Earthquake-damaged structure reinforcement scheme optimization method and device based on digital twin model

Abstract

The application relates to a seismic damage structure reinforcement scheme optimization method and equipment based on a digital twin model, and belongs to the technical field of computers. The method comprises the following steps: acquiring a three-dimensional geometric model of a seismic damage structure containing component semantic information, and crack characteristic data and measured dynamic response data associated with the three-dimensional geometric model; generating a digital twin basic model based on the three-dimensional geometric model of the seismic damage structure; determining a crack characteristic vector of an analysis object of the seismic damage structure according to the crack characteristic data, and constructing a seismic damage state digital twin model based on the digital twin basic model, a mapping function and the measured dynamic response data; the mapping function represents a mapping relationship from the crack characteristic vector to a material damage factor; the seismic damage state digital twin model is virtually reinforced by activating a zero stress of a reinforcement layer unit of a reinforcement scheme, and an optimized reinforcement scheme is determined according to reinforcement performance index constraints and reinforcement implementation optimization targets, so that the objectivity, rationality and adaptability of the reinforcement scheme determination are significantly improved.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a method and device for optimizing reinforcement schemes for earthquake-damaged structures based on digital twin models. Background Technology

[0002] my country is a country prone to earthquakes, and the repair and reinforcement of building structures after an earthquake is a crucial aspect of post-disaster reconstruction. Determining a suitable reinforcement scheme for earthquake-damaged structures is a key step in post-earthquake structural repair and reinforcement. The following are the main technical approaches in determining reinforcement schemes: The first type of technical approach is the traditional technique based on manual inspection and experience. This is currently the most commonly used method in engineering practice. Technicians obtain material strength and damage data by visually inspecting cracks on-site, conducting impact tests, and using portable equipment such as rebound hammers and ultrasonic detectors, in accordance with relevant specifications. They then determine reinforcement schemes based on experience, and the determination of reinforcement schemes is constrained by subjective experience.

[0003] The second technical approach is a building operation and maintenance management platform based on BIM technology. This approach uses BIM technology to create a three-dimensional geometric model of the building and attaches information such as crack photos and sensor data collected after an earthquake to the model to achieve visualized information management and determine reinforcement schemes based on the visualized information. Its core flaw lies in its emphasis on form over substance, lacking a mechanical core. Monitoring data is merely attached to the model as tags, failing to achieve true virtual-to-real mapping, thus causing the determined reinforcement scheme to deviate from the actual situation. Summary of the Invention

[0004] In view of the above analysis, the present invention aims to provide a method and device for optimizing the reinforcement scheme of earthquake-damaged structures based on a digital twin model, so as to solve the problem of insufficient virtual-real mapping in the determination of existing reinforcement schemes and the deviation between the determination of the reinforcement scheme and the actual earthquake damage.

[0005] On one hand, embodiments of the present invention provide a method for optimizing a reinforcement scheme for a seismically damaged structure based on a digital twin model. The method includes: acquiring a three-dimensional geometric model of the seismically damaged structure containing semantic information of the components, as well as crack feature data and measured dynamic response data associated with the three-dimensional geometric model of the seismically damaged structure; generating a digital twin base model based on the three-dimensional geometric model of the seismically damaged structure; determining a crack feature vector corresponding to a pre-set seismically damaged structure analysis object based on the crack feature data; constructing a seismically damaged state digital twin model based on the digital twin base model, a mapping function corresponding to the seismically damaged structure analysis object, and the measured dynamic response data; wherein the mapping function represents the mapping relationship between the crack feature vector and the material damage factor; performing virtual reinforcement on the seismically damaged state digital twin model by activating the reinforcement layer units of the reinforcement scheme with zero stress; iteratively adjusting the reinforcement scheme according to pre-set reinforcement performance index constraints and reinforcement implementation optimization objectives until the optimization result of the reinforcement scheme is determined.

[0006] Further, the acquisition of the three-dimensional geometric model of the earthquake-damaged structure containing component semantic information, as well as crack feature data and measured dynamic response data associated with the three-dimensional geometric model of the earthquake-damaged structure, includes: acquiring three-dimensional laser scanning point cloud data, UAV point cloud data, and building surface image data of the earthquake-damaged structure, dynamic time history data collected by pre-deployed sensors, and building geometric information data; generating a fused point cloud by performing point cloud registration on the three-dimensional laser scanning point cloud data and the UAV point cloud data, and performing reverse modeling on the fused point cloud in combination with the building geometric information data to generate the three-dimensional geometric model of the earthquake-damaged structure containing component semantic information; obtaining crack feature data by performing image processing on the building surface image data; acquiring measured dynamic response data based on the dynamic time history data; and associating and storing the crack feature data and the measured dynamic response data with the three-dimensional geometric model of the earthquake-damaged structure containing component semantic information.

[0007] Furthermore, the step of generating a digital twin base model based on the three-dimensional geometric model of the earthquake-damaged structure includes: establishing a first mechanical calculation model for the critical areas of the earthquake-damaged structure using mesoscopic solid elements based on the three-dimensional geometric model of the earthquake-damaged structure and pre-determined rules for dividing critical and non-critical areas; establishing a second mechanical calculation model for the non-critical areas of the earthquake-damaged structure using macroscopic solid elements; establishing a third mechanical calculation model at the interface between the critical and non-critical areas using non-coordinated mesh interface transfer technology; and generating the digital twin base model based on the first, second, and third mechanical calculation models; wherein the macroscopic solid elements include rod elements or shell elements.

[0008] Furthermore, at the interface between the critical region and the non-critical region, a third mechanical calculation model is established using non-coordinated mesh interface transfer technology. This includes: defining the nodes of the macroscopic solid element or the interface reference node as master nodes, and defining the nodes of the mesoscopic solid element as slave nodes; calculating the coupling weight of each slave node relative to the master node based on at least one of the interface area represented by the slave node, the distance from the slave node to the master node, and the interface shape function; generating a constraint coefficient matrix based on the coupling weight of each slave node; and establishing the third mechanical calculation model by generating a displacement compatibility equation by writing the constraint coefficient matrix into the MPC equation, the Lagrange multiplier equation, or the penalty function equation; wherein, in solving the third mechanical calculation model, the interface force is transferred by the constraint reaction force.

[0009] Furthermore, after generating the digital twin base model, the method further includes: setting one or more transitional solid units between the mesoscopic solid units and the macroscopic solid units of the digital twin base model; wherein the stiffness parameter of the transitional solid unit is determined by monotonic interpolation based on the position of the transitional solid unit in the mesoscopic solid unit and the macroscopic solid unit.

[0010] Further, the construction of the damaged state digital twin model based on the digital twin basic model, the mapping function corresponding to the damaged structural analysis object, and the measured dynamic response data includes: after determining the initial values ​​of the undetermined coefficients in the mapping function corresponding to each damaged structural analysis object, iteratively executing the following process until the value of the target error function is less than or equal to a preset threshold or the maximum number of iterations is reached, and outputting the damaged state digital twin model: determining the material damage factor corresponding to each damaged structural analysis object according to the mapping function corresponding to each damaged structural analysis object and the crack feature vector; wherein, the material damage factor includes A set of material constitutive parameter reduction coefficients; based on the digital twin basic model, the original material constitutive parameters are updated according to the material damage factors corresponding to each of the earthquake-damaged structural analysis objects and written into the material constitutive model of the earthquake-damaged structural analysis object to generate an earthquake-damaged digital twin model; modal analysis is performed on the earthquake-damaged digital twin model to obtain calculated dynamic response data; the value of the target error function is calculated based on the measured dynamic response data and the calculated dynamic response data, and when the value of the target error function is greater than the preset threshold, the undetermined coefficients of the mapping function corresponding to each of the earthquake-damaged structural analysis objects are adjusted by an optimization algorithm.

[0011] Furthermore, the target error function is expressed as:

[0012] in, This represents the target error function. Indicates the undetermined coefficients. , , These are pre-set non-negative weights, used to control the proportions of frequency error, mode shape correlation error, and time history response error in the target error function, respectively. , They represent the first The calculated frequency and measured frequency of each frequency characteristic term involved in the frequency error calculation; This indicates the number of frequency characteristic terms involved in the frequency error calculation; , They represent the first The calculated mode shape vector and the measured mode shape vector corresponding to the mode shape characteristic terms involved in the mode shape correlation calculation; This indicates the number of modal characteristic terms involved in the modal correlation calculation; , They represent the first The calculated time history response and the measured time history response of each response channel involved in the time history response error calculation; Indicates the number of response channels involved in the calculation of time history response error; It is a mode shape correlation index, used to measure the degree of consistency between the calculated mode shape vector and the measured mode shape vector under the same measurement point, the same direction, and the same degree of freedom sequence. This represents the root mean square error.

[0013] Furthermore, the reinforcement layer unit includes a reinforcement material unit, a connection unit, an interface unit, or an energy dissipation unit; the zero-stress activation of the reinforcement layer unit of the reinforcement scheme includes: solving the stress of the digital twin model of the seismic damage state, and calculating the spurious strain after the stress solution converges; at the moment of activation of the reinforcement layer unit of the reinforcement scheme, the strain reference field in the material definition of the reinforcement layer unit is set to be equal to the spurious strain.

[0014] Furthermore, adjusting the reinforcement scheme includes adjusting the reinforcement method and / or reinforcement parameters; wherein, the reinforcement parameters include interface behavior parameters between new and old materials that form a contact relationship during the reinforcement or repair process.

[0015] The present invention also provides an electronic device, including 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 optimization method for strengthening earthquake-damaged structures based on digital twin models as described above.

[0016] In this invention, the above-described technical solutions can be combined with each other to achieve more preferred combinations. Other features and advantages of this invention will be set forth in the following description, and some advantages may become apparent from the description or be learned by practicing the invention. The objects and other advantages of this invention can be realized and obtained from what is particularly pointed out in the description and drawings.

[0017] The present invention provides a method and equipment for optimizing earthquake-damaged structural reinforcement schemes based on digital twin models. By establishing a digital twin model of the earthquake-damaged state, it achieves a precise virtual-real mapping from apparent geometric damage to intrinsic mechanical properties. Furthermore, based on pre-set reinforcement performance index constraints and reinforcement implementation optimization objectives, it iteratively optimizes the reinforcement scheme, significantly improving the accuracy and objectivity of earthquake damage state identification, enhancing the objectivity, rationality, and adaptability of reinforcement scheme determination, avoiding insufficient or excessive reinforcement due to reliance on experience alone, and preventing waste of reinforcement materials and unsafe reinforcement schemes. Attached Figure Description

[0018] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts. Figure 1 This is a flowchart illustrating the optimization method for strengthening earthquake-damaged structures based on digital twin models provided by the present invention. Figure 2 A schematic diagram of the physical structure of an electronic device is provided. Detailed Implementation

[0019] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.

[0020] Figure 1 This is a flowchart illustrating the optimization method for earthquake-damaged structure reinforcement based on a digital twin model provided by the present invention. Figure 1 As shown, the method includes: Step S1: Obtain the three-dimensional geometric model of the earthquake-damaged structure containing semantic information of the components, as well as the crack feature data and measured dynamic response data associated with the three-dimensional geometric model of the earthquake-damaged structure.

[0021] This invention optimizes reinforcement schemes for earthquake-damaged structures based on digital twin technology and finite element / numerical mechanics analysis. To construct the digital twin model, it is first necessary to acquire a three-dimensional geometric model of the earthquake-damaged structure containing semantic information of its components, as well as crack feature data and measured dynamic response data associated with this model. Component semantic information may include component ID, component type (beam, column, wall, slab), floor level, spatial location, cross-sectional dimensions, material type, connection relationships, boundary constraint relationships, whether it is a critical component, and associated crack / sensor data, etc.

[0022] The three-dimensional geometric model of the earthquake-damaged structure, which includes semantic information about the components, can be determined based on multi-source data acquisition and fusion, and stored in association with crack feature data and measured dynamic response data. Crack feature data can be obtained through image acquisition and processing, while measured dynamic response data can be acquired through sensors.

[0023] Step S2: Generate a digital twin basic model based on the three-dimensional geometric model of the earthquake-damaged structure.

[0024] Based on digital twin technology, a basic digital twin model capable of mechanical calculation is generated from a three-dimensional geometric model of a seismically damaged structure containing semantic information of its components.

[0025] Step S3: Determine the crack feature vector corresponding to the pre-set seismic damage structure analysis object based on the crack feature data, and construct a seismic damage state digital twin model based on the digital twin basic model, the mapping function corresponding to the seismic damage structure analysis object, and the measured dynamic response data; wherein, the mapping function represents the mapping relationship between the crack feature vector and the material damage factor.

[0026] The digital twin basic model is generated based on a three-dimensional geometric model of the damaged structure containing semantic information of the components. However, it has not yet been associated with crack feature data. The crack feature data related to the damage and the measured dynamic response data fail to effectively drive the dynamic update of the material constitutive parameters in the virtual model. To achieve this, the crack feature data related to the damage and the measured dynamic response data can effectively drive the dynamic update of the material constitutive parameters in the virtual model. This invention determines the crack feature vector corresponding to the pre-set damaged structure analysis object based on the crack feature data, and constructs a damaged-state digital twin model based on the digital twin basic model, the mapping function corresponding to the damaged structure analysis object, and the measured dynamic response data.

[0027] The objects of seismic damage structural analysis can be components, damaged regions, elements, integration points, etc. An element is the basic computational domain formed after discretizing the structure, such as a small solid block, a shell plate, or a beam segment. An integration point is the location within an element used for numerical integration. Finite element programs typically calculate and store stress, strain, damage variables, and constitutive states at integration points.

[0028] Based on crack characteristic data, a crack feature vector corresponding to a pre-defined seismically damaged structural analysis object is determined. The crack feature vector includes at least the average crack width and crack areal density, and may also include crack type and crack connectivity. The average crack width measures the "depth" or severity of the damage. Crack areal density is typically defined as the total crack length per unit surface area, used to measure the "breadth" or distribution range of the damage. Crack types, such as shear cracks and bending cracks, are used to achieve anisotropic damage assessment. Crack connectivity can be the extension length, penetration degree, and interconnection relationship of single or multiple cracks, used to assess the extent of damage penetration.

[0029] When constructing a damaged-state digital twin model based on a digital twin foundation model, a mapping function corresponding to the damaged structural analysis object, and measured dynamic response data, the damaged-state digital twin model can be determined through multiple iterative optimizations. The goal is to ensure that the error between the calculated dynamic response data obtained from the damaged-state digital twin model and the measured dynamic response data meets set requirements. Determining the damaged-state digital twin model includes determining the mapping function corresponding to the damaged structural analysis object, calculating the material damage factor based on the mapping function and crack feature vector, and updating the material constitutive parameters based on the material damage factor.

[0030] Step S4: Virtual reinforcement is implemented on the digital twin model of the seismic damage state by activating the reinforcement layer unit of the reinforcement scheme with zero stress. The reinforcement scheme is iteratively adjusted according to the pre-set reinforcement performance index constraints and reinforcement implementation optimization objectives until the optimization result of the reinforcement scheme is determined.

[0031] The reinforcement scheme is a proposed method for reinforcing the earthquake-damaged structure, which may include reinforcement methods and parameters. Reinforcement methods may include increasing the cross-section, bonding CFRP or steel plates, external steel cladding, and adding energy-dissipating supports. Reinforcement parameters may include the type of reinforcement material, geometric dimensions, bonding range, interface performance parameters, and anchoring measures. The reinforcement layer units are related to the reinforcement scheme and may include reinforcement material units, connection units, interface units, or energy-dissipating units. Virtual reinforcement is implemented on the earthquake-damaged digital twin model by activating the reinforcement layer units of the reinforcement scheme with zero stress, simulating the reinforcement of the actual earthquake-damaged structure and ensuring that the internal stress of the reinforcement layer units is zero at the moment of activation. Subsequently, under continued applied loads (such as seismic action), the reinforcement layer units will only bear incremental deformation, realistically simulating the physical process of "secondary stress."

[0032] The reinforcement scheme is iteratively adjusted based on pre-set reinforcement performance constraints and reinforcement implementation optimization objectives until the optimized result is determined. Reinforcement performance constraints are the conditions that the reinforcement performance indicators must meet, such as constraints on maximum inter-story drift angle, residual deformation, component damage index, and interface delamination state. Reinforcement implementation optimization objectives include minimizing reinforcement material usage, minimizing engineering cost, and maximizing performance margin. Constrained multi-objective optimization can be performed by automatically adjusting the reinforcement scheme using optimization algorithms (such as genetic algorithms and particle swarm optimization) to ultimately determine the optimized result of the reinforcement scheme.

[0033] The present invention provides an optimization method for strengthening earthquake-damaged structures based on a digital twin model. By establishing a digital twin model of the earthquake-damaged state, it achieves a precise virtual-real mapping from apparent geometric damage to intrinsic mechanical properties. Furthermore, based on pre-set strengthening performance index constraints and strengthening implementation optimization objectives, it iteratively optimizes the strengthening scheme, significantly improving the accuracy and objectivity of earthquake damage state identification, and significantly improving the objectivity, rationality, and adaptability of strengthening scheme determination. It avoids insufficient or excessive strengthening caused by relying solely on experience, and prevents waste of strengthening materials and unsafe strengthening schemes.

[0034] According to the present invention, an optimization method for strengthening earthquake-damaged structures based on a digital twin model is provided. The step of acquiring a three-dimensional geometric model of the earthquake-damaged structure containing semantic information of its components, as well as crack feature data and measured dynamic response data associated with the three-dimensional geometric model of the earthquake-damaged structure, includes: acquiring three-dimensional laser scanning point cloud data, UAV point cloud data, and building surface image data of the earthquake-damaged structure; acquiring dynamic time history data and building geometric information data collected by pre-deployed sensors; generating a fused point cloud by point cloud registration of the three-dimensional laser scanning point cloud data and the UAV point cloud data; generating the three-dimensional geometric model of the earthquake-damaged structure containing semantic information of its components by reverse modeling the fused point cloud in conjunction with the building geometric information data; obtaining crack feature data by image processing of the building surface image data; acquiring measured dynamic response data based on the dynamic time history data; and associating and storing the crack feature data and the measured dynamic response data with the three-dimensional geometric model of the earthquake-damaged structure containing semantic information of its components.

[0035] Physical characterization data of earthquake-damaged structures includes, but is not limited to: 3D laser scanning point cloud data, UAV point cloud data, UAV-captured building surface images, dynamic time-history data collected by sensors deployed at predetermined key locations on the earthquake-damaged structure under earthquake or aftershock conditions, and building geometric information data. Dynamic time-history data is a record of dynamic response that varies over time, such as time series of acceleration, velocity, displacement, and strain. Building geometric information data includes Building Information Modeling (BIM) or 2D design drawings, serving as a reference for geometric and component information.

[0036] By employing techniques such as the Iterative Closest Point (ICP) algorithm, point cloud registration is performed on 3D laser scanning point cloud data and UAV point cloud data to generate a fused point cloud, achieving coordinate unification. To improve the robustness of point cloud registration in earthquake-damaged and spalled areas, higher weights can be assigned to geometric feature points such as beam and column edge lines extracted from the original BIM model during the point cloud configuration process. Point cloud registration can utilize building geometric information data as geometric prior information.

[0037] By combining building geometric information data with fused point cloud data, a 3D geometric model of the earthquake-damaged structure in a unified coordinate system, containing semantic information of the components, is generated through reverse modeling. Crack feature data is obtained by image processing of building surface image data, and measured dynamic response data is acquired based on dynamic time history data. The crack feature data and measured dynamic response data are then associated and stored with the 3D geometric model of the earthquake-damaged structure containing semantic information of the components. This association and storage includes linking crack features to corresponding components or units, and mapping the sensor measurement point positions of the measured dynamic response data to the model coordinate system.

[0038] Among them, dynamic time history data is a raw data form of dynamic response data; dynamic response data has a wider range, which can include dynamic time history data such as acceleration, displacement, and strain, as well as characteristic quantities such as natural frequency, damping ratio, mode shape, and peak response identified from dynamic time history data.

[0039] The overall processing of acquiring a 3D geometric model of a seismically damaged structure containing semantic information of its components, along with associated crack feature data and measured dynamic response data, can include: point cloud spatial registration → component segmentation → geometric fitting → component ID assignment → crack identification and crack feature extraction → linking crack features to corresponding components or elements → mapping sensor measurement point positions to the model coordinate system. The reverse modeling process includes steps such as component segmentation, geometric fitting, and component semantic assignment, where component ID assignment is used to establish the data index relationship between crack features, sensor measurement points, and elements.

[0040] The multi-scale model constructed in this invention can be solved efficiently in the same global coordinate system, and the calculation results (such as damage, internal forces, and deformation) can be accurately fed back to the component level, providing a reliable foundation for subsequent reinforcement optimization. The implementation path is as follows: all component, node, element, and sensor positions are mapped to the same global coordinate system; each element / node is bound to a component ID; after solving, the results such as element damage, node displacement, and internal forces are summarized back to the corresponding beams, columns, walls, etc., according to the component ID. This allows for solving within a unified model while also outputting component-level results.

[0041] The present invention provides an optimization method for strengthening earthquake-damaged structures based on digital twin models, which realizes the acquisition of a three-dimensional geometric model of the earthquake-damaged structure containing semantic information of components, as well as crack feature data and measured dynamic response data associated with the three-dimensional geometric model.

[0042] According to the present invention, an optimization method for strengthening earthquake-damaged structures based on a digital twin model is provided. The step of generating a digital twin base model based on the three-dimensional geometric model of the earthquake-damaged structure includes: establishing a first mechanical calculation model for the critical areas of the earthquake-damaged structure using mesoscopic solid elements based on the three-dimensional geometric model of the earthquake-damaged structure and pre-determined rules for dividing critical and non-critical areas; establishing a second mechanical calculation model for the non-critical areas of the earthquake-damaged structure using macroscopic solid elements; establishing a third mechanical calculation model at the interface between the critical and non-critical areas using a non-coordinated mesh interface transfer technique; and generating the digital twin base model based on the first, second, and third mechanical calculation models. The macroscopic solid elements include rod elements or shell elements.

[0043] When generating a digital twin model based on a 3D geometric model of a seismically damaged structure containing semantic information of its components, using a single-scale model and employing only macroscopic solid elements for modeling results in an inability to accurately describe the damage evolution at the microscopic component level; conversely, using only microscopic solid elements for modeling leads to excessively low computational efficiency. This invention addresses the contradiction between the timeliness requirements of rapid post-earthquake assessment for overall structural analysis and the preservation of the microscopic nonlinear damage evolution mechanism of key damaged areas (such as nodes and plastic hinge zones), thus resolving the difficulty of applying a single-scale model. It constructs a "macroscopic-microscopic" coupled multi-scale digital twin model that balances computational efficiency and analytical accuracy.

[0044] Based on a 3D geometric model of the earthquake-damaged structure containing semantic information of the components and pre-defined rules for dividing key and non-key regions, a first mechanical calculation model is established for key regions of the earthquake-damaged structure (such as severely damaged nodes, the bottom of shear walls, etc.) using mesoscopic solid elements that can finely describe the local stress state. The rules for dividing key and non-key regions can be based on crack density, component type, etc. A component or region is identified as a "key region" and modeled using mesoscopic solid elements if any of the following conditions are met: the maximum or average crack width identified in the component or region exceeds a crack width threshold; the total crack length per unit area of ​​the component or region exceeds a crack density threshold; or the component belongs to a predefined key component type, such as the core area of ​​a node, the plastic hinge area at the bottom of a shear wall, a connecting beam, or a short column. Other components or regions that do not meet the above conditions are identified as "non-key regions" and modeled using macroscopic solid elements.

[0045] For non-critical areas, macroscopic solid elements, i.e., computationally efficient rod or shell elements, are used for modeling. Mesoscopic solid elements refer to continuous solid elements used to finely describe the evolution of local three-dimensional stress, strain, and damage. For example, the core area of ​​a concrete column joint can be modeled using eight-node hexahedral solid elements C3D8R or C3D8, or tetrahedral solid elements C3D4 / C3D10, etc.

[0046] At the interface between critical and non-critical regions, i.e., at the interface between macro and micro models, a third mechanical calculation model is established through non-coordinated mesh interface transfer technology. For example, through non-coordinated mesh interface transfer technology based on kinematic coupling constraints or energy conservation, the continuous transfer of force and displacement between models of different scales is ensured.

[0047] By integrating the first, second, and third mechanical calculation models, a digital twin foundation model is generated, resulting in a multi-scale digital twin foundation model capable of mechanical calculations. This digital twin foundation model includes a complete material property library, boundary conditions, and load case definitions. The material property library can be derived from BIM / design drawings, test reports, standard values, material tests, or default databases, and includes parameters such as elastic modulus, strength, density, Poisson's ratio, yield parameter, and damage parameter. Boundary conditions refer to support constraints, floor slab constraints, component connection relationships, and foundation constraints. Load cases include self-weight, dead load, live load, seismic input, aftershock time history, and loads during the reinforcement construction phase.

[0048] The present invention provides an optimization method for strengthening earthquake-damaged structures based on digital twin models. By using microscopic and macroscopic solid elements to establish mechanical calculation models for key and non-key regions respectively, and by using non-coordinated mesh interface transfer technology to establish a mechanical calculation model at the interface between the macroscopic and microscopic models, a multi-scale digital twin model is finally obtained. This solves the problem of the difficulty in reconciling computational efficiency with the simulation accuracy of microscopic damage mechanisms in key parts, and achieves high-fidelity simulation with low computing power cost.

[0049] According to the present invention, an optimization method for strengthening earthquake-damaged structures based on a digital twin model is provided. The method involves establishing a third mechanical calculation model at the interface between the critical and non-critical regions using a non-coordinated mesh interface transfer technique. This includes: defining nodes of the macroscopic solid elements or interface reference nodes as master nodes and nodes of the mesoscopic solid elements as slave nodes; calculating the coupling weight of each slave node relative to the master node based on at least one of the interface area represented by the slave node, the distance from the slave node to the master node, and the interface shape function; generating a constraint coefficient matrix based on the coupling weight of each slave node; and establishing the third mechanical calculation model by generating displacement compatibility equations by incorporating the constraint coefficient matrix into MPC equations, Lagrange multiplier equations, or penalty function equations. In solving the third mechanical calculation model, interface forces are transferred by constraint reaction forces.

[0050] At the interface between critical and non-critical regions, that is, at the interface between macroscopic solid elements (such as beam elements) and mesoscopic solid elements (such as three-dimensional continuum solid elements), mesh nodes typically do not coincide (non-coordinated meshes). This invention ensures the continuous transmission of force and displacement by generating displacement compatibility equations to construct a third mechanical calculation model. The implementation steps are as follows: Define master and slave node sets: Nodes of macroscopic solid elements or interface reference nodes are defined as master nodes, whose degrees of freedom (e.g., 3 translations, 3 rotations) control the behavior of the entire interface; all nodes of mesoscopic solid elements on the interface are defined as slave nodes. Interface reference nodes, also known as reference point nodes or auxiliary control nodes, are not solid elements and do not represent actual material volume. Instead, they are control nodes located at the center of the interface section or the endpoints of the component axis, used to centrally express the displacement, rotation, or generalized degrees of freedom of the macroscopic solid element at that interface.

[0051] Establish displacement compatibility equations: Calculate the coupling weight of each slave node relative to the master node based on the interface area represented by the slave node, the distance from the slave node to the master node, and at least one of the interface shape functions. Generate a constraint coefficient matrix based on the coupling weight of each slave node. Establish a third mechanical calculation model by writing the constraint coefficient matrix into the MPC (Multipoint Constraint) equation, Lagrange multiplier equation, or penalty function equation. In solving the third mechanical calculation model, the interface force is transmitted by the constraint reaction force.

[0052] When calculating the coupling weight of each slave node relative to the master node, the coupling weight determined by the interface area represented by the slave node can be called the area weight, and the area weight determines the contribution coefficient according to the interface area represented by the slave node; the coupling weight determined by the distance from the slave node to the master node can be called the distance weight; the coupling weight determined by the interface shape function can be called the shape function weight, and the shape function weight refers to interpolation according to the shape function of the macroscopic solid unit or interface unit.

[0053] The present invention provides an optimization method for strengthening earthquake-damaged structures based on digital twin models. By reasonably setting the coupling weights of master and slave nodes and slave nodes relative to master nodes in the generation of displacement coordination equations, displacement coordination on both sides of the interface is achieved.

[0054] According to the present invention, an optimization method for strengthening earthquake-damaged structures based on a digital twin model is provided. After generating the digital twin basic model, the method further includes: setting one or more layers of transitional solid elements between the microscopic solid elements and the macroscopic solid elements in the digital twin basic model; wherein the stiffness parameters of the transitional solid elements are determined by monotonic interpolation based on the position of the transitional solid elements in the microscopic solid elements and the macroscopic solid elements.

[0055] Although macroscopic and mesoscopic elements are geometrically adjacent, their element scales, degrees of freedom types, and stiffness expressions differ. Abrupt changes can cause local stiffness jumps and spurious stress peaks, affecting damage assessment and internal force transmission. To avoid spurious stress concentrations at the interface due to abrupt changes in model scale, this invention provides a smooth stiffness transition mechanism. One or more "transitional elements" are set between macroscopic and mesoscopic elements. The stiffness parameters of the transitional elements, such as the elastic modulus, are determined by monotonic interpolation (e.g., linear or cubic spline interpolation) based on their positions within the mesoscopic and macroscopic elements, thus achieving a smooth stiffness transition and improving the accuracy of the calculation results. After setting the transition zone formed by the transitional elements, the stiffness and deformation modes gradually transition, improving convergence and increasing the reliability of stress / damage results in key areas.

[0056] The present invention provides an optimization method for strengthening earthquake-damaged structures based on digital twin models. By setting one or more transitional solid elements between microscopic and macroscopic solid elements, and determining the stiffness parameters of the transitional solid elements by monotonic interpolation based on the distance from the transitional solid elements to the microscopic and macroscopic solid elements, a smooth transition of stiffness is achieved, thereby improving the accuracy of the calculation results.

[0057] According to the present invention, an optimization method for strengthening earthquake-damaged structures based on a digital twin model is provided. The step of constructing a earthquake-damaged digital twin model based on the digital twin basic model, the mapping function corresponding to the earthquake-damaged structural analysis object, and the measured dynamic response data includes: after determining the initial values ​​of the undetermined coefficients in the mapping function corresponding to each earthquake-damaged structural analysis object, iteratively executing the following process until the value of the target error function is less than or equal to a preset threshold or the maximum number of iterations is reached, and outputting the earthquake-damaged digital twin model: determining the material damage factor corresponding to each earthquake-damaged structural analysis object according to the mapping function corresponding to each earthquake-damaged structural analysis object and the crack feature vector; The material damage factor includes a set of material constitutive parameter reduction coefficients. Based on the digital twin basic model, the original material constitutive parameters are updated according to the material damage factor corresponding to each of the earthquake-damaged structural analysis objects and written into the material constitutive model of the earthquake-damaged structural analysis object to generate an earthquake-damaged digital twin model. Modal analysis is performed on the earthquake-damaged digital twin model to obtain calculated dynamic response data. The value of the target error function is calculated based on the measured dynamic response data and the calculated dynamic response data. When the value of the target error function is greater than the preset threshold, the undetermined coefficients of the mapping function corresponding to each of the earthquake-damaged structural analysis objects are adjusted by an optimization algorithm.

[0058] When constructing a damaged state digital twin model based on a digital twin foundation model, a mapping function corresponding to the damaged structural analysis object, and measured dynamic response data, undetermined coefficients are set in the mapping function. These undetermined coefficients are determined through iterative processes, ultimately identifying the mapping function corresponding to each damaged structural analysis object. Based on the determined mapping function and crack feature vector for each damaged structural analysis object, the material damage factor corresponding to that object is determined. The material damage factor includes a set of material constitutive parameter reduction coefficients. Based on the digital twin foundation model, the corresponding original material constitutive parameters are updated according to the material damage factor for each damaged structural analysis object and written into the material constitutive model of the damaged structural analysis object, generating the damaged state digital twin model.

[0059] Material constitutive models are models that describe the stress-strain relationship of materials and their evolutionary laws such as damage, plasticity, cracking, yielding, and softening. For example, the concrete damage-plasticity model, the steel bilinear hardening model, and the CFRP linear elastic / brittle fracture model are all material constitutive models.

[0060] The following section details the generation process of the digital twin model of the seismic damage state.

[0061] Before the first iteration, a feature vector from the crack is established. To material damage factor mapping function ,in It is a set of undetermined coefficients. The mathematical form of the mapping function can be established based on existing damage mechanics, experimental regression, or machine learning. The mapping function corresponding to the seismically damaged structural analysis object can adopt a uniform form, or different undetermined coefficients can be configured according to the component type, damage mode, or material type.

[0062] After determining the initial values ​​of the undetermined coefficients in the mapping function corresponding to each seismically damaged structural analysis object, the iterative process is executed. The processing procedure for each iteration is as follows: The material damage factor for each seismically damaged structural analysis object is determined based on its mapping function and crack feature vector. The material damage factor includes a set of material constitutive parameter reduction coefficients, each containing at least one material constitutive parameter reduction coefficient. Material constitutive parameters include elastic modulus E, compressive strength fc, tensile strength ft, and fracture energy Gf. The material damage factor is a set of degradation coefficients representing the degree of degradation of these constitutive parameters. The mapping function can be a linear function, a nonlinear function, or a trained machine learning model (such as support vector regression or neural networks).

[0063] Based on the digital twin foundation model, the original material constitutive parameters of each damaged structural analysis object (such as components, damaged areas, elements, and integration points) are updated according to the material damage factors corresponding to each damaged structural analysis object and written into the material constitutive model of the damaged structural analysis object, generating a damaged-state digital twin model. Since different damaged structural analysis objects can correspond to different mapping functions, the material damage factors corresponding to different damaged structural analysis objects, that is, the reduction coefficients of a set of material constitutive parameters, can be different. The original material constitutive parameters can be obtained from the material property library of the digital twin foundation model.

[0064] The following verifies whether the generated digital twin model of the seismic damage state meets the requirements.

[0065] Modal analysis is performed on the damaged digital twin model to obtain calculated dynamic response data. The calculated dynamic response data corresponds to the measured dynamic response data, the difference being that one is a calculated result and the other a measured result. The generated damaged digital twin model meets the following condition: the value of the objective error function used to evaluate the difference between the calculated and measured dynamic response data is less than or equal to a preset threshold; if the value of the objective error function is greater than the preset threshold, the generated damaged digital twin model does not meet the requirement. If the generated damaged digital twin model meets the requirement, or reaches the maximum number of iterations, the iteration ends. If the generated damaged digital twin model does not meet the requirement and has not reached the maximum number of iterations, the undetermined coefficients of the mapping function corresponding to each damaged structural analysis object are adjusted using optimization algorithms, such as genetic algorithms, particle swarm optimization, Bayesian optimization, or gradient-based algorithms, to adjust the mapping model. Undetermined coefficients in Then, the next iteration begins.

[0066] In addition, the material damage factor can also be defined as a damage variable in continuous damage mechanics, where 0 represents no damage and 1 represents complete failure. By incorporating the damage variable into the material constitutive model of each seismically damaged structural analysis object, a digital twin model of the seismic damage state for each iteration period is generated.

[0067] To ensure the stability of numerical calculations, this invention can optimize the calculated material damage factor. Apply boundary constraints, i.e. This effectively prevents the damage factor from exceeding 1.0 (corresponding to negative material stiffness) due to model extrapolation or improper parameters, thus avoiding numerical collapse problems during finite element analysis. An example form of the mapping function is: Where α and β are undetermined coefficients. ρc represents the average width of the crack, and ρc represents the surface density of the crack. and These are the normalized baseline values ​​for crack width and crack surface density, respectively.

[0068] The present invention provides an optimization method for strengthening earthquake-damaged structures based on digital twin models. By iteratively determining the undetermined coefficients of the mapping function, the digital twin model of the earthquake-damaged state is determined. This ensures that the final generated digital twin model of the earthquake-damaged state can reproduce the real dynamic characteristics of the earthquake-damaged structure with high fidelity, providing a highly reliable mechanical model for subsequent performance evaluation and strengthening design.

[0069] According to the present invention, an optimization method for strengthening earthquake-damaged structures based on a digital twin model is provided, wherein the objective error function is expressed as:

[0070] in, This represents the target error function. Indicates the undetermined coefficients. , , These are pre-set non-negative weights, used to control the proportions of frequency error, mode shape correlation error, and time history response error in the target error function, respectively. , They represent the first The calculated frequency and measured frequency of each frequency characteristic term involved in the frequency error calculation; This indicates the number of frequency characteristic terms involved in the frequency error calculation; , They represent the first The calculated mode shape vector and the measured mode shape vector corresponding to the mode shape characteristic terms involved in the mode shape correlation calculation; This indicates the number of modal characteristic terms involved in the modal correlation calculation; , They represent the first The calculated time history response and the measured time history response of each response channel involved in the time history response error calculation; Indicates the number of response channels involved in the calculation of time history response error; It is a mode shape correlation index, used to measure the degree of consistency between the calculated mode shape vector and the measured mode shape vector under the same measurement point, the same direction, and the same degree of freedom sequence. This represents the root mean square error.

[0071] in, For frequency error, This is the mode shape correlation error. This refers to the time history response error.

[0072] The present invention provides an optimization method for strengthening earthquake-damaged structures based on digital twin models. By weighting the target error function based on frequency error, mode shape correlation error, and time history response error, the accuracy of the target error function value is improved.

[0073] According to the present invention, an optimization method for strengthening a seismically damaged structure based on a digital twin model is provided. The strengthening layer unit includes strengthening material units, connection units, interface units, or energy dissipation units. The zero-stress activation of the strengthening layer unit of the strengthening scheme includes: solving the stress of the seismically damaged digital twin model; calculating spurious strain after the stress solution converges; and setting the strain reference field in the material definition of the strengthening layer unit to be equal to the spurious strain at the moment of activation of the strengthening layer unit.

[0074] The method provided by this invention is applicable to various structural types and reinforcement methods. Structural types include reinforced concrete frame structures, masonry or concrete shear wall structures, etc. Reinforcement methods include external steel cladding, adding energy-dissipating supports, etc. Depending on the structural type and reinforcement method, the reinforcement layer unit includes reinforcement material units, connection units, interface units, or energy-dissipating units.

[0075] To address the challenge of existing technologies failing to realistically simulate the "post-stress" process after reinforcement, this invention proposes an incremental virtual reinforcement method. Its core is a "zero-stress activation" mechanism, achieved by writing a strain reference field. This is achieved through a mechanism that distinguishes it from the simple "add element" operation in conventional finite element software.

[0076] This invention defines the implementation time point for virtual reinforcement. The reinforcement operation must be performed in a separate, newly added analysis step (Load Step) after the seismic damage model has completed the analysis under preceding loads (such as structural self-weight and dead load) and reached a convergence state. This clearly defined triggering time... This ensures a strict distinction between the structural deformation before reinforcement and the new deformation after reinforcement.

[0077] During the reinforcement activation moment The present invention uses the following algorithm to write a strain reference field for all the reinforcement layer units to be activated. : Extracting nodal displacements: First, extract the displacements of the original structural nodes at the locations of the reinforced layer elements. displacement vector at time t .

[0078] Calculating spurious strain: If the reinforced layer element is directly activated, the finite element program will calculate the spurious strain based on the displacement vector. The strain-displacement matrix [B] calculates a huge initial strain that does not conform to physical facts, i.e., a spurious strain. .

[0079] Writing a reference strain: This invention incorporates the strain reference field into the material definition of the reinforced layer unit. Set with spurious strain They are equal. In commercial finite element software, this is typically achieved by defining an equivalent initial strain field or thermal expansion field.

[0080] By writing the strain reference field The constitutive relation of the reinforced layer element was corrected. The actual stress within it was... No longer determined by total strain It is not directly determined by the total strain, but by the difference between the total strain and the reference strain (i.e., the incremental strain). The core relationship is as follows:

[0081] in, It is the constitutive matrix (or stiffness operator) of the reinforced material. At the moment of activation... Total strain Equivalent to false strain The reference strain that has already been written It is also equal to Therefore, the incremental strain within the parentheses is zero, thus ensuring that the initial stress within the reinforced layer unit is zero at the moment of activation. Strictly zero (or below a very small numerical threshold), perfectly achieving "zero-stress activation".

[0082] Subsequently, under new loads (such as subsequent seismic action), It will continue to change, and If it remains unchanged, the reinforcement layer will only bear the load caused by incremental strain. The stress generated thus realistically and accurately simulates the physical fact of the "post-stress" of the reinforced material.

[0083] The present invention provides an optimization method for strengthening earthquake-damaged structures based on digital twin models. By solving the stress in the earthquake-damaged digital twin model and calculating the spurious strain after the stress solution converges, the strain reference field is set to be equal to the spurious strain in the material definition of the strengthening layer unit when the strengthening layer unit is activated. This achieves zero-stress activation of the strengthening layer unit, avoiding the erroneous inclusion of the original structural deformation into the stress of the strengthening material, thereby avoiding the insecurity of the strengthening scheme due to overestimation of the strengthening contribution.

[0084] According to the present invention, an optimization method for strengthening a seismically damaged structure based on a digital twin model is provided. The adjustment of the strengthening scheme includes adjusting the strengthening method and / or strengthening parameters; wherein, the strengthening parameters include interface behavior parameters between new and old materials that form a contact relationship during the strengthening or repair process.

[0085] When adjusting the reinforcement scheme, the reinforcement method and its parameters can be adjusted. This invention also incorporates interface behavior parameters (such as interface slip, interface damage variable, peel initiation index, peel propagation length, and anchorage zone stress) between the old and new materials that form a contact relationship during reinforcement or repair into the optimizable reinforcement parameters. Interface elements based on the traction-separation law can be set at the interface, or equivalent nonlinear spring elements or a custom bond-slip constitutive model can be used to simulate interface behavior.

[0086] The present invention provides an optimization method for strengthening earthquake-damaged structures based on digital twin models. By including interface behavior parameters between new and old materials in the strengthening parameters, it can predict and avoid potential failure modes such as interface peeling, thereby improving the reliability of the strengthening scheme.

[0087] Figure 2 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 2 As shown, the electronic device may include a processor 210, a communications interface 220, a memory 230, and a communication bus 240. The processor 210, communications interface 220, and memory 230 communicate with each other via the communication bus 240. The processor 210 can call logical instructions from the memory 230 to execute the optimization method for strengthening earthquake-damaged structures based on a digital twin model provided in the above embodiments.

[0088] Furthermore, the logical instructions in the aforementioned memory 230 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0089] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the optimization method for strengthening earthquake-damaged structures based on digital twin models provided in the above embodiments.

[0090] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform the optimization method for strengthening earthquake-damaged structures based on digital twin models provided in the above embodiments.

[0091] Those skilled in the art will understand that all or part of the processes of the methods described in the above embodiments can be implemented by a computer program instructing related hardware, and the program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a disk, optical disk, read-only memory, or random access memory, etc.

[0092] The device embodiments described above are merely illustrative. 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 modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0093] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0094] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

[0095] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for optimizing the reinforcement scheme of earthquake-damaged structures based on a digital twin model, characterized in that, include: Obtain a three-dimensional geometric model of the earthquake-damaged structure containing semantic information of the components, as well as crack feature data and measured dynamic response data associated with the three-dimensional geometric model of the earthquake-damaged structure; A digital twin basic model is generated based on the three-dimensional geometric model of the earthquake-damaged structure; Based on the crack feature data, a crack feature vector corresponding to a pre-set seismic damage structure analysis object is determined. A seismic damage state digital twin model is constructed based on the digital twin basic model, the mapping function corresponding to the seismic damage structure analysis object, and the measured dynamic response data. The mapping function represents the mapping relationship between the crack feature vector and the material damage factor. Virtual reinforcement is implemented on the seismic damage digital twin model by activating the reinforcement layer units of the reinforcement scheme with zero stress. The reinforcement scheme is iteratively adjusted according to the pre-set reinforcement performance index constraints and reinforcement implementation optimization objectives until the optimization result of the reinforcement scheme is determined.

2. The method for optimizing the reinforcement scheme of earthquake-damaged structures based on digital twin models according to claim 1, characterized in that, The acquisition of the three-dimensional geometric model of the earthquake-damaged structure containing semantic information of the components, as well as the crack feature data and measured dynamic response data associated with the three-dimensional geometric model of the earthquake-damaged structure, includes: Acquire three-dimensional laser scanning point cloud data, UAV point cloud data and building surface image data of the earthquake-damaged structure, as well as dynamic time history data and building geometric information data collected by pre-deployed sensors; By registering the three-dimensional laser scanning point cloud data and the UAV point cloud data to generate a fused point cloud, and combining the building geometric information data to perform reverse modeling on the fused point cloud to generate the three-dimensional geometric model of the earthquake-damaged structure containing component semantic information. Crack feature data are obtained by processing the image data of the building surface; The measured dynamic response data is obtained based on the dynamic time history data; The crack feature data and the measured dynamic response data are associated and stored with the three-dimensional geometric model of the earthquake-damaged structure containing semantic information of the components.

3. The method for optimizing the reinforcement scheme of earthquake-damaged structures based on digital twin models according to claim 1, characterized in that, The generation of the digital twin base model based on the three-dimensional geometric model of the earthquake-damaged structure includes: Based on the three-dimensional geometric model of the earthquake-damaged structure and the pre-determined rules for dividing key and non-key regions, a first mechanical calculation model is established for the key regions of the earthquake-damaged structure using mesoscopic solid elements; a second mechanical calculation model is established for the non-key regions of the earthquake-damaged structure using macroscopic solid elements; a third mechanical calculation model is established at the interface between the key and non-key regions using non-coordinated mesh interface transfer technology; and the digital twin basic model is generated based on the first, second, and third mechanical calculation models; wherein, the macroscopic solid elements include rod elements or shell elements.

4. The method for optimizing the reinforcement scheme of earthquake-damaged structures based on digital twin models according to claim 3, characterized in that, At the interface between the critical region and the non-critical region, a third mechanical calculation model is established using non-coordinated mesh interface transfer technology, including: Define the nodes or interface reference nodes of the macroscopic entity unit as master nodes, and define the nodes of the microscopic entity unit as slave nodes. The coupling weight of each slave node relative to the master node is calculated based on at least one of the interface area represented by the slave node, the distance from the slave node to the master node, and the interface shape function, and a constraint coefficient matrix is ​​generated based on the coupling weight of each slave node. The third mechanical calculation model is established by generating displacement compatibility equations by incorporating the constraint coefficient matrix into the MPC equation, Lagrange multiplier equation, or penalty function equation; wherein, in solving the third mechanical calculation model, the interface force is transmitted by the constraint reaction force.

5. The method for optimizing the reinforcement scheme of earthquake-damaged structures based on a digital twin model according to claim 3, characterized in that, After generating the digital twin base model, the method further includes: One or more transitional solid units are provided between the mesoscopic solid units and the macroscopic solid units in the digital twin basic model; wherein the stiffness parameter of the transitional solid unit is determined by monotonic interpolation based on the position of the transitional solid unit in the mesoscopic solid units and the macroscopic solid units.

6. The method for optimizing the reinforcement scheme of earthquake-damaged structures based on digital twin models according to claim 1, characterized in that, The construction of the seismic damage state digital twin model based on the digital twin basic model, the mapping function corresponding to the seismically damaged structural analysis object, and the measured dynamic response data includes: After determining the initial values ​​of the undetermined coefficients in the mapping function corresponding to each of the earthquake-damaged structural analysis objects, the following process is executed iteratively until the value of the target error function is less than or equal to a preset threshold or the maximum number of iterations is reached, and the earthquake-damaged digital twin model is output: The material damage factor corresponding to each of the earthquake-damaged structural analysis objects is determined based on the mapping function and the crack feature vector; wherein, the material damage factor includes a set of material constitutive parameter reduction coefficients; Based on the aforementioned digital twin model, the original material constitutive parameters are updated according to the material damage factors corresponding to each of the earthquake-damaged structural analysis objects and written into the material constitutive model of the earthquake-damaged structural analysis object to generate an earthquake-damaged digital twin model. Modal analysis was performed on the aforementioned damaged digital twin model to obtain calculated dynamic response data; The value of the target error function is calculated based on the measured dynamic response data and the calculated dynamic response data. When the value of the target error function is greater than the preset threshold, the undetermined coefficients of the mapping function corresponding to each of the seismically damaged structural analysis objects are adjusted by an optimization algorithm.

7. The method for optimizing the reinforcement scheme of earthquake-damaged structures based on a digital twin model according to claim 6, characterized in that, The target error function is expressed as: in, This represents the target error function. Indicates the undetermined coefficients. , , These are pre-set non-negative weights, used to control the proportions of frequency error, mode shape correlation error, and time history response error in the target error function, respectively. , They represent the first The calculated frequency and measured frequency of each frequency characteristic term involved in the frequency error calculation; This indicates the number of frequency characteristic terms involved in the frequency error calculation; , They represent the first The calculated mode shape vector and the measured mode shape vector corresponding to the mode shape characteristic terms involved in the mode shape correlation calculation; This indicates the number of modal characteristic terms involved in the modal correlation calculation; , They represent the first The calculated time history response and the measured time history response of each response channel involved in the time history response error calculation; Indicates the number of response channels involved in the calculation of time history response error; It is a mode shape correlation index, used to measure the degree of consistency between the calculated mode shape vector and the measured mode shape vector under the same measurement point, the same direction, and the same degree of freedom sequence. This represents the root mean square error.

8. The method for optimizing the reinforcement scheme of earthquake-damaged structures based on digital twin models according to claim 1, characterized in that, The reinforcement layer unit includes a reinforcement material unit, a connection unit, an interface unit, or an energy dissipation unit; the zero-stress activation of the reinforcement layer unit of the reinforcement scheme includes: The stress solution is performed on the digital twin model of the earthquake damage state, and the spurious strain is calculated after the stress solution converges; When the reinforcement layer unit of the reinforcement scheme is activated, the strain reference field in the material definition of the reinforcement layer unit is set to be equal to the dummy strain.

9. The method for optimizing the reinforcement scheme of earthquake-damaged structures based on digital twin models according to claim 1, characterized in that, The adjustment of the reinforcement scheme includes adjusting the reinforcement method and / or reinforcement parameters; wherein, the reinforcement parameters include interface behavior parameters between the old and new materials that form a contact relationship during the reinforcement or repair process.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the method for optimizing the reinforcement scheme of earthquake-damaged structures based on a digital twin model as described in any one of claims 1 to 9.

Images