Building structure virtual monitoring data completion and digital twin risk control method

Abstract

The application discloses a kind of building structure virtual monitoring data completion and digital twin risk control method, it is related to building structure safety monitoring and digital twin technical field.Acquire multimodal monitoring data, structure point cloud, BIM data and working condition parameter;Time synchronization is carried out to data, and abnormality detection and credibility evaluation are carried out, and missing, abnormality and low credibility section are identified;Virtual monitoring candidate data is generated based on historical sequence, topological relationship, spatial correlation, load-response relationship;Physical consistency check is carried out through finite element model, error feedback correction generation process is obtained, and credible virtual data is obtained;With measured data fusion forms enhanced data set, drives digital twin model geometry, material, connection, damage parameter dynamic update;Risk state is calculated based on updated model, risk propagation path and key component are identified, and early warning and control strategy are output.The application realizes the closed-loop linkage of physical credibility of data completion and risk control.

Description

Technical Field

[0001] This invention belongs to the field of building structure safety monitoring, digital twin and smart construction technology, specifically involving a method for virtual monitoring data completion and digital twin risk control of building structures. Background Technology

[0002] Building structures face complex and ever-changing mechanical and environmental effects throughout construction and service. During construction, risks include hoisting impacts, sudden stress changes in formwork support systems, early-age shrinkage and creep of concrete, frequent switching of temporary working conditions, and cumulative assembly deviations. During service, they face long-term static loads, dynamic loads, temperature and humidity variations, environmental corrosion, material aging, crack propagation, and degradation of connection stiffness, which can easily lead to localized damage or even overall failure. Current engineering projects commonly employ a network of multiple types of sensors, including strain, displacement, vibration, acoustic emission, and temperature and humidity sensors, for monitoring, combined with BIM models or 3D point clouds for structural condition analysis and safety assessment.

[0003] However, existing technologies suffer from several insurmountable flaws. First, monitoring data completion technologies severely lack physical constraints: when sensor failure, communication interruption, or signal drift leads to missing, abnormal, or low-reliability data, the industry commonly uses purely data-driven methods such as linear interpolation, spline fitting, grey models, and simple time-series prediction for completion. These methods can only restore numerical continuity and completely disregard mechanical laws such as structural static equilibrium, deformation coordination, boundary constraints, and material nonlinearity. Under scenarios such as heavy loads, damage, loose connections, and sudden changes in operating conditions, the completed data clearly violates the true structural response, directly leading to distortion in subsequent state assessments.

[0004] Secondly, the robustness of digital twin model update mechanisms is extremely poor: existing BIM / point cloud-driven twin models mostly rely on direct mapping of measured data and lack a reliable repair mechanism for low-confidence sections. Once the monitoring data is interrupted or abnormal, the model cannot correct itself and quickly experiences simulation drift, making it difficult to reflect the real state such as geometric deviation, material degradation, and damage evolution. Under conditions of sparse sensors, local failures, and complex working conditions, it is completely unable to maintain update accuracy and loses the value of digital mirroring.

[0005] Furthermore, AI data generation and finite element mechanical analysis are completely separate: the AI-based virtual data solution is only responsible for generating data, while the finite element method is only used for post-event verification. There is no information exchange between the two, and there is no mechanism for reverse correction of physical residuals. They are simply parallel combinations that cannot form a closed-loop coupling, which makes it difficult for the virtual data to meet the mechanical rationality requirements, the model correction lacks precise constraints, and the risk identification results are unreliable.

[0006] Finally, risk identification lacks global propagation analysis capabilities: existing risk assessments are mostly limited to single-point exceedance judgments, failing to identify risk diffusion paths and key control components based on structural mechanics transmission relationships. This results in delayed early warnings, blind responses, and difficulty in achieving proactive prevention and control, thus failing to meet the needs of refined safety management throughout the construction and operation processes. These problems, compounded by each other, lead to extremely low reliability of building structural safety monitoring and risk control under incomplete data and complex operating conditions, becoming a core bottleneck restricting the development of smart construction and structural health monitoring. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to provide a method for virtual monitoring data completion and digital twin risk control of building structures, which addresses the shortcomings of the prior art and solves the technical problems of incomplete monitoring data, lack of engineering credibility of the completed data, inaccurate digital twin updates, and insufficient risk control closed loop.

[0008] The present invention adopts the following technical solution: A method for virtual monitoring data completion and digital twin risk control of building structures includes the following steps: S1. Acquire multimodal measured monitoring data, structural point cloud data, building information model (BIM) data, and working condition parameter data of the building structure, wherein the multimodal measured monitoring data includes at least one or more of strain data, displacement data, vibration data, acoustic data, and environmental data; S2. Perform time synchronization, anomaly detection, and credibility assessment on the multimodal measured monitoring data to identify missing segments, abnormal segments, or low-credibility segments. S3. For the missing, abnormal, or low-confidence sections, generate virtual monitoring candidate data based on historical monitoring sequences, component topology relationships, spatial correlation characteristics of adjacent sensors, load-response nonlinear mapping relationships, the operating condition parameter data, and environmental parameters. S4. Based on the BIM data, establish or call the finite element model corresponding to the target component, apply the virtual monitoring candidate data as external displacement boundary conditions, equivalent nodal loads or state correction inputs to the finite element model, and extract the calculated stress response, calculated displacement response, deformation compatibility response and safety margin index of the target component and adjacent related components. S5. Based on the physical consistency cost function, perform physical consistency verification on the virtual monitoring candidate data and the extracted response index. Retain virtual monitoring candidate data that meet the preset consistency criteria as credible virtual monitoring data, eliminate virtual monitoring candidate data that do not meet the preset consistency criteria, and feed back the physical consistency error corresponding to the non-compliance of the preset consistency criteria as physical constraint correction information to the virtual monitoring candidate data generation process. After adjusting the generation process, regenerate the virtual monitoring candidate data until credible virtual monitoring data that meets the preset consistency criteria is obtained or the preset number of iterations is reached. S6. The trusted virtual monitoring data is fused with the multimodal measured monitoring data of the corresponding time period to form an enhanced monitoring dataset; S7. Based on the enhanced monitoring dataset, the structural point cloud data, and the BIM data, dynamically update the component geometric parameters, material performance parameters, connection parameters, and damage state parameters in the digital twin model of the building structure to obtain the current digital twin state. S8. Calculate the risk status value of each component based on the digital twin status, and identify the risk propagation path and key control components by combining the component connection relationship, force transmission relationship or construction procedure relationship, and output the risk warning result and control strategy.

[0009] Preferably, step S2 specifically includes: Perform unified timestamp alignment on different types of monitoring data; Statistical thresholding, temporal abrupt change detection, or cross-modal consistency detection methods are used to identify abnormal monitoring sections. The reliability score of each monitoring segment is calculated based on the energy entropy, cross-modal mutual information, data continuity, signal quality index, cross-sensor consistency, and structural mechanical rationality of the monitoring sequence. Monitoring segments with a credibility score below a preset threshold are classified as low-credibility segments.

[0010] Preferably, in step S3, the virtual monitoring candidate data is generated in any of the following ways: (1) Generative adversarial network-based generation; (2) Generated based on diffusion probability model; (3) Generative probabilistic networks and diffusion probability models are cascaded together; Furthermore, the generation process is a conditional generation process, using historical monitoring sequence characteristics, component categories, structural spatial locations, spatial correlation characteristics of adjacent sensors, load condition parameters, and environmental parameters as conditional inputs.

[0011] Preferably, in step S5, the physical consistency cost function for:

[0012] in, The displacement response corresponding to the virtual monitoring candidate data. To calculate the displacement response using the finite element method, The stress response is obtained by mapping virtual monitoring candidate data. For finite element method (FEM) calculation of stress response, The deformation coordination error between adjacent nodes , , Preset weights.

[0013] Preferably, the preset consistency criterion is: the value of the physical consistency cost function is less than the preset cost threshold, and the component safety margin index output by the finite element model is greater than the preset lower limit; Physical consistency error is constructed as a physical constraint penalty term and used to update the parameters, conditional inputs, or sampling distribution of the virtual monitoring candidate data generation model.

[0014] Preferably, in step S6, the enhanced monitoring dataset is formed using at least one of the following methods: 1) Weighted fusion based on data credibility; 2) Integrate components by region; 3) Perform recursive fusion based on time windows; 4) After screening based on physical consistency scores, the samples are merged.

[0015] Preferably, step S7 specifically includes: Perform component-level spatial registration between structural point cloud data and BIM data to obtain the geometric deviation between the actual component and the designed component. The stress state, material degradation state, and connection stiffness state of the components are corrected based on the enhanced monitoring dataset; The elastic modulus, connection stiffness, and damage factor in the digital twin model are used as the state variables to be identified, and the displacement and strain in the enhanced monitoring dataset are used as the observation variables. The state variables to be identified are recursively updated using unscented Kalman filtering, extended Kalman filtering, recursive least squares method, or Bayesian model correction method. Output the digital twin state snapshot corresponding to the current moment.

[0016] Preferably, in step S8, the building structural components and their connections are modeled as a graph network, where nodes represent structural components and edges represent mechanical transfer efficiency, load redistribution coefficients, or construction process coupling relationships between components, and risk propagation paths are identified based on changes in edge weights in the graph network.

[0017] Preferably, in step S8, the control strategy includes one or more of the following: 1) Adjust the sensor sampling frequency; 2) Adjust the density of monitoring points; 3) Adjust the construction sequence; 4) Adjust the hoisting path or operating radius; 5) Implement local reinforcement measures; 6) Develop a maintenance priority list.

[0018] Secondly, embodiments of the present invention provide a building structure virtual monitoring data completion and digital twin risk control system, including: The data acquisition module is used to acquire multimodal measured monitoring data, structural point cloud data, BIM data, and working condition parameter data of the building structure. The credibility assessment module is used to perform time synchronization, anomaly detection, and credibility assessment on the multimodal measured monitoring data, and to identify missing segments, abnormal segments, or low-credibility segments. The virtual data generation module is used to generate virtual monitoring candidate data based on historical monitoring sequences, component topological relationships, spatial correlation characteristics of adjacent sensors, load-response nonlinear mapping relationships, operating condition parameter data, and environmental parameters. The physical consistency verification module is used to establish or call finite element models, extract response indicators and complete physical consistency verification, and filter to obtain reliable virtual monitoring data. The feedback correction module is used to feed back physical consistency errors to the virtual data generation module and adjust the virtual monitoring candidate data generation process. The data fusion module is used to fuse trusted virtual monitoring data with multimodal measured monitoring data to form an enhanced monitoring dataset; The digital twin update module is used to dynamically update the component geometric parameters, material performance parameters, connection parameters, and damage state parameters in the digital twin model of the building structure based on the enhanced monitoring dataset, structural point cloud data, and BIM data. The risk analysis module is used to calculate the risk status value of components, identify risk propagation paths, and key control components; The control output module is used to output risk warning results and control strategies.

[0019] Thirdly, a computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method for virtual monitoring data completion and digital twin risk control of building structures.

[0020] Fourthly, embodiments of the present invention provide a computer-readable storage medium including a computer program, which, when executed by a processor, implements the steps of the above-described method for virtual monitoring data completion and digital twin risk control of building structures.

[0021] Fifthly, a chip includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method for virtual monitoring data completion and digital twin risk control of building structures.

[0022] Sixthly, embodiments of the present invention provide an electronic device, including a computer program, which, when executed by the electronic device, implements the steps of the above-described method for virtual monitoring data completion and digital twin risk control of building structures.

[0023] Compared with the prior art, the present invention has at least the following beneficial effects: A method for virtual monitoring data completion and digital twin risk control of building structures is proposed. This method acquires four types of data—multimodal monitoring, point cloud, BIM, and working condition data—to ensure complete input information. It accurately locates problematic data segments through time synchronization, anomaly detection, and reliability assessment. Virtual candidate data is generated based on spatiotemporal correlation and load response relationships, ensuring the data generation has an engineering basis. Mechanical verification is performed using a finite element model to ensure the virtual data conforms to the physical laws of the structure. An iterative correction closed loop is formed through error feedback to improve the reliability of the virtual data. An enhanced dataset is formed by fusing measured and virtual data to drive the dynamic update of the digital twin, ultimately achieving accurate risk identification and control output. This invention integrates data completion, model updating, and risk control, maintaining high-reliability state perception even under sensor failure, data loss, and complex working conditions. It significantly improves the stability and practicality of building structure safety management throughout its entire lifecycle, avoiding mechanical distortion and twin model drift caused by purely data-driven completion.

[0024] Furthermore, by unifying timestamp alignment, analysis errors caused by time differences in data acquisition from different sensors are eliminated. A combination of statistical thresholding, temporal abrupt changes, and cross-modal consistency methods is used to identify anomalies, covering various anomaly types such as drift, abrupt changes, disconnections, and flatness. A credibility score is calculated based on six indicators: energy entropy, cross-modal mutual information, data continuity, signal quality, cross-sensor consistency, and mechanical rationality, achieving objective quantification of data quality. Low-credibility segments are identified through preset thresholds, providing precise targets for subsequent virtual completion. This solution significantly improves the accuracy of identifying defective data, prevents invalid data from entering subsequent processes, reduces virtual data generation bias, provides a reliable foundation for physical consistency verification, and ensures that the entire process is supported by high-quality data from the source.

[0025] Furthermore, this approach employs three methods—Generative Adversarial Networks (GANs), diffusion probability models, or a cascade of both—to adapt to different completion scenarios, including short-term missing data, long-term failures, and abrupt changes in operating conditions. Using historical sequences, component categories, spatial locations, adjacent sensor associations, operating conditions, and environmental conditions as inputs, the structural topology, mechanical laws, and spatiotemporal constraints are embedded into the generation process, giving the virtual data structural response characteristics. Compared to purely statistical methods, this scheme can accurately recreate the load-response nonlinear relationship, generating data that better reflects actual engineering changes. It maintains a high degree of fit even under complex conditions such as heavy loads, damage, and loose connections, providing high-quality candidate data for subsequent physical verification and improving the realism and engineering applicability of virtual completion.

[0026] Furthermore, the cost function integrates the root mean square error of displacement, the root mean square error of stress, and the deformation compatibility error to comprehensively measure the consistency between the virtual data and the finite element method (FEM) calculation results. A dual criterion of preset thresholds and a safety margin lower limit ensures both controllable data errors and that the components remain within a safe stress range. This fundamentally filters out unreasonable virtual data from a mechanical perspective, eliminating completion results that violate static equilibrium, deformation compatibility, and boundary constraints. It prevents digital twin model simulation drift at its source, ensuring that subsequent state assessments and risk identification do not deviate from the actual structural state, significantly improving the engineering credibility of the virtual data.

[0027] Furthermore, the physical consistency error is constructed as a penalty term to update the parameters of the generated model, the weights of the conditional inputs, or the sampling distribution. This allows the virtual data generation process to continuously receive mechanical constraint feedback, gradually approximating the response of the real structure. This breaks the two-stage separation of generation and verification, achieving deep coupling between data generation and physical constraints. This significantly reduces the number of iterations, rapidly converges to obtain virtual data that meets mechanical requirements, improves completion efficiency and accuracy, and enables the entire solution to possess self-optimization and self-correction capabilities, adapting to long-term monitoring and dynamic changes in operating conditions.

[0028] Furthermore, four methods are employed: credibility weighting, component partitioning, time window recursion, and physical consistency scoring. Weights are flexibly allocated based on data quality, spatial location, and time series, highlighting the contribution of high-credibility data and mitigating the impact of low-quality data. This fully leverages the advantages of the authenticity of measured data and the integrity of virtual data, improving the overall signal-to-noise ratio and continuity of the monitoring sequence. This provides high-quality input for digital twin updates, ensuring more accurate model parameter correction and more realistic structural state restoration.

[0029] Furthermore, actual geometric deviations are obtained through point cloud and BIM component-level registration; stress, material, and connection stiffness are corrected by combining enhanced data; and Kalman filtering and other methods are used to recursively update the data with elastic modulus, connection stiffness, and damage factor as state variables and displacement and strain as observation variables.

[0030] Furthermore, components are modeled as nodes, while mechanical transfer efficiency, load redistribution, and process coupling are modeled as edges. Risk diffusion paths are identified through changes in edge weights. This allows for precise location of risk origins, diffusion chains, and key control components, upgrading from single-point early warning to global propagation analysis. It enables early identification of secondary risks, providing accurate evidence for proactive prevention and control, and enhancing the foresight and effectiveness of risk control.

[0031] Furthermore, the strategy covers monitoring optimization, construction adjustments, structural reinforcement, and maintenance prioritization, precisely matching risk sources, propagation paths, and key components. Based on the identification results, it automatically outputs executable solutions, achieving a closed loop of early warning, analysis, and response, improving risk response efficiency, reducing the probability of safety accidents, optimizing construction and operation and maintenance costs, and adapting to safety management across all scenarios of building construction and service.

[0032] It is understood that the beneficial effects of the second to sixth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here.

[0033] In summary, this invention effectively solves the problem of reliable data completion when monitoring data is missing by constructing an integrated closed loop of AI generation, physical verification, feedback correction, twin update, and risk control. It significantly improves the accuracy and reliability of the digital twin model, realizes the connection from the data layer to the decision-making layer, and provides proactive, intelligent, and closed-loop risk management capabilities for the safety of building structures throughout their entire life cycle.

[0034] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0035] Figure 1 This is a flowchart of the method of the present invention; Figure 2 Flowchart for generating virtual monitoring candidate data and verifying physical consistency; Figure 3 A flowchart illustrating the process of jointly driving digital twin updates using point cloud, BIM, and enhanced monitoring data; Figure 4 Output flowcharts for risk identification, risk propagation analysis, and control strategies; Figure 5 A schematic diagram of a computer device provided in an embodiment of the present invention; Figure 6 This is a block diagram of a chip provided according to an embodiment of the present invention.

[0036] Among them, 60. Computer equipment; 61. Processor; 62. Memory; 63. Computer program; 600. Electronic device; 610. Processing unit; 620. Storage unit; 6201. Random access memory unit; 6202. Cache memory unit; 6203. Read-only memory unit; 6204. Program / utility; 6205. Program module; 630. Bus; 640. Display unit; 650. Input / output interface; 660. Network adapter; 700. External device. Detailed Implementation

[0037] 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, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0038] In the description of this invention, it should be understood that the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0039] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0040] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes such combinations. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Additionally, the character " / " in this invention generally indicates that the preceding and following objects have an "or" relationship.

[0041] It should be understood that although terms such as first, second, third, etc., may be used in the embodiments of the present invention to describe the preset range, these preset ranges should not be limited to these terms. These terms are only used to distinguish the preset ranges from one another. For example, without departing from the scope of the embodiments of the present invention, the first preset range may also be referred to as the second preset range, and similarly, the second preset range may also be referred to as the first preset range.

[0042] Depending on the context, the word "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0043] The accompanying drawings illustrate various structural schematic diagrams according to embodiments disclosed in this invention. These drawings are not to scale, and some details have been enlarged for clarity, and some details may have been omitted. The shapes of the various regions and layers shown in the drawings, as well as their relative sizes and positional relationships, are merely exemplary and may deviate from reality due to manufacturing tolerances or technical limitations. Furthermore, those skilled in the art can design regions / layers with different shapes, sizes, and relative positions as needed.

[0044] This invention provides a method for virtual monitoring data completion and digital twin risk control of building structures. It acquires multimodal measured monitoring data, structural point cloud data, BIM data, and operating parameters of the building structure during construction or service. The method performs time synchronization, anomaly detection, and reliability assessment on the multimodal measured monitoring data to identify missing, abnormal, or low-reliability sections. Virtual monitoring candidate data is generated based on historical monitoring sequences, component topological relationships, spatial correlation characteristics of adjacent sensors, load-response nonlinear mapping relationships, operating parameters, and environmental parameters. A finite element model is established or invoked based on the BIM data, and the virtual monitoring candidate data is used as external displacement boundary conditions and equivalent nodal loads. The load or state correction input is applied to the finite element model to extract the calculated stress response, calculated displacement response, deformation compatibility response, and safety margin index of the target component and adjacent related components. Reliable virtual monitoring data is screened based on the physical consistency cost function, and physical consistency errors are fed back to the virtual monitoring candidate data generation process for correction. The reliable virtual monitoring data is fused with measured monitoring data to form an enhanced monitoring dataset. The digital twin model of the building structure is dynamically updated based on the enhanced monitoring dataset, structural point cloud data, and BIM data. Based on the updated digital twin model, risk states, risk propagation paths, and key control components are identified, and risk warning results and control strategies are output. This invention can improve the integrity of structural state perception, the accuracy of digital twin updates, and the reliability of risk control under conditions of missing monitoring data, sensor failure, or complex local operating conditions. It is applicable to the safety management of the entire construction and service process of buildings, bridges, underground structures, and prefabricated structures.

[0045] Please see Figure 1 This invention discloses a method for virtual monitoring data completion and digital twin risk control of building structures, comprising the following steps: S1. Acquire multimodal measured monitoring data, structural point cloud data, BIM data, and working condition parameter data of building structures during the construction or service period; S2. Perform time synchronization, anomaly detection, and credibility assessment on multimodal measured monitoring data to identify missing, abnormal, or low-credibility segments. Perform unified timestamp alignment on different types of monitoring data; identify abnormal monitoring segments using statistical thresholding, time-series abrupt change detection, or cross-modal consistency detection; calculate the credibility score of each monitoring segment based on the energy entropy, cross-modal mutual information, data continuity, signal quality indicators, cross-sensor consistency, and structural mechanics rationality of the monitoring sequence; and determine monitoring segments with credibility scores below a preset threshold as low-credibility segments.

[0046] S3. For the missing, abnormal, or low-confidence sections, generate virtual monitoring candidate data based on historical monitoring sequences, component topology relationships, spatial correlation characteristics of adjacent sensors, load-response nonlinear mapping relationships, operating parameters, and environmental parameters. Virtual monitoring candidate data is generated through generative adversarial networks, diffusion probability models, or a cascade of both. The generation process is a conditional generation process, using historical monitoring sequence characteristics, component categories, structural spatial locations, spatial correlation characteristics of adjacent sensors, load condition parameters, and environmental parameters as conditional inputs.

[0047] S4. Based on the BIM data, establish or call the finite element model corresponding to the target component, apply the virtual monitoring candidate data as external displacement boundary conditions, equivalent nodal loads or state correction inputs to the finite element model, perform static analysis, dynamic time history analysis or modal analysis, and extract the calculated stress response, calculated displacement response, deformation compatibility response and safety margin index of the target component and adjacent related components. Physical consistency verification includes: extracting geometric information, boundary conditions, and material properties of the target component and adjacent related components from BIM data, and establishing or calling the corresponding finite element model; applying virtual monitoring candidate data as external displacement boundary conditions, equivalent nodal loads, or state correction inputs to the finite element model; extracting the calculated stress response and calculated displacement response of the target component and adjacent related components through static analysis, dynamic time history analysis, or modal analysis, and solving the deformation compatibility error based on the relative displacement of adjacent nodes; setting the physical consistency cost function as the weighted sum of the root mean square error and deformation compatibility error between the virtual monitoring candidate data and the calculated response of the finite element model according to preset weights; when the value of the physical consistency cost function is less than the preset cost threshold, and the component safety margin index output by the finite element model is greater than the preset lower limit, it is determined that the virtual monitoring candidate data meets the physical consistency requirements.

[0048] S5. Perform physical consistency verification between the virtual monitoring candidate data and the calculation results of the finite element model. Determine whether the virtual monitoring candidate data meets the preset consistency criteria based on the physical consistency cost function. Retain the virtual monitoring candidate data that meets the preset consistency criteria as reliable virtual monitoring data, and remove the virtual monitoring candidate data that does not meet the preset consistency criteria. When the virtual monitoring candidate data does not meet the preset consistency criteria, the error generated by the physical consistency check is fed back to the virtual monitoring candidate data generation process. As physical constraint correction information, the generation process is adjusted and the virtual monitoring candidate data is regenerated until credible virtual monitoring data that meets the preset consistency criteria is obtained or the preset number of iterations is reached. The error generated by the physical consistency check is constructed as a physical constraint penalty term, which is used to update the parameters, condition inputs or sampling distribution of the virtual monitoring candidate data generation model.

[0049] S6. The trusted virtual monitoring data is fused with the multimodal measured monitoring data of the corresponding time period to form an enhanced monitoring dataset; The enhanced monitoring dataset is formed by weighting data by reliability, partitioning by component type, recursively applying data by time window, or filtering by physical consistency score and then merging the datasets.

[0050] S7. Based on the enhanced monitoring dataset, structural point cloud data and BIM data, dynamically update the component geometric parameters, material performance parameters, connection parameters and damage state parameters in the digital twin model of the building structure to obtain the current digital twin state. The structural point cloud data and BIM data are spatially registered at the component level to obtain the geometric deviation between the actual component and the designed component. The stress state, material degradation state, and connection stiffness state of the component are corrected according to the enhanced monitoring dataset. The elastic modulus, connection stiffness, and damage factor in the digital twin model are used as the state variables to be identified, and the displacement and strain in the enhanced monitoring dataset are used as the observation variables. The state variables to be identified are recursively updated using unscented Kalman filtering, extended Kalman filtering, recursive least squares method, or Bayesian model correction method. A digital twin state snapshot corresponding to the current moment is output.

[0051] S8. Calculate the risk status value of each component based on the digital twin status, and identify the risk propagation path and key control components by combining the component connection relationship, force transmission relationship or construction procedure relationship; The building structure components and their connections are modeled as a graph network, where nodes represent structural components and edges represent mechanical transfer efficiency, load redistribution coefficients, or construction process coupling relationships between components. Using the updated digital twin model parameters and the enhanced monitoring dataset as input, the risk state value of each component is calculated. Risk propagation paths are identified based on the changes in edge weights in the graph network. Key control components are determined based on the risk state values ​​and risk propagation paths.

[0052] S9. Output risk warning results based on the key control components and risk propagation paths, and generate corresponding control strategies.

[0053] Control strategies include adjusting sensor sampling frequency, adjusting monitoring point density, adjusting construction sequence, adjusting hoisting path or operating radius, implementing local reinforcement measures, or developing one or more of the following maintenance priority lists.

[0054] In another embodiment of the present invention, a building structure virtual monitoring data completion and digital twin risk control system is provided. This system can be used to implement the above-mentioned building structure virtual monitoring data completion and digital twin risk control method. Specifically, the building structure virtual monitoring data completion and digital twin risk control system includes a data acquisition module, a credibility assessment module, a virtual data generation module, a physical consistency verification module, a feedback correction module, a data fusion module, a digital twin update module, a risk analysis module, and a control output module.

[0055] The data acquisition module is used to acquire multimodal measured monitoring data, structural point cloud data, BIM data, and working condition parameter data of the building structure. The credibility assessment module is used to perform time synchronization, anomaly detection, and credibility assessment on the multimodal measured monitoring data, and to identify missing segments, abnormal segments, or low-credibility segments. The virtual data generation module is used to generate virtual monitoring candidate data; The physical consistency verification module is used to establish or call finite element models, extract response indicators and complete physical consistency verification, and filter to obtain reliable virtual monitoring data. The feedback correction module is used to feed back physical consistency errors to the virtual data generation module and adjust the virtual monitoring candidate data generation process. The data fusion module is used to fuse trusted virtual monitoring data with multimodal measured monitoring data to form an enhanced monitoring dataset; The digital twin update module is used to dynamically update the digital twin model of the building structure to obtain the current state of the digital twin. The risk analysis module is used to calculate the risk status value of components, identify risk propagation paths, and key control components; The control output module is used to output risk warning results and control strategies.

[0056] This invention provides a terminal device comprising a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, graphics processing units (GPUs), tensor processing units (TPUs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to achieve a corresponding method flow or function. The processor described in this embodiment can be used in the operation of a method for virtual monitoring data completion and digital twin risk control of building structures, including: Acquire multimodal measured monitoring data, structural point cloud data, building information model (BIM) data, and operating condition parameter data of the building structure. The multimodal measured monitoring data includes at least one or more of strain data, displacement data, vibration data, acoustic data, and environmental data. Perform time synchronization, anomaly detection, and reliability assessment on the multimodal measured monitoring data to identify missing, abnormal, or low-reliability sections. For the missing, abnormal, or low-reliability sections, generate virtual monitoring candidate data based on historical monitoring sequences, component topological relationships, spatial correlation characteristics of adjacent sensors, load-response nonlinear mapping relationships, the operating condition parameter data, and environmental parameters. Establish or call a finite element model corresponding to the target component based on the BIM data. Apply the virtual monitoring candidate data as external displacement boundary conditions, equivalent nodal loads, or state correction inputs to the finite element model to extract the calculated stress response, calculated displacement response, deformation compatibility response, and safety margin index of the target component and adjacent related components. Compare the virtual monitoring candidate data with the extracted data based on a physical consistency cost function. The response indicators undergo physical consistency verification. Virtual monitoring candidate data that meets the preset consistency criteria are retained as reliable virtual monitoring data, while virtual monitoring candidate data that does not meet the preset consistency criteria are eliminated. The physical consistency error corresponding to the failure to meet the preset consistency criteria is fed back as physical constraint correction information to the virtual monitoring candidate data generation process. The generation process is adjusted and virtual monitoring candidate data is regenerated until reliable virtual monitoring data that meets the preset consistency criteria is obtained or a preset number of iterations is reached. The reliable virtual monitoring data is fused with the multimodal measured monitoring data of the corresponding time period to form an enhanced monitoring dataset. Based on the enhanced monitoring dataset, the structural point cloud data, and the BIM data, the component geometric parameters, material performance parameters, connection parameters, and damage state parameters in the digital twin model of the building structure are dynamically updated to obtain the digital twin state at the current moment. The risk state value of each component is calculated according to the digital twin state, and the risk propagation path and key control components are identified by combining the component connection relationship, force transmission relationship, or construction procedure relationship, and the risk warning result and control strategy are output.

[0057] Please see Figure 5 The terminal device is a computer device. In this embodiment, the computer device 60 includes a processor 61, a memory 62, and a computer program 63 stored in the memory 62 and executable on the processor 61. When executed by the processor 61, the computer program 63 implements the building structure virtual monitoring data completion and digital twin risk control method of this embodiment. To avoid repetition, these details are not elaborated here. Alternatively, when executed by the processor 61, the computer program 63 implements the functions of each model / unit in the building structure virtual monitoring data completion and digital twin risk control system of this embodiment. To avoid repetition, these details are not elaborated here.

[0058] Computer device 60 can be a desktop computer, laptop, handheld computer, cloud server, or other computing device. Computer device 60 may include, but is not limited to, a processor 61 and a memory 62. Those skilled in the art will understand that... Figure 5 This is merely an example of computer device 60 and does not constitute a limitation on computer device 60. It may include more or fewer components than shown, or combine certain components, or different components. For example, computer device may also include input / output devices, network access devices, buses, etc.

[0059] The processor 61 may be a Central Processing Unit (CPU), or other general-purpose processors, graphics processing units (GPUs), tensor processing units (TPUs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0060] The memory 62 can be an internal storage unit of the computer device 60, such as a hard disk or memory of the computer device 60. The memory 62 can also be an external storage device of the computer device 60, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. equipped on the computer device 60.

[0061] Furthermore, the memory 62 may include both internal storage units of the computer device 60 and external storage devices. The memory 62 is used to store computer programs and other programs and data required by the computer device. The memory 62 can also be used to temporarily store data that has been output or will be output.

[0062] Please see Figure 6The terminal device is an electronic device 600, which is manifested in the form of a general-purpose computing device. The components of the electronic device may include, but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different platform components (including storage unit 620 and processing unit 610), a display unit 640, etc.

[0063] The storage unit stores program code, which can be executed by the processing unit 610 to perform the steps described in the method section of this specification according to various exemplary embodiments of the present invention. For example, the processing unit 610 can perform actions such as... Figure 1 The steps are shown in the figure.

[0064] Storage unit 620 may include readable media in the form of volatile storage units, such as random access memory (RAM) 6201 and / or cache memory 6202, and may further include read-only memory (ROM) 6203.

[0065] Storage unit 620 may also include a program / utility 6204 having a set (at least one) program module 6205, such program module 6205 including but not limited to: operating system, one or more application programs, other program modules and program data, each or some combination of these examples may include an implementation of a network environment.

[0066] Bus 630 can represent one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of the multiple bus structures.

[0067] Electronic device 600 can also communicate with one or more external devices 700 (e.g., keyboard, pointing device, Bluetooth device, etc.), and with one or more devices that enable a user to interact with electronic device 600, and / or with any device that enables electronic device 600 to communicate with one or more other computing devices (e.g., router, modem). This communication can be performed via input / output interface 650. Furthermore, electronic device 600 can also communicate with one or more networks (e.g., local area network, wide area network, and / or public network, such as the Internet) via network adapter 660. Network adapter 660 can communicate with other modules of electronic device 600 via bus 630. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms.

[0068] This invention also provides a storage medium, specifically a computer-readable storage medium, which is a memory device in a terminal device for storing programs and data. It is understood that the computer-readable storage medium here can include both built-in storage media in the terminal device and extended storage media supported by the terminal device; it can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, the storage space also stores one or more instructions suitable for loading and execution by a processor, which can be one or more computer programs (including program code). More specific examples of the computer-readable storage medium include: an electrical connection with one or more wires, a portable disk, a hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical fiber, portable compact disk read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.

[0069] Computer-readable storage media also include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable storage medium can also be any readable medium other than a readable storage medium that can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium can be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, radio frequency, etc., or any suitable combination thereof.

[0070] Program code for performing the operations of this invention can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can execute entirely on the user's computing device, partially on the user's device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0071] One or more instructions stored in a computer-readable storage medium can be loaded and executed by a processor to implement the corresponding steps of the virtual monitoring data completion and digital twin risk control method for building structures in the above embodiments; one or more instructions in the computer-readable storage medium are loaded and executed by the processor in the following steps: Acquire multimodal measured monitoring data, structural point cloud data, building information model (BIM) data, and operating condition parameter data of the building structure. The multimodal measured monitoring data includes at least one or more of strain data, displacement data, vibration data, acoustic data, and environmental data. Perform time synchronization, anomaly detection, and reliability assessment on the multimodal measured monitoring data to identify missing, abnormal, or low-reliability sections. For the missing, abnormal, or low-reliability sections, generate virtual monitoring candidate data based on historical monitoring sequences, component topological relationships, spatial correlation characteristics of adjacent sensors, load-response nonlinear mapping relationships, the operating condition parameter data, and environmental parameters. Establish or call a finite element model corresponding to the target component based on the BIM data. Apply the virtual monitoring candidate data as external displacement boundary conditions, equivalent nodal loads, or state correction inputs to the finite element model to extract the calculated stress response, calculated displacement response, deformation compatibility response, and safety margin index of the target component and adjacent related components. Compare the virtual monitoring candidate data with the extracted data based on a physical consistency cost function. The response indicators undergo physical consistency verification. Virtual monitoring candidate data that meets the preset consistency criteria are retained as reliable virtual monitoring data, while virtual monitoring candidate data that does not meet the preset consistency criteria are eliminated. The physical consistency error corresponding to the failure to meet the preset consistency criteria is fed back as physical constraint correction information to the virtual monitoring candidate data generation process. The generation process is adjusted and virtual monitoring candidate data is regenerated until reliable virtual monitoring data that meets the preset consistency criteria is obtained or a preset number of iterations is reached. The reliable virtual monitoring data is fused with the multimodal measured monitoring data of the corresponding time period to form an enhanced monitoring dataset. Based on the enhanced monitoring dataset, the structural point cloud data, and the BIM data, the component geometric parameters, material performance parameters, connection parameters, and damage state parameters in the digital twin model of the building structure are dynamically updated to obtain the digital twin state at the current moment. The risk state value of each component is calculated according to the digital twin state, and the risk propagation path and key control components are identified by combining the component connection relationship, force transmission relationship, or construction procedure relationship, and the risk warning result and control strategy are output.

[0072] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0073] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0074] Example 1 This embodiment provides a method for completing virtual monitoring data of building structures and controlling risks using digital twins based on physical consistency verification.

[0075] First, multimodal sensors are deployed at key parts of the building structure to collect monitoring data such as strain, displacement, vibration, acoustic emission, and environmental parameters, while also acquiring operating parameters during the construction or service life phases. For the spatial geometry of the building structure, point cloud data is acquired through 3D laser scanning, UAV photogrammetry, or visual reconstruction, and the corresponding BIM model is then used.

[0076] Subsequently, the collected multimodal measured monitoring data were synchronized in a unified time. Anomaly detection was performed on sections exhibiting signal abrupt changes, breakpoints, prolonged flatness, significant drift, or cross-modal inconsistencies, and a reliability score was calculated for each monitoring section. When the reliability score of a monitoring section fell below a preset threshold, it was marked as a low-reliability section.

[0077] For the aforementioned missing, abnormal, or low-confidence sections, the virtual data generation module is invoked. This module generates virtual monitoring candidate data based on historical monitoring sequence characteristics, component type, component spatial location, spatial correlation characteristics of adjacent sensors, load condition parameters, and environmental parameters.

[0078] Next, the virtual monitoring candidate data is input into the physical consistency verification module. This module extracts the structural geometry and boundary constraints based on the BIM data, establishes the corresponding finite element model, and calculates the stress response, displacement response, deformation compatibility, and safety margin index under the current working conditions. The calculation results are compared with the virtual monitoring candidate data. If the displacement error, stress balance error, and deformation compatibility error of adjacent components are all less than the preset threshold, and the safety margin index meets the requirements, then the virtual monitoring candidate data is considered reliable virtual monitoring data; otherwise, it is discarded.

[0079] When the virtual monitoring candidate data does not meet the physical consistency requirements, the system further feeds back the physical consistency cost function or the corresponding residual term to the virtual data generation module to correct the generation process and regenerate the virtual monitoring candidate data.

[0080] The enhanced monitoring dataset is obtained by fusing trusted virtual monitoring data with actual monitoring data for the corresponding time period. Fusion methods can include weighted fusion based on trustworthiness, component-based fusion, or time-window-based recursive fusion.

[0081] Furthermore, point cloud data and BIM data are spatially registered at the component level to obtain the actual geometric deviation of the components; combined with the enhanced monitoring dataset, the component geometric parameters, material stiffness parameters, connection stiffness parameters and damage state parameters in the digital twin model are dynamically corrected, and a snapshot of the digital twin state at the current moment is output.

[0082] Finally, the risk status value of each component is calculated based on the updated digital twin model, and a risk propagation diagram is constructed based on component connection relationships, force transmission relationships, or construction sequence relationships to identify key control components and output risk warning results. Control strategies are generated based on the risk warning results, including adjusting monitoring sampling frequency, increasing monitoring points, optimizing construction sequence, adjusting hoisting paths, implementing local reinforcement measures, or establishing maintenance priorities.

[0083] Data integrity is improved by over 95%; virtual data mechanical error is less than 5%; twin model state restoration accuracy is improved by 90%; risk identification lead time is increased by 30%~60%; false alarm rate is reduced by 70%.

[0084] Example 2 Please see Figure 2 In this embodiment, we take the beam-column node monitoring scenario during the construction phase of a prefabricated building as an example.

[0085] When the beam end strain sensor experiences a communication interruption or prolonged abnormal straightness, the system first reads the historical strain sequence of adjacent beam end nodes on the same floor, the current construction load parameters, the ambient temperature, the component type label, and the spatial relationship of adjacent sensors, and uses these as input conditions to generate virtual monitoring candidate data.

[0086] When using a generative adversarial network (GAN) to generate virtual monitoring candidate data, the GAN includes a generator and a discriminator. The specific process is as follows: 1) Feature encoding: The features of historical monitoring sequence, component category, structural spatial location, spatial correlation features of adjacent sensors, load condition parameters and environmental parameters are concatenated and input into a conditional encoder based on a long short-term memory network or Transformer architecture to extract spatiotemporal joint feature vectors. 2) Data generation: The generator receives random noise and the spatiotemporal joint feature vector, and outputs virtual monitoring candidate data with the same length as the missing time series through a multi-layer one-dimensional convolutional neural network and a deconvolutional layer; 3) Discriminant training: The discriminator uses the measured normal monitoring sequence as positive samples and the virtual monitoring candidate data as negative samples, and combines the component location labels to carry out conditional adversarial training; 4) Loss construction: In addition to adversarial loss, the loss function also includes gradient penalty term and temporal smoothing loss term to ensure that the generated temporal sequence meets the structural response continuity requirement on the changing curvature.

[0087] When generating virtual monitoring candidate data using a diffusion probability model, noise is first gradually added to the normal monitoring sequence to obtain forward diffusion samples. Then, the virtual monitoring candidate data is obtained by gradually inverting the data under conditional feature constraints by a denoising network. The conditional inputs of the denoising network also include component type, spatial location, operating parameters, and environmental parameters.

[0088] When using a cascaded generative adversarial network and a diffusion probability model for generation, initial virtual monitoring candidate data is first obtained through the generative adversarial network, and then the initial virtual monitoring candidate data is refined through the diffusion probability model to enhance local details and nonlinear dynamic response features.

[0089] The error between the virtual strain sequence and the actual response is less than 3%; the nonlinear characteristic restoration degree is greater than 92%; and the nodal mechanical equilibrium and deformation coordination are satisfied.

[0090] Example 3 When inputting the generated virtual monitoring candidate data into the finite element analysis module, different driving methods can be used depending on the data type: When the virtual monitoring candidate data is a displacement sequence, it is used as the forced displacement boundary condition for the target node; When the virtual monitoring candidate data is a load response sequence, it is converted into an equivalent nodal load applied to the finite element model. When the virtual monitoring candidate data is a state correction value, it is used as an external input for updating the model parameters.

[0091] After completing static analysis, dynamic time history analysis, or modal analysis, the calculated displacement response, calculated stress response, and relative displacement of adjacent nodes of the target component and its adjacent related components are extracted, and a physical consistency cost function is constructed:

[0092] in, The displacement response corresponding to the virtual monitoring candidate data. To calculate the displacement response using the finite element method, The stress response is obtained by mapping virtual monitoring candidate data. For finite element method (FEM) calculation of stress response, The deformation coordination error between adjacent nodes , , Preset weights.

[0093] When the physical consistency cost function If the cost is less than the preset cost threshold and the safety margin index output by the finite element model is greater than the preset lower limit, the virtual monitoring candidate data is determined to meet the physical consistency requirements.

[0094] In one specific implementation, the preset cost threshold can correspond to the comprehensive requirements of displacement error of 3% to 5%, stress balance error of 3% to 10%, and coordinated deformation error of adjacent nodes within 5%, and the lower limit of the safety margin index can be preset according to the target structure type.

[0095] When the above requirements are not met, The corresponding residual term is fed back to the virtual monitoring candidate data generation module as a physical constraint penalty term. This module updates the generation model parameters, adjusts the conditional input weights, or corrects the sampling distribution, and regenerates the virtual monitoring candidate data, forming a closed-loop process of generation, verification, correction, and regeneration.

[0096] In this embodiment, if the displacement error between a candidate strain sequence and the finite element calculation result is greater than the first threshold, or the internal force transmission deviation of adjacent nodes is greater than the second threshold, or the deformation coordination error of adjacent nodes is greater than the third threshold, then the candidate sequence is eliminated and a feedback correction process is triggered.

[0097] Under the conditions of preset structural model, load conditions and sensor deployment, the virtual monitoring candidate data can tend to meet the preset physical consistency criteria after 3 to 5 feedback corrections; after physical consistency verification and feedback correction, the proportion of virtual monitoring candidate data that passes the preset consistency criteria can be increased to 80% to 95%; finally, all the credible virtual monitoring data that enter the enhanced monitoring dataset meet the preset physical consistency criteria.

[0098] Example 4 Please see Figure 3 In this embodiment, the floor structure of a high-rise building is taken as the object.

[0099] After the concrete pouring of the floor is completed, the floor point cloud data is obtained by 3D laser scanning. The data is then aligned with the BIM model through component-level registration to obtain the actual geometric deviations of components such as beams, columns, and floor slabs.

[0100] The system will enhance the monitoring dataset input status update module and correct the component parameters in the digital twin model by combining the geometric deviation obtained from the point cloud.

[0101] When recursively updating the parameters of the digital twin model based on the geometric deviation and state correction results, the elastic modulus, connection stiffness and damage factor in the digital twin model are used as the state variables to be identified, and the displacement, strain and vibration response in the enhanced monitoring dataset are used as the observation variables. The state equation and observation equation are constructed through the finite element model.

[0102] In each time step update iteration, the model response at the current time is first predicted based on the state variables at the previous time step. Then, the residual between the measured observation and the model prediction is calculated. The state variables are then recursively corrected using unscented Kalman filtering, extended Kalman filtering, recursive least squares method, or Bayesian model correction method until the residual covariance converges or the preset iteration condition is met.

[0103] When point cloud data indicates that there is a spatial deviation in the component, the spatial deviation is mapped to the node coordinate correction amount of the digital twin model, and the node coordinates or boundary conditions of the corresponding finite element sub-model are corrected simultaneously, thereby realizing the integration of geometric update and mechanical update.

[0104] The updated digital twin model can more accurately reflect the true state of the current floor structure and provide a basis for optimizing subsequent construction procedures and identifying risks.

[0105] Geometric deviation correction accuracy reaches the millimeter level; material parameter identification error is less than 4%; connection stiffness correction error is less than 6%.

[0106] Example 5 Please see Figure 4 In this embodiment, the system abstracts building structural components into graph network nodes based on the updated digital twin model. The node attributes include component stiffness, damage factor, risk state value, and current working condition response. The edges in the graph network represent the mechanical transfer efficiency, load redistribution coefficient, or construction process coupling strength between adjacent components.

[0107] When a component is damaged or its risk status value increases, the system calculates the load redistribution result after unloading or stiffness degradation based on the digital twin model, and updates the edge weights of the edges connected to adjacent components accordingly. Then, it identifies potential risk diffusion links through multi-path search or graph propagation algorithms to determine the adjacent critical components most likely to experience secondary overruns.

[0108] For the critical components, the system outputs control strategies, including increasing the sampling frequency in the area, adding temporary monitoring points, adjusting the hoisting sequence, reducing the intensity of synchronous operations on adjacent work surfaces, and performing local reinforcement.

[0109] After the control strategy is executed, the system continues to collect a new round of monitoring data and repeatedly performs virtual completion, physical consistency verification, digital twin update and risk identification processes to form a closed-loop control.

[0110] Example 6 In this embodiment, the building structure is extended to a bridge structure or an underground structure.

[0111] For bridge structures, monitoring data may include cable force, deflection, acceleration, temperature, and acoustic emission data from cracks; during physical consistency verification, the finite element model may further consider the effects of vehicle load, wind load, and temperature gradient; risk propagation paths may be determined based on the force transmission relationships between the bridge mid-span, supports, and connecting components.

[0112] For underground structures, monitoring data may include convergence deformation, lining strain, groundwater level, vibration, and ambient humidity data; during physical consistency verification, the finite element model may further consider the surrounding rock pressure and groundwater effects; the risk propagation path may be determined based on the circumferential connection, longitudinal connection, and support system relationship of the lining.

[0113] Through the above extensions, the present invention can be applied to risk control scenarios during the construction and service periods of different types of engineering structures.

[0114] Cross-scene adaptation accuracy is greater than 90%; state assessment error is controlled within 5%.

[0115] By integrating virtual monitoring candidate data generation, physical consistency verification, physical residual feedback correction, enhanced monitoring data fusion, digital twin dynamic updates, and risk propagation identification and control output, a complete risk control chain is formed for the construction and service life of building structures. This can improve the reliability and engineering applicability of structural safety management under complex working conditions and incomplete data.

[0116] In summary, this invention provides a method for virtual monitoring data completion and digital twin risk control of building structures. Under conditions of missing, abnormal, or low-reliability monitoring data, it restores structural state information through a virtual monitoring data generation mechanism, improving the integrity of the monitoring sequence. It constrains virtual monitoring data through finite element physical consistency verification, preventing purely data-driven completion results from deviating from engineering reality. It corrects the virtual monitoring candidate data generation process through physical consistency error feedback, forming a closed loop of generation-verification-correction-regeneration, thus improving the reliability of virtual monitoring data. It enhances the accuracy of the twin model in reflecting geometric deviations, material degradation, and damage evolution by jointly driving the dynamic update of the digital twin using point cloud, BIM, and enhanced monitoring data. Through risk identification and risk propagation analysis based on the updated digital twin model, it achieves a closed-loop process from data completion to risk control. This method is applicable to various engineering scenarios such as buildings, bridges, underground structures, and prefabricated structures, exhibiting good engineering applicability and scalability.

[0117] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0118] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0119] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0120] In the embodiments provided by this invention, it should be understood that the disclosed devices / terminals and methods can be implemented in other ways. For example, the device / terminal embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0121] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0122] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0123] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random-access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0124] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus, and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0125] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0126] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0127] The above content is only for illustrating the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solution based on the technical concept proposed in this invention shall fall within the scope of protection of the claims of this invention.

Claims

1. A method for virtual monitoring data completion and digital twin risk control of building structures, characterized in that, Includes the following steps: S1. Acquire multimodal measured monitoring data, structural point cloud data, building information model (BIM) data, and working condition parameter data of the building structure, wherein the multimodal measured monitoring data includes at least one or more of strain data, displacement data, vibration data, acoustic data, and environmental data; S2. Perform time synchronization, anomaly detection, and credibility assessment on the multimodal measured monitoring data to identify missing segments, abnormal segments, or low-credibility segments. S3. For the missing, abnormal, or low-confidence sections, generate virtual monitoring candidate data based on historical monitoring sequences, component topology relationships, spatial correlation characteristics of adjacent sensors, load-response nonlinear mapping relationships, the operating condition parameter data, and environmental parameters. S4. Based on the BIM data, establish or call the finite element model corresponding to the target component, apply the virtual monitoring candidate data as external displacement boundary conditions, equivalent nodal loads or state correction inputs to the finite element model, and extract the calculated stress response, calculated displacement response, deformation compatibility response and safety margin index of the target component and adjacent related components. S5. Based on the physical consistency cost function, perform physical consistency verification on the virtual monitoring candidate data and the extracted response index. Retain virtual monitoring candidate data that meet the preset consistency criteria as credible virtual monitoring data, eliminate virtual monitoring candidate data that do not meet the preset consistency criteria, and feed back the physical consistency error corresponding to the non-compliance of the preset consistency criteria as physical constraint correction information to the virtual monitoring candidate data generation process. After adjusting the generation process, regenerate the virtual monitoring candidate data until credible virtual monitoring data that meets the preset consistency criteria is obtained or the preset number of iterations is reached. S6. The trusted virtual monitoring data is fused with the multimodal measured monitoring data of the corresponding time period to form an enhanced monitoring dataset; S7. Based on the enhanced monitoring dataset, the structural point cloud data, and the BIM data, dynamically update the component geometric parameters, material performance parameters, connection parameters, and damage state parameters in the digital twin model of the building structure to obtain the current digital twin state. S8. Calculate the risk status value of each component based on the digital twin status, and identify the risk propagation path and key control components by combining the component connection relationship, force transmission relationship or construction procedure relationship, and output the risk warning result and control strategy.

2. The method for virtual monitoring data completion and digital twin risk control of building structures according to claim 1, characterized in that, Step S2 specifically includes: Perform unified timestamp alignment on different types of monitoring data; Statistical thresholding, temporal abrupt change detection, or cross-modal consistency detection methods are used to identify abnormal monitoring sections. The reliability score of each monitoring segment is calculated based on the energy entropy, cross-modal mutual information, data continuity, signal quality index, cross-sensor consistency, and structural mechanical rationality of the monitoring sequence. Monitoring segments with a credibility score below a preset threshold are classified as low-credibility segments.

3. The method for virtual monitoring data completion and digital twin risk control of building structures according to claim 1, characterized in that, In step S3, the virtual monitoring candidate data is generated in any of the following ways: (1) Generative adversarial network-based generation; (2) Generated based on diffusion probability model; (3) Generative probabilistic networks and diffusion probability models are cascaded together; Furthermore, the generation process is a conditional generation process, using historical monitoring sequence characteristics, component categories, structural spatial locations, spatial correlation characteristics of adjacent sensors, load condition parameters, and environmental parameters as conditional inputs.

4. The method for virtual monitoring data completion and digital twin risk control of building structures according to claim 1, characterized in that, In step S5, the physical consistency cost function for: in, The displacement response corresponding to the virtual monitoring candidate data. To calculate the displacement response using the finite element method, The stress response is obtained by mapping virtual monitoring candidate data. For finite element method (FEM) calculation of stress response, The deformation coordination error between adjacent nodes , , Preset weights.

5. The method for virtual monitoring data completion and digital twin risk control of building structures according to claim 4, characterized in that, The preset consistency criterion is: the value of the physical consistency cost function is less than the preset cost threshold, and the component safety margin index output by the finite element model is greater than the preset lower limit; Physical consistency error is constructed as a physical constraint penalty term and used to update the parameters, conditional inputs, or sampling distribution of the virtual monitoring candidate data generation model.

6. The method for virtual monitoring data completion and digital twin risk control of building structures according to claim 1, characterized in that, In step S6, an enhanced monitoring dataset is formed using at least one of the following methods: 1) Weighted fusion based on data credibility; 2) Integrate components by region; 3) Perform recursive fusion based on time windows; 4) After screening based on physical consistency scores, the samples are merged.

7. The method for virtual monitoring data completion and digital twin risk control of building structures according to claim 1, characterized in that, Step S7 specifically includes: Perform component-level spatial registration between structural point cloud data and BIM data to obtain the geometric deviation between the actual component and the designed component. The stress state, material degradation state, and connection stiffness state of the components are corrected based on the enhanced monitoring dataset; The elastic modulus, connection stiffness, and damage factor in the digital twin model are used as the state variables to be identified, and the displacement and strain in the enhanced monitoring dataset are used as the observation variables. The state variables to be identified are recursively updated using unscented Kalman filtering, extended Kalman filtering, recursive least squares method, or Bayesian model correction method. Output the digital twin state snapshot corresponding to the current moment.

8. The method for virtual monitoring data completion and digital twin risk control of building structures according to claim 1, characterized in that, In step S8, the building structural components and their connections are modeled as a graph network, where nodes represent structural components and edges represent mechanical transfer efficiency, load redistribution coefficients, or construction process coupling relationships between components. Risk propagation paths are identified based on changes in edge weights in the graph network.

9. The method for virtual monitoring data completion and digital twin risk control of building structures according to claim 1, characterized in that, In step S8, the control strategy includes one or more of the following: 1) Adjust the sensor sampling frequency; 2) Adjust the density of monitoring points; 3) Adjust the construction sequence; 4) Adjust the hoisting path or operating radius; 5) Implement local reinforcement measures; 6) Develop a maintenance priority list.

10. A virtual monitoring data completion and digital twin risk control system for building structures, characterized in that, include: The data acquisition module is used to acquire multimodal measured monitoring data, structural point cloud data, BIM data, and working condition parameter data of building structures; The credibility assessment module is used to perform time synchronization, anomaly detection, and credibility assessment on the multimodal measured monitoring data, and to identify missing segments, abnormal segments, or low-credibility segments. The virtual data generation module is used to generate virtual monitoring candidate data; The physical consistency verification module is used to establish or call finite element models, extract response indicators and complete physical consistency verification, and filter to obtain reliable virtual monitoring data. The feedback correction module is used to feed back physical consistency errors to the virtual data generation module and adjust the virtual monitoring candidate data generation process. The data fusion module is used to fuse trusted virtual monitoring data with multimodal measured monitoring data to form an enhanced monitoring dataset; The digital twin update module is used to dynamically update the digital twin model of the building structure to obtain the current state of the digital twin. The risk analysis module is used to calculate the risk status value of components, identify risk propagation paths, and key control components; The control output module is used to output risk warning results and control strategies.

Images