CN-121977770-A - Digital twin monitoring method and system for fatigue of steel bridge deck

CN121977770ACN 121977770 ACN121977770 ACN 121977770ACN-121977770-A

Abstract

The invention discloses a steel bridge deck fatigue digital twin monitoring method and system, which comprise the steps of responding to received traffic operation video information on a steel bridge deck acquired by camera equipment, identifying vehicle parameters according to the traffic operation video information, generating vehicle load data based on the vehicle parameters, inputting the vehicle load data into a preset steel bridge deck structural model, calculating stress response data of a preset bridge deck fatigue sensitive part, calculating fatigue parameters based on the stress response data by adopting an equivalent structural stress method, and visually displaying the fatigue parameters and/or fatigue early warning information on a three-dimensional digital twin platform of the steel bridge. Therefore, the full-field, real-time and non-contact monitoring of the fatigue state of the key part of the bridge deck is realized by constructing a full-flow closed loop of data acquisition, load calculation, dynamic simulation, equivalent structure stress method damage calculation, visualization and early warning.

Inventors

YE XIAOWEI
WANG MINGYANG
SU YOUHUA
DING YANG
JIN TAO

Assignees

浙江大学

Dates

Publication Date: 20260505
Application Date: 20260403

Claims (10)

1. A digital twin monitoring method for steel deck fatigue, comprising: In response to receiving traffic operation video information on a steel bridge deck acquired by the camera equipment, identifying vehicle parameters according to the traffic operation video information, and generating vehicle load data based on the vehicle parameters; inputting the vehicle load data into a preset steel bridge deck structural model and calculating stress response data of a fatigue sensitive part of a preset bridge deck, wherein the preset steel bridge deck structural model is an orthotropic steel bridge deck three-dimensional entity model established based on a finite element method; Based on the stress response data, adopting an equivalent structural stress method to integrate stress components in two mutually orthogonal directions of the fatigue sensitive part of the preset bridge deck, and calculating fatigue parameters, wherein the fatigue parameters comprise equivalent stress amplitude and corresponding cycle times; And/or visually displaying fatigue early warning information triggered according to the comparison result of the fatigue parameter and a preset threshold value.
2. The digital twin steel deck fatigue monitoring method according to claim 1, wherein the identifying vehicle parameters from the traffic operation video information and generating vehicle load data based on the vehicle parameters comprises: Detecting and tracking vehicles according to the traffic operation video information by utilizing a deep learning algorithm to extract vehicle parameters of the vehicles, wherein the vehicle parameters comprise appearance size, wheelbase, wheel axle number, speed and track; Estimating the load parameters of the vehicle in real time based on the vehicle parameters, wherein the load parameters are structured data with unique labels and time coordinates, the load parameters are obtained by matching the vehicle parameters with a preset vehicle database or by inputting the vehicle parameters into a preset machine learning model for inference, and the load parameters comprise total mass, axle load and wheel track; and integrating the vehicle parameters and the load parameters according to the time sequence and the space track to generate a continuous moving load sequence as a concrete expression form of the vehicle load data.
3. The digital twin monitoring method for fatigue of the steel bridge deck according to claim 2, wherein the step of inputting the vehicle load data into a preset steel bridge deck structural model and calculating stress response data of a preset bridge deck fatigue sensitive part comprises the steps of inputting the moving load sequence into the preset steel bridge deck structural model, simulating a dynamic moving process of a vehicle and carrying out dynamic calculation, and outputting dynamic simulation stress time course original data of the preset bridge deck fatigue sensitive part as concrete expression forms of the stress response data, wherein the preset steel bridge deck structural model is established based on a design drawing and actual measurement data of a steel bridge.
4. The digital twin monitoring method for steel deck fatigue according to claim 3, wherein the step of integrating two stress components of the preset deck fatigue sensitive part in mutually orthogonal directions by adopting an equivalent structural stress method based on the stress response data and calculating fatigue parameters comprises the steps of: screening stress time course data of a preset bridge deck fatigue sensitive part which meets preset conditions from stress time course raw data output by dynamic simulation, wherein the preset bridge deck fatigue sensitive part comprises U ribs, a steel bridge roof weld joint and a cross slab and steel bridge roof junction; Based on the stress time interval data, adopting an equivalent structural stress method to integrate the normal stress and the shearing stress of the two mutually orthogonal directions of the fatigue sensitive part of the preset bridge deck so as to eliminate the stress concentration effect, and carrying out fatigue damage calculation so as to obtain equivalent structural stress time interval data; Based on the equivalent structure stress time-course data, extracting an equivalent stress amplitude and corresponding cycle times by adopting a rain flow counting method; Selecting a corresponding S-N curve according to the actual structure of the steel bridge and the category of the preset bridge deck fatigue sensitive part, and combining a Miner accumulated damage theory, and calculating the accumulated damage coefficient of the preset bridge deck fatigue sensitive part by using the extracted equivalent stress amplitude and the corresponding cycle times, wherein the equivalent stress amplitude, the corresponding cycle times and the accumulated damage coefficient form a fatigue parameter; And predicting the residual fatigue life of the fatigue sensitive part of the preset bridge deck based on the accumulated damage coefficient and by combining a preset linear prediction model and the design life of the steel bridge, and/or judging that the fatigue failure risk exists in the fatigue sensitive part of the preset bridge deck when the accumulated damage coefficient reaches or exceeds a preset critical value.
5. The steel bridge deck fatigue digital twin monitoring method according to claim 4, wherein the fatigue damage calculation formula is: ; Wherein, the Is the stress of the equivalent structure, which is that, In order to preset the positive stress of the fatigue sensitive part of the bridge deck in the stress direction, Is the shearing stress perpendicular to the stress direction of the fatigue sensitive part of the preset bridge deck, The weight coefficient determined based on the material characteristics of the steel bridge is in the range of 1.0-1.2.
6. The steel bridge deck fatigue digital twin monitoring method according to claim 4, wherein the calculation formula of the cumulative damage coefficient is: ; Wherein, the In order to accumulate the damage coefficient(s), Is the first The number of cycles corresponding to the magnitude of the equivalent stress, Is the first Fatigue life corresponding to the magnitude of the level equivalent stress, the fatigue life being determined from the S-N curve, Is the number of steps of the equivalent stress amplitude.
7. The steel bridge deck fatigue digital twin monitoring method according to claim 4, wherein the step of visually displaying the fatigue parameters on a three-dimensional digital twin platform of the steel bridge comprises the steps of mapping the fatigue parameters of the preset bridge deck fatigue sensitive part to the three-dimensional digital twin platform of the steel bridge and performing visual dynamic display in a cloud image or thermodynamic diagram mode, wherein the three-dimensional digital twin platform is constructed based on a BIM and GIS fusion framework and is constructed through combination of a finite element model and a BIM model.
8. The steel bridge deck fatigue digital twin monitoring method according to claim 7, wherein the visually displaying fatigue early warning information triggered according to the comparison result of the fatigue parameter and a preset threshold on the three-dimensional digital twin platform of the steel bridge comprises: presetting a preset threshold value and fatigue early warning information, wherein the preset threshold value comprises a preset equivalent stress amplitude threshold value, a preset cycle number threshold value and a preset accumulated damage coefficient threshold value, and the fatigue early warning information comprises first early warning information, second early warning information and third early warning information; When the equivalent stress amplitude reaches or exceeds a preset equivalent stress amplitude threshold value, visually displaying the first early warning information corresponding to the equivalent stress amplitude on a three-dimensional digital twin platform of the steel bridge through a platform popup window; when the cycle number reaches or exceeds a preset cycle number threshold, visually displaying the second early warning information corresponding to the cycle number on a three-dimensional digital twin platform of the steel bridge through a platform popup window; and when the accumulated damage coefficient reaches or exceeds a preset accumulated damage coefficient threshold value, visually displaying the third early warning information corresponding to the accumulated damage coefficient on the three-dimensional digital twin platform of the steel bridge through a platform popup window.
9. A digital twin monitoring system for steel bridge deck fatigue is characterized by comprising an information acquisition and load generation module, a calculation analysis module, a fatigue analysis module and a display early warning module, The information acquisition and load generation module is used for responding to the received traffic operation video information on the steel bridge deck acquired by the camera equipment, identifying vehicle parameters according to the traffic operation video information and generating vehicle load data based on the vehicle parameters; The calculation analysis module is used for inputting the vehicle load data into a preset steel bridge deck structural model and calculating stress response data of a fatigue sensitive part of a preset bridge deck, wherein the preset steel bridge deck structural model is an orthotropic steel bridge deck three-dimensional entity model established based on a finite element method; The fatigue analysis module is used for integrating stress components of the fatigue sensitive part of the preset bridge deck in the mutually orthogonal direction by adopting an equivalent structural stress method based on the stress response data and calculating fatigue parameters, wherein the fatigue parameters comprise equivalent stress amplitude and corresponding cycle times; the display early warning module is used for visually displaying the fatigue parameters on the three-dimensional digital twin platform of the steel bridge and/or visually displaying fatigue early warning information triggered according to the comparison result of the fatigue parameters and a preset threshold value.
10. The steel deck slab fatigue digital twin monitoring system according to claim 9, wherein the information acquisition and load generation module comprises an information acquisition unit and a load generation unit, wherein, The information acquisition unit is used for receiving traffic operation video information on the steel bridge deck acquired by the camera equipment; The load generating unit is used for identifying vehicle parameters according to the traffic operation video information and generating vehicle load data based on the vehicle parameters.

Description

Digital twin monitoring method and system for fatigue of steel bridge deck Technical Field The invention relates to the technical field of bridge structure health monitoring and digital twinning, in particular to a digital twinning monitoring method and system for fatigue of a steel bridge deck. Background Bridges are important components of traffic infrastructure and play a key role in modern urban traffic and regional economic development. With the rapid growth of highways, urban highways and heavy traffic, the traffic load borne by bridges is increasingly complex and huge. In particular to an orthotropic steel bridge deck structure widely used in a large-span steel structure bridge, and the service environment and the running load of the orthotropic steel bridge deck structure show diversified and high dynamic characteristics. Long-term repeated loading of the vehicle can cause local stress concentration, fatigue crack, early failure and other safety risks of the steel bridge deck structure. Therefore, real-time monitoring of fatigue stress amplitude and cycle number and scientific life prediction for bridge deck key parts have become important technical directions for urgent breakthrough in the civil engineering field. The existing bridge health monitoring mainly depends on physical sensor layout modes such as strain gauges, pressure sensors and the like, and long-term tracking is carried out on structural stress, deformation, damage and other states. Although the method can provide a certain data support for the safe operation of the bridge structure, the method has the problems of limited distribution points, limited monitoring range, easy damage of the sensor, high maintenance cost and the like. The on-site layout of a large number of sensors also affects the normal operation of the bridge, and is difficult to realize comprehensive, continuous and high-resolution monitoring of large-area and complex parts. In addition, the sensor data generally only represent specific monitoring points, and cannot completely reflect stress field distribution and fatigue evolution rules of the whole bridge deck and key parts (such as U ribs and diaphragm weld zones). Dynamic weighing systems are another technique for obtaining information on the load of a vehicle on a deck. The system adopts a pressure sensor buried under a bridge deck or a road to measure parameters such as weight, axle load, running speed and the like of a passing vehicle in real time. Although the dynamic weighing system is applied to partial bridges, the equipment cost is high, the installation and maintenance workload is large, the dynamic weighing system is easily influenced by temperature, road surface state and vehicle running conditions, and the long-term stability and accuracy have uncertainty, so that the popularization of the dynamic weighing system in a large-scale and complex traffic environment is limited. In recent years, the rapid development of computer vision and artificial intelligence technology provides a new opportunity for traffic monitoring and structural health assessment. The video monitoring technology based on the camera is combined with a deep learning algorithm (such as YOLO, deepSORT and the like), and functions of vehicle type identification, target tracking, track reconstruction, vehicle speed measurement and the like can be realized in a multi-lane and high-density traffic environment. By analyzing the continuous video stream data, the number, the type and the operation rule of vehicles passing through the bridge deck can be captured in real time without contact and damage. This creates conditions for efficient acquisition of traffic load data and large data driven structural analysis. However, technologies based on visual recognition currently cannot directly measure the actual weight of a vehicle and the actual stress state of a bridge structure. The estimation of the vehicle weight mainly depends on the matching of parameters such as the appearance size, the number of axles, the wheel track and the like of the vehicle and a standard vehicle type database, certain errors exist in the method, and even if more accurate vehicle load estimation is realized, how to efficiently couple the data with a bridge deck finite element model, drive structural simulation in real time, directly obtain fatigue stress amplitude and cycle number distribution of bridge deck key parts, and the fatigue stress amplitude and cycle number distribution are used for fatigue state estimation and risk early warning, and still face technical challenges. Traditional finite element analysis is a main means for researching stress and fatigue life of bridge structures. Although finite element modeling can finely analyze the complex structure and stress characteristics of an orthotropic steel bridge deck, most of the current analysis relies on standard load spectrum or idealized traffic conditions, and the time-varying and spatial distribution charac