital twinning-based stub method cantilever construction bridge linear prediction and regulation method, which comprises the steps of obtaining geometric data, physical attribute data, construction behavior data and engineering rule data of a stub method precast beam section, establishing an initial digital twinning model, calibrating initial digital twinning model parameters, constructing and operating a multi-scale mixed prediction model, outputting a predicted value and an uncertainty interval of a whole construction process linear based on the digital twinning model, judging whether the predicted value has overrun deviation, generating and optimizing a regulation scheme when the overrun deviation exists, converting an executable instruction into an executable instruction and transmitting the executable instruction to field construction equipment, collecting result data after instruction execution, and refluxing the result data to update the digital twinning model and the mixed prediction model to form closed loop control. The invention solves the hysteresis problem of the traditional method for correcting deviation after the accident, realizes the real-time prediction, intelligent diagnosis and active regulation of the construction line shape, and forms a complete prediction-regulation-optimization closed loop.

Inventors

GAO CHEN
CHEN XIAOTING
SU YINGCHAO
HUANG JUN
ZHANG LELE
ZHANG MINGBIN
ZHANG BAO
CHEN HAIFA
GENG WEIMING
ZHOU XIAOYI
LI JIA

Assignees

中建七局第六建筑有限公司
东南大学

Dates

Publication Date: 20260512
Application Date: 20260127

Claims (10)

1. A method for predicting and regulating the linear shape of a cantilever construction bridge based on digital twinning includes such steps as S1, obtaining the geometric data, physical attribute data, construction behavior data and engineering rule data of the segment of the cantilever, S2, creating an initial digital twinning model integrating geometric model, structural mechanical model and construction procedure rule, S3, setting up multi-source sensor network to collect real-time data in construction period, dynamically calibrating the parameters of said initial digital twinning model by data assimilation algorithm to realize synchronous evolution of model and solid bridge, S4, creating and running multi-scale mixed prediction model, outputting the predicted value and uncertainty interval of the whole construction process based on said digital twinning model, S5, judging if there is an overrun deviation, S6, intelligent diagnosing to identify the cause of deviation, automatically generating and optimizing the scheme, converting the optimized scheme to executable instruction to field construction equipment, S7, and updating the data after the instruction is executed to form a closed-loop control model.
2. The method for predicting and regulating the linear shape of the cantilever construction bridge based on the digital twin stub method according to claim 1 is characterized in that in the step S1, the obtaining of the geometric data comprises the steps of carrying out three-dimensional laser scanning on each prefabricated segment to obtain point cloud data, carrying out denoising, simplification and feature extraction on the point cloud data, calculating geometric parameters of the end faces of the segments and the assembly reference plane, and matching with a design model to generate a three-dimensional error field for guiding assembly.
3. The method for predicting and regulating the linear shape of a cantilever construction bridge based on a digital twin-line method according to claim 1 or 2, wherein in the step S1, the construction behavior data comprises a sequence of working procedures driven by an event, a tensioning sequence, a loss value, a hanging basket pushing and locking state and a maintenance system, and the engineering rule data comprises a target linear shape of a control point, a pre-camber value, a deviation threshold value and a risk level set according to specifications.
4. The method according to claim 1 or 2, wherein in step S2, the initial digital twin model couples the segment geometry model, the structural mechanics finite element model considering the time-varying effect, and the executable logic encoded with the construction behavior data and the engineering rule data under a unified space-time reference.
5. The method for predicting and regulating the linear shape of the cantilever construction bridge based on the digital twin stub method according to claim 1 or 2 is characterized in that in the step S3, the data collected by the multi-source sensor network comprises geometric displacement data obtained by a total station, a displacement meter and an inclinometer, mechanical data obtained by a strain gauge, a cable force meter and a jack pressure sensor, environmental data obtained by a temperature and humidity sensor and an anemometer, and construction management data reflecting the state of tensioning equipment and the age of concrete.
6. The method for predicting and regulating the linear shape of the cantilever construction bridge based on the digital twin-line method according to claim 1 or 2 is characterized in that the step S3 further comprises the steps of cleaning, time-space alignment, anomaly identification and reliability assessment pretreatment on collected real-time data, and the data assimilation algorithm adopts an extended Kalman filtering, unscented Kalman filtering, particle filtering or Bayesian updating method to calibrate material parameters, boundary rigidity and prestress loss on line.
7. The method for predicting and regulating the linear shape of a cantilever construction bridge based on a digital twin stub method according to claim 1 or 2, wherein in the step S4, the multi-scale mixed prediction model comprises: The embedded entity model is used for carrying out rapid incremental analysis by using a process event driver; The high-fidelity physical simulation sub-model is used for carrying out fine time course calculation in construction stages; the data driving sub-model is used for learning the historical prediction residual error of the physical simulation sub-model and correcting the historical prediction residual error; and the mixed prediction model fuses the output of the three models in a weighting or Bayesian mode to obtain a final linear prediction value with a confidence interval.
8. The method for predicting and regulating the linear shape of the cantilever construction bridge based on the digital twin short-line method according to claim 7, wherein the data driving sub-model adopts a long-short-period memory network, a gradient lifting tree or a Gaussian process regression algorithm, and the input characteristics comprise concrete age, structural temperature gradient, prestress tensioning actual measurement value, historical linear deviation and construction equipment state data.
9. The method for predicting and regulating the linear shape of the cantilever construction bridge based on the digital twin stub method according to claim 1,2 or 8 is characterized in that in the step S6, the intelligent diagnosis identifies a deviation main factor based on sensitivity analysis or variable importance evaluation, the automatically generated regulating scheme comprises one or more of adjusting the subsequent prestress tensioning force or sequence, adjusting the elevation of a temporary support, controlling the thickness of a cementing layer among sections, implementing local temperature control and correcting the forward parameters of the hanging basket, and the generated scheme is preferentially converted into an operation instruction capable of directly controlling the tensioning system, the jack or the hanging basket after being evaluated by a multi-objective optimizing algorithm.
10. The method for predicting and controlling the linear shape of a cantilever construction bridge based on a digital twin method according to claim 1,2 or 8, wherein in the step S7, the reflowed data is used for updating the geometric calibration parameters and the mechanical parameters of the digital twin model, further training the data driving sub-model, and precipitating the diagnosis, decision and execution of the current control as a reusable knowledge base.

Description

Digital twinning-based short-line cantilever construction bridge linear prediction and regulation method Technical Field The invention relates to the technical field of building construction, in particular to a digital twinning-based method for predicting and regulating the linear shape of a cantilever construction bridge by a stub method. Background The short-line method prefabricated cantilever assembling is a common construction method for superstructure such as long-span continuous beams and cable-stayed bridges. In the construction method, in the multi-stage of segment prefabrication, transportation, temporary assembly, hanging basket propelling, prestress tensioning and closure, the actual geometry, materials and construction conditions are different from the design assumption, and the factors such as temperature, creep shrinkage, tensioning loss, hanging basket deflection, temporary support settlement, glue layer thickness fluctuation, construction time sequence and equipment state can be coupled to influence the structural linearity. The existing monitoring is based on offline theoretical calculation before construction and post correction of field scattered measurement, and has the problems of model and real bridge disjoint, prediction lag, unexplained deviation cause, dependence on experience in regulation and control and insufficient closed loop, and is easy to cause linear overrun and reworking, quality risk and potential safety hazard. The invention discloses a method for controlling the assembly line shape of a high-speed rail continuous beam stub-line matching precast cantilever, which relates to the technical field of high-speed rail bridge construction control and comprises the steps of marking a pair of measuring points on the central line of a matching section, marking a pair of measuring points at equal distances on two sides of the pair of measuring points, marking three pairs of measuring points at the same position of a section to be installed, assembling and adjusting the beam section, installing the section to be installed, then measuring the deviation values of the measuring points of the six pairs of measuring points after adjusting the elevation and plane position of the six pairs of measuring points, then reinforcing, assembling and correcting the beam section, and calculating the error of each pair of measuring points of the section to be installed and correcting. However, the method cannot predict the line shape before or during construction, is difficult to identify and prevent deviation in advance, mainly depends on measurement and manual correction after construction, has the problems of control lag, experience dependence, lack of predictability and closed loop optimization capability, and is difficult to realize high-precision and active control of the line shape. Disclosure of Invention Aiming at the technical problems, the invention provides a digital twinning-based method for predicting and regulating the linearity of a cantilever construction bridge by a stub method, which is used for solving the problem that the linear prediction cannot be carried out before or during construction in the cantilever assembly linear control method in the prior art. In order to achieve the above purpose, the technical scheme of the invention is realized as follows: A method for predicting and regulating the linear shape of a cantilever construction bridge based on digital twinning includes such steps as S1, obtaining the geometric data, physical attribute data, construction behavior data and engineering rule data of the segment of prefabricated beam, S2, creating an initial digital twinning model integrating geometric model, structural mechanical model and construction procedure rule, S3, collecting real-time data by multi-source sensor network in construction period, dynamically calibrating the parameters of said initial digital twinning model by data assimilation algorithm, S4, creating and running multi-scale mixed prediction model, outputting the predicted value and uncertainty interval of the whole construction process based on said digital twinning model, S5, judging if there is an overrun, if so, intelligent diagnosis is carried out to identify the cause of the overrun, automatic generation and regulation scheme, converting the optimized scheme into executable instruction to field construction equipment, S7, updating the reflux data after instruction execution to form a closed loop with the digital twinning model. Further, in the step S1, the step of obtaining the geometric data comprises the steps of carrying out three-dimensional laser scanning on each prefabricated segment to obtain point cloud data, carrying out denoising, simplification and feature extraction on the point cloud data, calculating geometric parameters of end faces of the segments and assembly reference planes, and matching with a design model to generate a three-dimensional error field for guiding assembly. Fur