A bridge stress state diagnosis method based on digital twinning and movable sensing

Abstract

This invention belongs to the field of bridge engineering technology, specifically relating to a bridge stress state diagnosis method based on digital twins and mobile sensors, aiming to achieve real-time diagnosis of the stress state of bridges with defects. The method includes acquiring spatial data of existing bridge defects, embedding defect parameters into a twin model, and generating an initial twin of the defective bridge. A modular monitoring station composed of mobile sensing devices is deployed on the bridge, with data transmitted between devices via a mesh self-organizing network to an edge computing terminal, where data preprocessing is performed. During the monitoring period, vehicle load, temperature and humidity, displacement, or vibration data are injected in real time into the initial twin of the defective bridge. An FEM-defect coupling algorithm is established to dynamically correct the structural stiffness matrix, forming a digital twin. A stress state diagnosis index system is established, comparing the responses of the digital and healthy twins under the same load, calculating the variation amplitude of each diagnostic index, and marking the structural health spectrum to determine whether the bridge has an abnormal stress state.

Description

Technical Field

[0001] This invention relates to the field of bridge engineering technology, and in particular to a method for diagnosing the stress state of bridges based on digital twins and mobile sensors. Background Technology

[0002] As a crucial transportation infrastructure, the operational safety and stress state of bridges are closely related. Existing bridge monitoring systems suffer from high initial investment, long equipment deployment cycles, and limitations due to their specificity and poor reusability, making it difficult to quickly respond to real-time monitoring needs. Furthermore, bridges are prone to damage such as pier displacement, main girder shifting, and cracks during service due to complex environments. Traditional load-bearing capacity assessment methods require traffic closures, are inefficient, and rely heavily on experience-based judgments, making it difficult to quantify the dynamic impact of these damages on structural stress, resulting in a lack of data support for safety assessments.

[0003] Current applications of digital twin technology in bridge engineering largely rely on building idealized models based on design drawings. These models often calculate stress, displacement, and other responses that deviate from the actual structural values. Particularly during the deterioration stage, they fail to accurately reflect the amplifying effect of defects on the stress state, leading to delayed risk warnings. Therefore, there is an urgent need for a method that balances rapid deployment, data-driven modeling, and accurate stress diagnosis to address the challenge of safety monitoring for bridges with defects. Summary of the Invention

[0004] The purpose of this invention is to provide a bridge stress state diagnosis method based on digital twins and mobile sensing. It aims to quantify the abnormal stress state of bridges caused by existing defects by driving the digital twin with short-term monitoring data, and to achieve accurate diagnosis of the real-time stress state of bridges with defects.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] This invention provides a bridge stress state diagnosis method based on digital twins and mobile sensors, comprising: S1: acquiring spatial data of existing bridge defects through UAV laser scanning or close-range photogrammetry, embedding defect parameters into the twin model, and generating an initial twin of the defective bridge; S2: deploying a modular monitoring station composed of mobile sensing devices on the bridge, transmitting data between devices to an edge computing terminal via a Mesh self-organizing network, and performing data preprocessing at the edge computing terminal; S3: during the monitoring period, injecting vehicle load, temperature and humidity, displacement, or vibration data into the initial twin of the defective bridge in real time, establishing an FEM-defect coupling algorithm to dynamically correct the structural stiffness matrix, simulating the stress state of the bridge under different working conditions, forming a digital twin, and realizing the dynamic mapping of the digital twin to the actual bridge; S4: establishing a stress state diagnosis index system, comparing the response of the digital twin and a healthy twin under the same load, calculating the change amplitude of each diagnostic index, marking the structural health spectrum, and determining whether the bridge has an abnormal stress state.

[0007] In step S1, a drone equipped with a laser scanning device is used to perform laser scanning on the bridge structure and its surrounding construction clearance to establish the geophysical coordinate system of the bridge and form a physical model of the bridge. Based on the bridge inspection data, the technical condition of the bridge components is marked on the physical model of the bridge, a component inspection list is established, and the closest similar component in physical location is selected as a reference component. Taking into account the longitudinal position, lateral position, vertical position and three-dimensional rotation angle of the component, the inspected component is subdivided into sub-components. The drone flight route is planned, and the drone equipped with a close-up camera takes close-up pictures of the defects, extracts the defects and their locations, determines the area affected by the defects, defines the stiffness damage factor, and generates an initial twin of the defective bridge.

[0008] The component inspection checklist includes components with a technical condition score of less than 60 points; the defect characteristics include the length and maximum width of cracks.

[0009] In step S2, taking into account the bridge structure type and bridge site environmental factors, the measuring points are divided into multiple application scenarios, and the sensor categories and quantities under different application scenarios are designed; the sensor unit with the highest mobile sensing score is determined, a modular monitoring station construction plan is formulated and deployed; the sensors transmit data to the edge computing terminal through the Mesh self-organizing network, and the edge computing terminal performs data preprocessing.

[0010] When determining the sensor unit with the highest mobile sensing score, a corresponding scoring and weighting system is established. The weighting system includes the sensor's ease of installation, reusability, asset depreciation, testing accuracy, and consumable price.

[0011] The data preprocessing performed by the edge computing terminal includes performing FFT transformation and removing noise components whose frequencies are not in the effective frequency band of the bridge.

[0012] In step S3, the noise-reduced vehicle load spectrum, ambient temperature and humidity, and vibration frequency parameters are input into the initial twin of the bridge with defects; the structural stiffness matrix is ​​dynamically corrected through the FEM-disease coupling algorithm; the time-varying influence coefficient is inverted and optimized based on short-term monitoring data, and the converged dynamic stiffness matrix is ​​input into the initial twin of the bridge with defects to form a digital twin and realize dynamic mapping.

[0013] Short-term monitoring data is for one week; the dynamic correction of the structural stiffness matrix is ​​based on the initial stiffness matrix of the healthy structure, the stiffness damage factor, and the time-varying influence coefficient.

[0014] In step S4, a cloud-based twin engine is deployed to periodically update the structural response distribution cloud map; the change range of each diagnostic indicator between the digital twin and the healthy twin under the same load is calculated; high-risk, medium-risk, and low-risk areas are marked according to the change range, and corresponding early warnings are issued; the remaining carrying capacity coefficient is calculated, and its change curve and envelope diagram over time are plotted to visualize the priority of architecture reinforcement.

[0015] The time interval for regular updates is 10 minutes; the residual bearing capacity coefficient is the ratio of the residual bearing capacity of the section to the bearing capacity in the healthy state, and is a function of the stiffness damage factor and the time-varying influence coefficient.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0017] 1. This application provides a bridge stress state diagnosis method based on digital twins and mobile sensors. By combining UAV laser scanning and close-range photogrammetry, actual fault parameters are embedded into the basic model of the bridge's digital twin to generate an initial twin with faults. This establishes a quantitative mapping between faults and stiffness damage, overcoming the problems of traditional digital twin models ignoring actual faults and having large deviations from the real structure, thus laying a precise foundation for subsequent stress analysis. Modular mobile monitoring stations are adopted, and the optimal sensor units are selected using a scoring and weighting system. The equipment is reusable and easy to deploy, significantly reducing asset investment and depreciation losses. Data preprocessing through Mesh self-organizing networks and edge computing ensures efficient and stable monitoring, solving the problems of high cost and poor reusability of traditional fixed monitoring systems.

[0018] 2. Based on short-term monitoring data, time-varying parameters are inverted and optimized to drive the dynamic correction of the stiffness matrix of the twin. Only one week's worth of data is needed to achieve dynamic mapping of the actual bridge. By comparing the response differences between the digital twin and the healthy twin, and combining the visualization analysis of the remaining bearing capacity coefficient, the structural health is quantified and graded early warning is given. This provides data support for reinforcement decisions, avoids risk lag caused by relying on experience judgment, and improves the scientific nature and timeliness of bridge safety management. Attached Figure Description

[0019] Figure 1This is a flowchart of a bridge stress state diagnosis method based on digital twin and mobile sensing provided in an embodiment of this application;

[0020] Figure 2 This is a flowchart illustrating the generation process of a digital twin, as provided in an embodiment of this application.

[0021] Figure 3 This is a flowchart illustrating the formulation of a motion monitoring scheme provided in an embodiment of this application;

[0022] Figure 4 This is a flowchart of bridge structure diagnosis and decision-making provided in an embodiment of this application. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0024] This application provides a method for diagnosing the stress state of bridges based on digital twins and mobile sensing, exemplified by, for example... Figure 1 As shown. The method includes:

[0025] S1: Obtain spatial data of existing bridge defects through UAV laser scanning or close-range photogrammetry, implant defect parameters into the twin model, and generate an initial twin of the defective bridge.

[0026] Combination Figure 1 and Figure 2 In step S1, a UAV equipped with a laser scanning device is used to perform laser scanning on the bridge structure and its surrounding construction clearance. For example, the scanning range should cover the UAV's obstacle avoidance zone. The geophysical coordinate system of the bridge is established to form a bridge physical model. Based on the bridge inspection data, the technical condition of the bridge components is marked on the bridge physical model, a component inspection list is established, and the closest similar component in physical location is selected as a reference component. Taking into account the longitudinal position, lateral position, vertical position, and three-dimensional rotation angle of the component, the inspected component is subdivided into sub-components. The flight route of the UAV is planned, and the UAV equipped with a close-up camera takes close-up pictures of the defects, extracts the defects and their locations, determines the area affected by the defects, defines the stiffness damage factor, and generates an initial twin of the defective bridge.

[0027] The component inspection checklist includes components with a technical condition score of less than 60. Defect characteristics include the length and maximum width of cracks.

[0028] For example, taking L / 4 to 3L / 4 of the box girder as an example, the analysis elements are divided, and the length of the analysis element is... The width can be taken as 1 / 2 to 1 times the main beam width; with transverse cracks as the first statistical item, analyze the longitudinal location and crack length of each crack. Crack width Are there vertical cracks within the ±1 / 2 stirrup spacing range, and what is the length of the vertical cracks? ,width Further calculation of the crack-affected area. Initial value, its length ,Width ,high It can be calculated according to formula (1);

[0029] (1)

[0030] in, The effective height of a component (such as a main beam) is usually the distance from the neutral axis of the cross section to the tension or compression edge. This refers to the flange width of a component (such as a main beam). This refers to the width of the web of a structural member (such as a main beam).

[0031] The stiffness damage factor D is defined based on the type of damage. First, the first... Stiffness of the relative cross-section neutral axis of the affected area and the stiffness of the original cross section Then, the stiffness damage factor of the analysis unit is calculated according to formula (2);

[0032] (2)

[0033] S2: Deploy modular monitoring stations consisting of mobile sensing devices on the bridge. The devices transmit data to the edge computing terminal through a Mesh self-organizing network, and the edge computing terminal performs data preprocessing.

[0034] Reference Figure 3 In step S2, taking into account the bridge structure type and bridge site environmental factors, the measuring points are divided into multiple application scenarios, and the sensor categories and quantities under different application scenarios are designed; the sensor unit with the highest mobile sensing score is determined, and a modular monitoring station construction plan is formulated and deployed; the sensors transmit data to the edge computing terminal through the ZigBee Mesh network, and the edge computing terminal performs data preprocessing.

[0035] For example, taking into account the bridge structure type, site environment, materials, test type, and test accessibility, the test points were divided into... 1. Application scenarios, and design sensor categories for different application scenarios. quantity .

[0036] When determining the sensor unit with the highest mobile sensing score, a corresponding scoring and weighting system is established. The weighting system includes the sensor's ease of installation, reusability, asset depreciation, testing accuracy, and consumable price.

[0037] One possible approach is to create a list of different types of sensor devices and then pair each sensor individually. In application scenarios The scores are assigned based on factors such as ease of installation, reusability, asset depreciation, testing accuracy, and consumable prices. One by one, the sensors In application scenarios The importance of factors such as ease of installation, reusability, asset depreciation, testing accuracy, and consumable prices should be weighted. .

[0038] All sensors that meet the sensor categories required for this test will be included. These are called sensor units, which can be arranged and combined from the sensor list. For each type of sensor unit, the movable sensing score is calculated according to formula (3).

[0039] (3)

[0040] After calculation, the sensor unit with the highest mobile sensing score was identified as the modular monitoring station. A construction plan for the modular monitoring station was developed based on the sensor's function, installation and disassembly requirements, and data acquisition requirements. The sensor automatically transmits data to the edge computing terminal through the ZigBee Mesh network.

[0041] The data preprocessing performed by the edge computing terminal includes performing an FFT transform to remove noise components whose frequencies are outside the bridge's effective frequency band. The FFT transform converts the raw signal acquired by the sensor from the time domain to the frequency domain, thereby accurately identifying the frequency components in the signal. Since bridge structures have specific effective frequency bands related to their stiffness, mass, and other characteristics when subjected to stress and vibration, and environmental noise signals such as wind noise and unrelated vibrations typically exist outside this effective frequency band, removing these noise components can significantly improve the signal-to-noise ratio of the monitoring data.

[0042] The processed clean data can more realistically reflect the actual stress response of the bridge (such as vibration and displacement), providing reliable input for subsequent steps such as stiffness matrix correction and time-varying parameter inversion in the digital twin, ensuring the accuracy of the twin's dynamic mapping of the bridge's real state, and ultimately improving the accuracy of stress state diagnosis.

[0043] S3: During the monitoring period, vehicle load, temperature and humidity, displacement or vibration data are injected into the initial twin of the bridge with defects in real time. The FEM-disease coupling algorithm is established to dynamically correct the structural stiffness matrix, simulate the stress state of the bridge under different working conditions, form a digital twin, and realize the dynamic mapping of the digital twin to the actual bridge.

[0044] In step S3, the noise-reduced vehicle load spectrum, ambient temperature and humidity, and vibration frequency parameters are input into the initial twin of the bridge with defects; the structural stiffness matrix is ​​dynamically corrected through the FEM-disease coupling algorithm; the time-varying influence coefficient is inverted and optimized based on short-term monitoring data, and the converged dynamic stiffness matrix is ​​input into the initial twin of the bridge with defects to form a digital twin and realize dynamic mapping.

[0045] For example, the formula for dynamic correction of the structural stiffness matrix is ​​shown in equation (4):

[0046] (4)

[0047] In the formula: The initial stiffness matrix of the healthy structure is derived from the as-built 3D model of the bridge. The first calculated according to equation (2) Stiffness damage factor of the affected area; For the first The time-varying influence coefficient of each disease-affected area at time t.

[0048] As one possible approach, short-term monitoring data is collected over one week. Dynamic correction of the structural stiffness matrix is ​​based on the initial stiffness matrix of the healthy structure, the stiffness damage factor, and the time-varying influence coefficient.

[0049] For example, the influence coefficient of time-varying data based on short-term (one-week) monitoring data. Inverse optimization involves first constructing the objective function as shown in equation (5), then solving it using gradient descent according to equation (6), and iterating until convergence.

[0050] (5)

[0051] (6)

[0052] in, This refers to the structural response data obtained from the kth actual monitoring. Based on the current time-varying coefficients The k-th structural response data obtained from twin simulation. Let be the value of the i-th time-varying influence coefficient in the j-th iteration, used for gradient descent optimization. The "learning rate" of gradient descent controls the step size for each iteration; for example, it is set to 0.01.

[0053] The converged dynamic stiffness matrix is ​​input into the digital twin to achieve dynamic mapping of the digital twin to the actual bridge structure.

[0054] Step S3 involves injecting real-time monitored vehicle load, environmental parameters, and structural response data into the initial digital twin of the bridge with defects. The stiffness matrix is ​​then dynamically corrected using the FEM-disease coupling algorithm, enabling the digital twin model to keep pace with the actual state of the bridge in real time. This overcomes the limitation of traditional static models in reflecting the time-varying characteristics of the structure. By correcting the stiffness matrix and integrating the stiffness damage factor with the real-time time-varying influence coefficient, the dynamic impact of defects on structural stiffness under different working conditions is quantified. This solves the problem of the disconnect between defects and stress state in traditional analysis, making the simulation results more closely match the actual stress conditions of the bridge.

[0055] In addition, by inverting and optimizing time-varying parameters based on a week of short-term monitoring data, the digital twin model can be quickly converged, which greatly improves the construction efficiency of the digital twin and meets the timeliness requirements of bridge real-time monitoring.

[0056] S4: Establish a stress state diagnostic index system, compare the response of the digital twin and the healthy twin under the same load, calculate the change range of each diagnostic index, mark the structural health spectrum, and determine whether the bridge has an abnormal stress state.

[0057] Reference Figure 4 In step S4, a cloud twin engine is deployed to periodically update the structural response distribution cloud map.

[0058] As one possible implementation, the update interval is 10 minutes.

[0059] Calculate the variation range of various diagnostic indicators between the digital twin and the healthy twin under the same load. Mark high-risk, medium-risk, and low-risk areas based on the variation range and issue corresponding early warnings; calculate the remaining carrying capacity coefficient, plot its variation curve and envelope diagram over time, and visualize the priority of architecture hardening.

[0060] Among them, the residual bearing capacity coefficient Remaining bearing capacity of the cross section With health status carrying capacity The ratio is the stiffness damage factor. and time-varying influence coefficient function , It can be calculated based on the principles of structural mechanics. Taking bending moment bearing capacity as an example, .

[0061] For example, the responses of diseased twins and healthy twins under the same load are analyzed and calculated one by one, and the changes in each diagnostic indicator are calculated. ;in For twins with disease in time The response For healthy twins in time The response.

[0062] Will Locations with responses exceeding 1.2, 1.1, and 1.0 times those of a healthy twin model are marked as high-risk, medium-risk, and low-risk areas, respectively, and are given red, orange, and yellow alerts. The user interface uses a WebGL compressed model for lightweight rendering and supports viewing key structural responses such as stress at the upper and lower edges of cross-sections.

[0063] draw Over time The change curves are analyzed to identify characteristic points. Further investigation is conducted to determine the effects of vehicle load spectra, environmental temperature and humidity at those characteristic points, and to analyze the causes of anomalies. The remaining bearing capacity coefficients of each analysis unit of the bridge structure are plotted. Within the monitoring period, regarding time The envelope diagram, and according to the remaining bearing capacity coefficient The size of the envelope graph is used for color mapping to visualize the architecture and prioritize hardening. The smallest reinforcement has the highest priority.

[0064] In summary, this application's embodiments establish a bridge physical model using a UAV equipped with laser scanning equipment and plan a flight path for a close-range camera based on bridge maintenance data. This improves the efficiency of UAV close-range camera defect scanning and enhances the quality of defective twins of the bridge. Furthermore, a mapping model of defect parameters and stiffness damage factors is established, quantifying the impact of bridge defects on structural stress and providing a more accurate mapping of the actual structure. A modular monitoring station design is proposed, comprehensively considering factors such as ease of installation, reusability, testing accuracy, and economy. A method for formulating a modular monitoring station scheme using mobile sensor scores as an indicator is also proposed, significantly reducing bridge monitoring costs and making the widespread application of this method possible.

[0065] Furthermore, the time-varying influence coefficient can be retrieved using only one week's worth of monitoring data. This technology drives the correction of defective twin models, significantly improving the efficiency of establishing defective twins. By monitoring the simulation results of bridge twins over a period of time, the remaining bearing capacity coefficient of components is quantified. The user end visualizes the reinforcement priority of components through lightweight rendering, reducing the user threshold and improving ease of use. It solves three major pain points in the industry: high cost, poor timeliness, and high technical threshold for users, providing a new method for intelligent management and monitoring of infrastructure.

[0066] In the description of this specification, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0067] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for diagnosing the stress state of bridges based on digital twins and mobile sensing, characterized in that, include: S1: Obtain spatial data of existing bridge defects through UAV laser scanning or close-range photogrammetry, implant defect parameters into twin model, and generate initial twin of the defective bridge; S2: Deploy modular monitoring stations consisting of mobile sensing devices on the bridge. The devices transmit data to the edge computing terminal through a Mesh self-organizing network, and the edge computing terminal performs data preprocessing. S3: During the monitoring period, vehicle load, temperature and humidity, displacement or vibration data are injected into the initial twin of the bridge with defects in real time. The FEM-disease coupling algorithm is established to dynamically correct the structural stiffness matrix, simulate the stress state of the bridge under different working conditions, form a digital twin, and realize the dynamic mapping of the digital twin to the actual bridge. S4: Establish a stress state diagnostic index system, compare the response of the digital twin and the healthy twin under the same load, calculate the change range of each diagnostic index, mark the structural health spectrum, and determine whether the bridge has an abnormal stress state. In step S3, the noise-reduced vehicle load spectrum, ambient temperature and humidity, and vibration frequency parameters are input into the initial twin of the bridge with defects; the structural stiffness matrix is ​​dynamically corrected through the FEM-disease coupling algorithm; the time-varying influence coefficient is inverted and optimized based on short-term monitoring data, and the converged dynamic stiffness matrix is ​​input into the initial twin of the bridge with defects to form a digital twin and realize dynamic mapping. The formula for the structural stiffness matrix is: in: The initial stiffness matrix of the healthy structure is derived from the as-built 3D model of the bridge. The first one defined according to the disease type Stiffness damage factor of each affected area; stiffness damage factor is calculated according to the formula. get; For the first The stiffness of the relative cross-section neutral axis of the affected area; The original cross-sectional stiffness; The length of the analysis cell; For the first The time-varying influence coefficient of each disease-affected area at time t; Based on short-term monitoring data, the time-varying influence coefficient Inversion optimization is performed, specifically including: Construct the objective function: according to Solve using gradient descent and iterate until convergence; in, This refers to the structural response data obtained from the kth actual monitoring. Based on the current time-varying coefficients The k-th structural response data obtained from twin simulation; Let be the value of the i-th time-varying influence coefficient in the j-th iteration; is the learning rate for gradient descent.

2. The bridge stress state diagnosis method based on digital twin and mobile sensing according to claim 1, characterized in that, In step S1, a drone equipped with a laser scanning device is used to perform laser scanning on the bridge structure and its surrounding construction clearance to establish the geophysical coordinate system of the bridge and form a physical model of the bridge. Based on the bridge inspection data, the technical condition of the bridge components is marked on the physical model of the bridge, a component inspection list is established, and the closest similar component in physical location is selected as the reference component. Taking into account the longitudinal, lateral, vertical, and three-dimensional rotation angles of the components, the components to be inspected are subdivided into sub-components; the flight path of the UAV is planned, and the UAV is equipped with a close-up camera to take close-up pictures of the defects, extract the defects and their locations, determine the affected areas of the defects, define the stiffness damage factor, and generate the initial twin of the bridge with defects.

3. The method for diagnosing the stress state of a bridge based on digital twin and mobile sensing according to claim 2, characterized in that, The component inspection checklist includes components with a technical condition score of less than 60 points; the defect characteristics include the length and maximum width of cracks.

4. The bridge stress state diagnosis method based on digital twin and mobile sensing according to claim 1, characterized in that, In step S2, taking into account the bridge structure type and bridge site environmental factors, the measuring points are divided into multiple application scenarios, and the types and quantities of sensors for different application scenarios are designed. Identify the sensor unit with the highest mobile sensing score, formulate and deploy a modular monitoring station construction plan; the sensor transmits data to the edge computing terminal through a Mesh self-organizing network, and the edge computing terminal performs data preprocessing.

5. The bridge stress state diagnosis method based on digital twin and mobile sensing according to claim 4, characterized in that, When determining the sensor unit with the highest mobile sensing score, a corresponding scoring and weighting system is established. The weighting system includes the sensor's ease of installation, reusability, asset depreciation, testing accuracy, and consumable price.

6. The bridge stress state diagnosis method based on digital twin and mobile sensing according to claim 4, characterized in that, The data preprocessing performed by the edge computing terminal includes performing FFT transformation to remove noise components whose frequencies are not in the effective frequency band of the bridge.

7. The bridge stress state diagnosis method based on digital twin and mobile sensing according to claim 4, characterized in that, In step S4, a cloud-based twin engine is deployed to periodically update the structural response distribution cloud map; the variation range of each diagnostic indicator between the digital twin and the healthy twin under the same load is calculated; high-risk, medium-risk, and low-risk areas are marked according to the variation range, and corresponding early warnings are issued. Calculate the remaining carrying capacity coefficient, plot its change curve and envelope diagram over time, and visualize the priority of architecture hardening.

8. The bridge stress state diagnosis method based on digital twin and mobile sensing according to claim 7, characterized in that, The time interval for the periodic update is 10 minutes; the remaining bearing capacity coefficient is the ratio of the remaining bearing capacity of the section to the bearing capacity in the healthy state, and is a function of the stiffness damage factor and the time-varying influence coefficient.

Images