Targeted sensing-oriented twin substructure interaction method and system and use

Abstract

The present invention relates to a targeted sensing-oriented twin substructure interaction method and system and the use. The method comprises: acquiring multi-element inspection and monitoring data and finite element influence line data of a main structure; on the basis of the inspection and monitoring data and the finite element influence line data of a non-key area, using a sparsity adaptive matching pursuit algorithm to solve a boundary condition of a key area, so as to achieve data fusion of a first level; for the key area, establishing a refined twin substructure finite element model taking into account the effect of structural degradation and, on the basis of the inspection and monitoring data and the finite element influence line data of the key area, correcting the material attribute of the refined twin substructure finite element model to achieve data fusion of a second level; and calculating a correction force on the basis of the boundary condition of the key area and the material attribute, and using the correction force as an equivalent external load to act on a node of a global finite element model, thus achieving the interaction between a substructure and the main structure. The present invention achieves mutual fusion and common communication between the monitoring data and the finite element model and the interaction between each substructure and the main structure.

Description

A targeted perception-guided twin structure interaction method, system and application Technical Field

[0001] The present invention relates to the fields of structural safety assessment and data processing technology, and in particular to a targeted perception-guided twin structure interaction method, system and application. Background Art

[0002] Bridge structural health monitoring and inspection technologies can provide timely and effective information on structural defects. Many scholars have conducted extensive research on directly inverting structural conditions based on inspection and monitoring data. However, for large structures with complex stress states, it is difficult to truly and effectively analyze and evaluate the service performance of the structure using only structural inspection and monitoring data. The rapid development of finite element theory has facilitated bridge structure analysis, and finite element improvement methods using monitoring data have been widely studied. However, existing methods still have the following problems:

[0003] (1) It is difficult to fully integrate the internal and external structural damage in key areas based on existing finite element model improvement methods, and it is difficult to fully integrate inspection and monitoring data to carry out structural analysis and evaluation;

[0004] (2) The actual structure is large in size and the stress state is complex. Carrying out detailed analysis of the entire structure easily leads to low analysis efficiency. Therefore, if the bridge inspection and monitoring information can be fully utilized in the simulation and analysis process, it is expected to break through the bottleneck of bridge structure status analysis and evaluation by combining structural mechanics model analysis (forward modeling) and inspection and monitoring information fusion (inversion).

[0005] Finite element model modification has been widely studied and applied in the field of civil engineering. It is one of the effective ways to integrate finite element models with monitoring data. However, finite element model modification technology mainly achieves consistency between theoretical values ​​and monitoring values ​​at measurement points by modifying internal model parameters. The modified finite element model parameters often reflect the overall structural properties, and the application and mining of monitoring data are not sufficient. The hybrid simulation method in the field of earthquake engineering uses interactive technology to achieve synchronous coupling of numerical substructure analysis calculations and experimental substructure dynamic loading, truly realizing the deep integration of experimental and numerical analysis. Inspired by hybrid simulation, the concept of hybrid monitoring was proposed in the field of structural health, combining bridge monitoring data with finite element models to achieve rapid reconstruction of structural response.

[0006] Integrating inspection and monitoring data to conduct detailed analysis of key areas is more suitable for meeting the needs of bridge structural health systems and structural safety assessments. Existing structural health monitoring systems, however, adhere to the principle of economic rationality in sensor placement. Typically, more sensors are deployed in key structural areas, such as those experiencing greater stress, deformation, or damage, while fewer sensors are deployed in less critical areas.

[0007] However, most existing methods do not consider the interaction between global and local models, and there is still a bottleneck problem that inspection and monitoring data and finite element models are difficult to integrate with each other.

[0008] In summary, the structural finite element analysis method used in the existing structural safety assessment process has a very limited degree of integration of detection and monitoring data. When conducting structural analysis and assessment of existing structures with large size and complex stress conditions, there are problems such as inaccurate assessment structure.

[0009] Summary of the Invention

[0010] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a targeted perception-guided twin structure interaction method, system and application, so as to overcome or partially overcome the bottleneck problem that the inspection and monitoring data and the finite element model are difficult to integrate with each other.

[0011] The purpose of the present invention can be achieved by the following technical solutions:

[0012] One aspect of the present invention provides a targeted perception-guided twin structure interaction method, comprising the following steps:

[0013] Acquire multi-element inspection monitoring data and finite element influence line data of the main structure, wherein the main structure is divided into key areas and non-key areas;

[0014] Based on the inspection and monitoring data of non-key areas and the finite element influence line data, the adaptive sparsity matching tracking algorithm is used to solve the boundary conditions of the key areas to achieve the first level of data fusion;

[0015] A refined twin substructure finite element model is established for the key areas, taking into account the effects of structural degradation. Based on the inspection and monitoring data and finite element influence line data of the key areas, the material properties of the refined twin substructure finite element model are modified to achieve the second level of data fusion.

[0016] The correction force is calculated based on the boundary conditions of the key area and the material properties, and the correction force is applied to the nodes of the global finite element model as an equivalent external load to complete the interaction between the refined twin structure finite element model and the global finite element model.

[0017] As a preferred technical solution, the process of solving the boundary conditions of the key areas includes:

[0018] Establish mathematical equations for inspection data and finite element influence line data in non-key areas;

[0019] The mathematical equation is transformed into an NP-hard non-convex combinatorial optimization problem and solved using the adaptive sparsity matching tracking algorithm.

[0020] As a preferred technical solution, the adaptive sparsity matching tracking algorithm includes the following steps:

[0021] Constructing an influence line matrix based on the finite element influence line data and performing singular value decomposition, projecting the multivariate monitoring data into the subspace formed by the column tensor of the influence line matrix;

[0022] Based on the projection of the multivariate monitoring data on the subspace formed by the influence line matrix column tensor, the sparsity of the item to be solved is changed through iterative calculation, and the sparsity with the highest solution accuracy after multiple iterations is used as the sparsity of the item to be solved.

[0023] As a preferred technical solution, the adaptive sparsity matching tracking algorithm specifically includes:

[0024] Step 1: Input influence line matrix A and monitoring data Y;

[0025] Step 2, perform singular value decomposition on the influence line matrix, and project the monitoring data Y into the subspace spanned by the column vectors of the influence line matrix A, that is, y = Proj A (Y);

[0026] Step 3, initialize r0 = y, Λ0 = φ, t = 1;

[0027] Step 4, calculate the correlation coefficient u=abs[A T r t-1 ], select the 2K maximum values ​​in u, and form the column number set J0 with the column number j corresponding to the maximum value in A;

[0028] Step 5, let Λ t =Λ t-1 ∪J0,A t =A t-1 ∪a j (j∈J0);

[0029] Step 6, Calculation

[0030] Step 7, A t The corresponding K item is recorded as A tK , the column number corresponding to A is recorded as Λ tK , update the set Λ t =Λ tK ;

[0031] Step 8: Calculate and update the error

[0032] Step 9, t = t + 1, if t ≤ 2K, return to Step 2 and continue iteration, otherwise go to Step 10;

[0033] Step 10, update sparsity K = K + ceil (0.02 * size (A, 2));

[0034] Step 11: If the sparsity exceeds K=size(A, 2)*0.5 or the error is less than the preset threshold, then the equivalent node force is output. As the boundary condition of the key area, otherwise execute Step 4,

[0035] Where t is the number of iterations, is an empty set, J0 represents the index obtained in each iteration, Λ t represents the index set of the t-th iteration, and Λ t The number of elements is L t , a j is the jth column of the influence line matrix A, A t ={a j}(j∈Λ t ) represents the index set Λ t The selected column set of the influence line matrix A, θ t For L t ×1 column vector, and the symbol ∪ represents a set union operation.

[0036] As a preferred technical solution, the structural deterioration effects include external crack damage effects and internal corrosion damage effects.

[0037] As a preferred technical solution, the process of constructing a refined twin structure finite element model considering the influence of structural degradation includes:

[0038] The crack weakening element method is used to establish the first reduction relationship between the width of the main structure's external cracks and the weakening element stiffness, thus modeling the impact of external crack damage.

[0039] Based on the material parameters of the steel bar corrosion degradation constitutive model, a second reduction relationship between the steel bar corrosion rate inside the main structure and the structural degradation constitutive model is established to model the impact of internal corrosion damage.

[0040] A refined twin structure finite element model is constructed based on the first reduction relationship and the second reduction relationship.

[0041] As a preferred technical solution, the key areas are the areas with multiple diseases found during the inspection process or the vulnerable areas found in the mechanical analysis, and the non-key areas are the other parts of the main structure except the key areas.

[0042] As a preferred technical solution, the multi-element monitoring data includes node displacement values, node rotation values ​​and strain displacement values, and the material properties include at least one of the constitutive parameters of concrete, steel bars and steel.

[0043] Another aspect of the present invention provides an application of the aforementioned targeted perception-guided twin structure interaction method, comprising the following steps:

[0044] The targeted perception-guided twin structure interaction method is used to complete the interaction between the refined twin structure finite element model and the global finite element model;

[0045] The theoretical displacements of nodes are calculated using the interactively completed global finite element model;

[0046] Get the measured displacement of the node;

[0047] A safety assessment of the bearing capacity of the main structure is performed based on the theoretical displacement and the measured displacement.

[0048] Another aspect of the present invention provides a targeted perception-guided twin structure interaction system, comprising:

[0049] Finite element information extraction module, used to obtain finite element influence line data;

[0050] a boundary condition solving module, configured to solve boundary conditions of key areas based on finite element influence line data of non-key areas of the main structure and acquired multi-element inspection monitoring data, using a preset storage medium, wherein the storage medium includes instructions for implementing an adaptive sparsity matching tracking algorithm;

[0051] The twin substructure fine identification module is used to establish a refined twin substructure finite element model for key areas of the main structure, taking into account the effects of structural degradation;

[0052] The correction feedback module is used to correct the material properties of the refined twin structure finite element model based on the inspection and monitoring data and finite element influence line data of the key areas, calculate the correction force based on the boundary conditions of the key areas and the material properties, and apply the correction force as an equivalent external load to the nodes of the global finite element model.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] (1) Realize the effective integration and communication of inspection and monitoring data and finite element models: To address the bottleneck problem of the difficulty in integrating inspection and monitoring data and finite element models, this application uses the inspection and monitoring data and finite element influence line data of non-key areas, and uses the adaptive sparsity matching tracking algorithm to solve the boundary conditions of key areas, thereby realizing the integration of the first-level inspection and monitoring data and the finite element model; based on the inspection and monitoring data and finite element influence line data of key areas, the material properties of the refined twin structure finite element model are corrected to realize the integration of the second-level inspection and monitoring data and the finite element model. Through the two-level interaction, the information between the global model and the local model of the key areas that need to be finely identified is integrated and communicated, effectively improving the modeling and analysis accuracy of the finite element model of key parts of the structure, and can be widely used in application scenarios such as safety assessment.

[0055] (2) Convenient solution of boundary conditions in key areas: To address the problem of difficulty in deploying sensors at the boundaries of key areas of some structures to measure the physical conditions of their boundaries, the present invention establishes a mathematical equation based on the intrinsic relationship between monitoring data from non-key areas and finite element influence lines, and achieves high-precision solution of boundary conditions in key areas through an adaptive sparsity matching tracking algorithm. This method only requires a small amount of monitoring data from non-key areas to effectively solve the boundary conditions of key areas, solving the problem of difficulty in deploying sensors and measuring the boundaries of complex structures.

[0056] (3) Improve the analysis efficiency of complex structures: To address the problem of low efficiency in conducting structural analysis by establishing an overall refined finite element model of a structure that is large in size and has complex stress states, the present invention converts the time-consuming analysis of establishing a refined model of the overall structure with global damage considerations into a nonlinear analysis of key areas and an equivalent linear elastic analysis of a simplified main structure. There is no need to establish a refined model of the entire structure. The analysis and evaluation of the structure can be completed through interactive analysis of key areas and the main structure, which has the advantages of fast analysis speed and accurate analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] FIG1 is a schematic diagram of the targeted perception-guided substructure interaction method in Example 1;

[0058] FIG2 is a flow chart of a detailed identification process of key structural regions using two-level interaction between inspection and monitoring data and finite element models in Example 1;

[0059] FIG3 is a diagram showing the principle of interactive analysis between the nonlinear analysis of the substructure and the equivalent linear elastic analysis of the main structure in Example 1;

[0060] FIG4 is a flow chart of the adaptive sparsity matching tracking algorithm in Example 1;

[0061] FIG5 is a schematic diagram of the interaction process between the main structure and the substructure in Example 1;

[0062] FIG6 is a flow chart of a structural safety assessment method based on a finite element model of a main structure in Example 1;

[0063] FIG7 is a schematic diagram of an analysis and calculation system for integrating inspection and monitoring data with a finite element model in Example 3;

[0064] FIG8 is a diagram of the experimental device and sensor layout in Example 1;

[0065] FIG9 is a schematic diagram of substructure fine identification by two-level fusion of inspection monitoring data and finite element model in Example 1;

[0066] FIG10 is a diagram showing the results of the refined analysis and evaluation of key regional structures in Example 1;

[0067] FIG11 is a diagram showing the evaluation results of the main structure in Example 1, wherein (a) is the result of displacement meter 3 and (b) is the result of displacement meter 4;

[0068] Figure 12 is a diagram of the experimental device and sensor layout in Example 2;

[0069] FIG13 is a flow chart of the targeted perception-guided substructure interaction method in Example 2;

[0070] FIG14 is a diagram of the substructure analysis results of Example 2, wherein (a) is the result of working condition 1, (b) is the result of working condition 2, (c) is the result of working condition 3, and (d) is the result of working condition 4;

[0071] Figure 15 is a diagram showing the results of the main structure interaction analysis in Example 2, where (a) is the result of working condition 1, (b) is the result of working condition 2, (c) is the result of working condition 3, and (d) is the result of working condition 4.

[0072] Among them, 1. Finite element information extraction module, 2. Mathematical equation construction module, 3. Boundary condition solving module, 4. Twin structure fine identification module, 5. Correction feedback module, 6. Loading device, 7. Camera, 8. Reinforced concrete beam, 9. Displacement meter, 10. Calibration plate. DETAILED DESCRIPTION

[0073] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0074] Some of the definitions involved in this application are as follows:

[0075] Key areas and non-key areas: The main structure consists of two parts: key areas and non-key areas. The key areas are the parts that require special attention and detailed identification, and are selected according to specific application scenarios.

[0076] Substructure and main structure: The main structure is the overall structure of the target building, and the substructure is the structure corresponding to the aforementioned key areas.

[0077] Twin substructure finite element model and global model: The twin substructure finite element model is a finite element model established for the aforementioned substructure, and the global model is a finite element model established for the aforementioned main structure. Compared with the global model, the twin substructure finite element model is established using higher-precision solid units, and the finite element model includes multiple set nodes.

[0078] Example 1

[0079] In response to the problems existing in the aforementioned prior art, this embodiment provides a targeted perception-guided twin substructure interaction method to achieve accurate analysis of important areas and rapid evaluation of the overall main structure. First, in order to address the bottleneck problem of the difficulty in integrating inspection and monitoring data with finite element models, the two-level fusion of inspection and monitoring data and finite element models is achieved through mathematical equations, mathematical equation solving, substructure model enhancement and other technologies, thereby achieving refined analysis of key areas of the structure. Then, in order to address the problem of low efficiency in conducting structural analysis by establishing an overall refined finite element model for structures with large size and complex stress states, an interactive analysis theory of substructure nonlinear analysis and main structure equivalent linear elastic analysis is provided.

[0080] Referring to Figure 1, this method mainly includes two parts: a fusion method of measured data and finite elements and an interactive analysis method of substructure and main structure. The fusion method of measured data and finite elements corresponds to steps S1-S7, and the interactive analysis method of substructure and main structure corresponds to steps S8-S10.

[0081] Referring to Figure 2, the targeted perception-guided twin structure interaction method includes the following steps:

[0082] Step S1: Divide the main structure to be analyzed into key areas and several non-key areas. The key areas can be divided according to the areas with multiple diseases found during the structural inspection process or the vulnerable areas (usually the parts with the greatest stress) found in the mechanical analysis.

[0083] Step S2 , extracting multi-source monitoring data and finite element influence line information (i.e., the overall measured data in FIG1 ) of key areas and non-key areas respectively.

[0084] Step S3, first-level fusion of non-key area monitoring data and finite element model: establish mathematical equations of a small amount of monitoring data from non-key areas and finite element influence line information, corresponding to the mathematical model of overall data and mechanical information in Figure 1. This step includes the following sub-steps:

[0085] Step S31: perform influence line analysis on the simplified global model (i.e., large-scale finite element model), obtain the influence lines of its physical parameters such as rotation angle, displacement, and strain, and construct its influence line matrix W = [D n×n ; R n×n ;E n×n ]. Among them, D n×n 、R n×n and E n×n They represent the displacement influence line matrix, rotation influence line matrix and strain influence line matrix respectively;

[0086] Step S32, establish a mathematical model between the structural response such as rotation angle, displacement and strain, the influence line matrix and the load: [x n×1 θ n×1 ε n×1 ]=[D n×n ; R n×n ;E n×n ]·F n×1

[0087] Where x n×1 is the node displacement value, θ n×1 is the node rotation value, ε n×1 is the strain displacement value, F n×1 is the node load.

[0088] In step S33, only the row and column data related to non-key areas in the above mathematical equation are retained. On this basis, the influence of various monitoring data on the solution of the equation is eliminated through the normalization coefficient (α, β, γ are normalization coefficients), and finally the mathematical equation for the fusion of non-key area monitoring data and finite element model is established:

[0089] Where, is the M displacement monitoring data of the non-key area of ​​the structure, N corner monitoring data of non-key areas of the structure, L strain monitoring data of non-key areas of the structure.

[0090] Step S4, as shown in FIG3 , is to transform the mathematical equation into an NP-hard non-convex combinatorial optimization problem through mathematical deduction based on the construction of the mathematical equation.

[0091] Considering that there is usually a certain error in solving the above underdetermined equation, a small error term is introduced to further express the above mathematical equation as: Y = AF + E. In the formula, A=[αD M×n βR N×n γE L×n ], E is the small error term in solving the underdetermined equation.

[0092] Perform singular value decomposition on the influence line matrix A, and the decomposition process can be expressed as: A=U∑V H By singular value decomposition, the (M+N+L)*n matrix A is decomposed into a (M+N+L)*(M+N+L) unitary matrix U, a (M+N+L)*n diagonal matrix ∑ and a conjugate transposed matrix VH of an n*n unitary matrix V;

[0093] The multi-source monitoring data Y is projected onto the subspace span(A) spanned by the column vectors of the influence line matrix A. The projection process can be expressed as: A (Y)=U(U H U) -1 U H Y;

[0094] After the above steps, the original mathematical equation Y=AF+E can be transformed into: Proj A (Y)=Proj A (AF)+Proj A (E), the expanded expression is: U(U H U) -1 U H Y=U(U H U) -1 U H AF+U(U H U) -1 U H E=AF+U(U H U) -1 U H E

[0095] Let y = U(U H U) -1 U H Y,e=U(U H U) -1 U H E, then the original mathematical equation Y = AF + E can be expressed as: y = AF + e. H U) -1 U H E||≤||E||, which means that after projection, the errors caused by noise and other factors can be effectively suppressed, further improving the accuracy of solving the equation.

[0096] Because F is sparse, the above mathematical equation has a unique solution and can be transformed into a minimum l0-norm optimization problem: min||F||0, st||y-AF||<<e. Since this problem is an NP-hard non-convex combinatorial optimization problem, it can be solved using the adaptive sparsity matching pursuit algorithm described later.

[0097] Step S5, as shown in FIG4, is based on the adaptive sparsity matching pursuit algorithm provided by the present invention to solve the above mathematical equations, thereby achieving a high-precision solution of the boundary conditions of the key areas (i.e., the substructure boundary conditions in FIG1). The algorithm includes the following steps:

[0098] Step S501, input: influence line matrix A, monitoring data Y;

[0099] Step S502: perform singular value decomposition on the influence line matrix and project Y onto the subspace spanned by the column vectors of matrix A, i.e., y = Proj A (Y);

[0100] Step S503, initialize r0=y, Λ0=φ, t=1, error max =1E-2, K=size(A, 2)*0.1;

[0101] Step S504, calculate the correlation coefficient u=abs[A T r t-1 ], select the 2K largest values ​​in u, and form the set J0 (column number set) with the column numbers j corresponding to these values ​​in A;

[0102] Step S505, let Λ t =Λ t-1 ∪J0,A t =A t-1 ∪a j (j∈J0);

[0103] Step S506, calculate y=A t θ t ,Right now

[0104] Step S507, A t The corresponding K item is recorded as A tK , the column number corresponding to A is recorded as Λ tK , update the set Λ t =Λ tK ;

[0105] Step S508, calculate and update the error:

[0106] Step S509, t=t+1, if t≤2K, return to step S503 to continue iteration;

[0107] Step S510: If the condition t≤2K is not satisfied, update the sparsity K=K+ceil(0.02*size(A,2)), where size(A,2) refers to the number of columns of the matrix A.

[0108] Step S511: If the sparsity exceeds K=size(A, 2)*0.5 or the error is less than the threshold Threshold, proceed to step S6; otherwise, return to step S503.

[0109] Where t is the number of iterations, is an empty set, J0 represents the index found in each iteration, Λ t Represents the index set of the tth iteration (assuming the number of elements is L t ), a j is the jth column of the influence line matrix A, A t ={a j}(j∈Λ t ) represents the index set Λ t The selected column set of the influence line matrix A, θ t For L t ×1 column vector, and the symbol ∪ represents a set union operation.

[0110] Conventional matching tracking algorithms require prior knowledge of the sparsity of the item to be solved. However, the sparsity in practical applications is usually difficult to obtain and needs to be estimated. If the value of K is underestimated, the algorithm's ability to accurately solve the problem will decrease or may even be eliminated, causing the algorithm to no longer converge; if the value of K is overestimated, the algorithm's robustness and solution accuracy will decrease, causing the solution error to increase. To address this problem, this embodiment provides an adaptive sparsity method, which iteratively changes the sparsity of the item to be solved, so that the sparsity with the highest solution accuracy after multiple iterations is used as the sparsity of the item to be solved, effectively avoiding the problem of needing to know the sparsity in advance and achieving accurate solutions to mathematical equations.

[0111] Step S6: In the finite element analysis software, a higher-precision solid element is used to establish a twin structure finite element model corresponding to the key area, and a finer finite element mesh is divided according to actual analysis needs:

[0112] Step S7, the second-level fusion of the inspection and monitoring data of key areas and the finite element model: establish the reduction relationship between the inspection data such as the width of the external cracks of the structure, the internal steel corrosion rate and the structural degradation constitutive model, and consider the existing cracks, corrosion and other structural defects in the key areas through the crack weakening unit method and the steel corrosion degradation constitutive model to realize the construction of the twin substructure, and carry out preliminary calculation of the substructure based on the boundary conditions solved in step S5; on this basis, based on the multi-type monitoring data such as strain and displacement densely distributed in the key areas, the model correction method is used to correct the material properties of the key areas, and realize the refined modeling of the substructure considering cracks and corrosion. This step corresponds to the construction of the reduction relationship based on the local measured data in Figure 1 to perform intelligent crack detection. This step includes the following substeps:

[0113] Step S71: Establish a reduction relationship between the detection data such as the corrosion rate of the internal steel bars of the structure and the structural degradation constitutive model. Based on the degradation constitutive model of the corroded steel bars and the corrosion rate detection data, reduce the material parameters such as the interface area, yield strength, tensile strength and elastic modulus of the internal steel bars of the structure to take into account the impact of the internal steel bars on the structure.

[0114] Step S72 establishes a reduction relationship between the structural external crack width and other test data and the weakened element stiffness. Based on the specifications and crack width, the weakened element reduction coefficient is inferred. The crack is then simulated at the crack site using the weakened element method. It is worth noting that the CDP model is still used to simulate the plastic behavior of the concrete beam. The Young's modulus and tensile strength in the cracked area are both based on weakened material properties, while the concrete compressive strength remains unchanged. The reduction relationship refers to the material properties of the structure, the material constitutive parameters required for finite element calculations.

[0115] Step S73: On this basis, based on the densely distributed strain, displacement and other types of monitoring data in the key area, the material properties of the key area are corrected using the existing model correction method.

[0116] The substructure model updating method provided in this embodiment can meet the needs of various structural analyses. This method fully considers measured data such as structural corrosion rate and crack width, and updates the material parameters of structural defects by integrating the structural degradation mechanics model.

[0117] Step S8, based on the above steps, by establishing a twin substructure finite element model for the key area taking into account the influence of structural degradation, a refined analysis of the structural mechanical state of the key area is achieved.

[0118] Step S9: Use the refined twin structure finite element model to obtain the boundary node reaction force and material properties of the key area, and transmit the boundary reaction force and material properties back to the main structure.

[0119] As shown in FIG5 , the interaction analysis between the substructure and the main structure is performed in steps S9 - S10 .

[0120] Step S10, calculate the correction force, and apply the correction force as an equivalent external load to the corresponding nodes of the main structure to complete the displacement coordination between the main structure and the substructure, complete the interaction between the refined twin substructure finite element model and the global model, and realize the equivalent linear elastic analysis of the main structure (corresponding to the global finite element model).

[0121] The theoretical derivation process of steps S9-S10 is as follows:

[0122] For the unit integration area of ​​the structure, the virtual work principle can be obtained

[0123] Where σ is the Cauchy stress tensor, ε is the Green strain tensor, b is the body force, ρ is the material density, η is the damping coefficient, u is the displacement, and Ω is the e is the unit integration domain, S is the unit integration domain Ω e The boundary of , t is the boundary force.

[0124] Add both sides of the above formula (D e is the material stiffness), then:

[0125] In local coordinates, the relationship between the displacement of any point in the unit and the displacement of the unit node is expressed as: u = N e v e , the relationship between strain and unit node displacement is: ε=B e v e Where u is the displacement of any point in the unit, N e is the shape function, v e is the displacement in local coordinates, ε is the strain at any point in the unit, B e is the strain matrix.

[0126] From this, the equivalent motion equation of the unit in global coordinates can be obtained:

[0127] Where u e 、 and Represent node displacement, velocity and acceleration respectively, m e 、c e 、k e 、f e 、 and r e Represents the unit's mass matrix, damping matrix, initial stiffness matrix, external forces, nonlinear correction forces, and unit internal forces. Specifically:

[0128] By integrating the equivalent motion equations of all units, we can obtain the control equation of the equivalent linear elastic analysis of the entire structure in global coordinates: K0U=F+F c

[0129] Where U represents the node displacement of the structure, K0 and F are the stiffness matrix and external force of the structure respectively, and N e Indicates the total number of structural units, N s Represents the number of nonlinear units. Specifically, it can be expressed as:

[0130] The solution of this embodiment is described below by constructing an experimental scenario.

[0131] As shown in Figure 8, the experiment used an MTS hydraulic loading device to conduct a four-point bending test on a reinforced concrete beam with coupled cracks and corrosion. The test used two-point loading to ensure pure bending deformation at the mid-span. The bending bearing capacity of the beam was evaluated by integrating detection and monitoring data. The experimental loading process was carried out using load control, and the crack development was recorded every 5kN. Data from long-gauge fiber optic sensors, strain gauges, and displacement gauges were also recorded. (a) is a schematic diagram of the scene structure, and (b) is a schematic cross-sectional diagram of the reinforced concrete beam.

[0132] The sensor layout adheres to the principle of regional distribution, with densely distributed sensors in key areas and fewer sensors in less critical areas. Figure 8 shows the sensor layout, with 12 long-gauge fiber optic sensors deployed, designated LS1 through LS12. To monitor the deformation of the reinforced concrete beam, five displacement gauges, designated 1 through 5, were placed at the bottom of the beam. To prioritize the performance of corrosion cells, five strain gauges were placed on one side of the beam, spaced 40 cm apart. An industrial camera was used to record the development of cracks in the reinforced concrete beam.

[0133] As shown in FIG9 , according to the method for fine-grained analysis of key structural areas by integrating two levels of inspection and monitoring data and finite element models disclosed in the present invention, the following operations are carried out on the reinforced concrete structure in Example 1: (1) according to the inspection results of the reinforced concrete beam, the structure to be analyzed is divided into key areas (substructures) and several non-key areas; (2) multi-source monitoring data and finite element influence line information of key areas and non-key areas are extracted respectively; (3) mathematical equations of a small amount of monitoring data and finite element influence line information of non-key areas are established; (4) based on the derivation of mathematical equations, the above mathematical equations are converted into NP-hard non-convex combinatorial optimization problems; (5) based on the adaptive sparsity matching tracking algorithm provided in this embodiment, the optimization problem is realized. The above mathematical equations are solved to achieve high-precision solution of the boundary conditions of the key areas; (6) A more precise solid element is used to establish a refined finite element model of the key areas, and a finer finite element grid is divided according to the actual analysis needs; (7) The reduction relationship between the detection data such as the width of the external cracks of the structure and the internal steel corrosion rate and the structural degradation constitutive model is established, and the existing cracks, corrosion and other structural defects in the key areas are considered through the crack weakening element method and the steel corrosion degradation constitutive model to establish a twin structure; on this basis, based on the multi-type monitoring data such as strain and displacement densely distributed in the key areas, the model correction method is used to correct the material properties of the key areas; (8) Based on the above steps, a refined analysis of the structural mechanical state of the key areas is achieved.

[0134] As shown in Figure 10, the finite element model of the key areas established using the method disclosed in the present invention is highly accurate, and the analysis results for the key areas are generally consistent with the experimental structure. Furthermore, the crack distribution in the finite element model of the key areas is generally consistent with the experimental distribution. Cracks are primarily distributed at mid-span, with prominent main cracks appearing near the rust, where the element stiffness and bearing capacity are significantly reduced.

[0135] According to the method disclosed above, the results of the main structure interaction analysis are shown in Figure 11. The load-displacement curves of the substructure nonlinear analysis and the main structure equivalent linear elastic analysis disclosed in this embodiment are basically consistent with the measured load-displacement curves and the load-displacement curves of the refined finite element model, verifying the accuracy of the bearing capacity assessment results of the concrete beam using the method proposed in this article. In terms of analysis efficiency, the global refined finite element model requires 15 minutes and 53 seconds to complete a single analysis, while the substructure interaction analysis method proposed in this article only requires 9 minutes and 3 seconds for a single calculation, greatly improving the analysis and evaluation efficiency.

[0136] This method has the following advantages:

[0137] (1) To address the bottleneck problem of the difficulty in integrating inspection and monitoring data with finite element models, by establishing mathematical equations for the internal connection between non-key areas and the finite element model influence line model, and improving the finite element model of key areas considering internal corrosion and external cracks, the two-level fusion of inspection and monitoring data and finite element models effectively improves the accuracy of finite element modeling and analysis of key structural parts. Through the fusion of inspection and monitoring data, the analysis accuracy of key structural areas is improved by more than 10%;

[0138] (2) When it is difficult to deploy sensors at the boundaries of key areas of some structures to measure the physical conditions of their boundaries, the present invention establishes a mathematical equation based on the intrinsic relationship between the monitoring data of non-key areas and the finite element influence line, and achieves high-precision solution of the boundary conditions of key areas through an adaptive sparsity matching tracking algorithm. This method only requires a small amount of monitoring data from non-key areas to effectively solve the boundary conditions of key areas, solving the problem of difficult deployment of sensors at the boundaries of complex structures and inability to measure the boundaries;

[0139] (3) To address the problem of low efficiency in conducting structural analysis by establishing an overall refined finite element model for a large structure with complex stress states, the present invention innovatively proposes an interactive analysis method of refined analysis of key areas and equivalent linear elastic analysis of the main structure, which converts the time-consuming analysis of establishing a refined model of the entire structure with global damage considerations into nonlinear analysis of key areas and equivalent linear elastic analysis of a simplified main structure. This method does not require the establishment of a refined model of the entire structure, and can complete the analysis and evaluation of the structure through interactive analysis of key areas and the main structure. It has the advantages of fast analysis speed and accuracy, and improves efficiency by more than 15%;

[0140] Example 2

[0141] In order to further verify the feasibility of the method in steel structure, this paper uses a scaled model of a steel arch bridge with a tie rod to carry out experimental research. As shown in Figure 12, the total length of the arch bridge is about 5.2m, the width is about 0.875m, and the height is about 0.646m. The main material of the arch bridge is Q235 steel, and it mainly includes tie beams, cross beams, arch ribs and supports. Among them, the tie beam has an H-section of 125*125*6.5*9, the cross beam has an H-section of 100*100*6*8, and the arch rib has a cross section of Strain sensors were primarily deployed using a regional distribution method, with densely packed sensors in key mid-span areas and relatively fewer sensors in other locations. Strain sensors used long-gauge fiber optic sensors with a gauge length of 0.26m. Nine displacement and rotation measurement points were located at the beam nodes. A loading device was used between beams P4 and P7, with four load levels set: 20kN, 40kN, 52kN, and 63kN.

[0142] As shown in Figure 13, based on the aforementioned targeted perception-guided twin substructure interaction method, the following steps are performed: (1) a mathematical model of bridge monitoring data and mechanical information of large-scale finite element model is established; (2) a new compressed sampling matching pursuit algorithm is proposed by integrating singular value decomposition (SVD) and adaptive sparsity to accurately solve the underdetermined equation and achieve high-precision solution of structural boundary displacement; (3) based on large-scale finite element simulation, combined with the characteristics of structural stress analysis and structural inspection results, a refined substructure model is established using solid elements in the key areas of the structure, and an independent analysis is performed on the refined substructure model based on the boundary conditions solved above; (4) the boundary reaction of the substructure is converted into an equivalent external load of the main structure and an equivalent linear elastic analysis is performed on the main structure to achieve interactive analysis of the substructure and the main structure. Finally, the information transmission between "monitoring data-finite element model" and "overall large-scale model-local refined model" is achieved.

[0143] As shown in Figure 14, the stress state of the substructure can be effectively analyzed based on the method proposed in this paper, and the strains under four different loading conditions are in good agreement with the experimental results. As shown in Figure 15, based on the numerical substructure theory, the boundary force of the substructure is converted into an equivalent external load of the main structure and an equivalent linear elastic analysis is performed on the main structure. Based on the modified finite element model of the main structure, a global analysis is performed to obtain its displacement response curve. The load-displacement curve based on the large-scale finite element model is basically consistent with the measured load-displacement curve. Among them, when the bridge structure is under the maximum load (63kN), the measured displacement value of the bridge is less than the theoretical calculated displacement value, the structural verification coefficient η is less than 1, the structural bearing capacity meets the requirements, and the structure is safe. This example further verifies the effectiveness of the proposed method in the field of steel structures.

[0144] The present invention adopts the concept of "targeted perception", that is, according to the characteristics of the existing structural health monitoring system, it divides the areas of focus into numerical substructures in a targeted manner, and then uses more refined solid units to construct a twin substructure finite element model to achieve "secondary analysis". In the process of "secondary analysis", on the one hand, the corresponding structural damage mechanism can be introduced for specific diseases, so as to achieve focused analysis of key areas. On the other hand, the monitoring data of key areas can be fully utilized, and the monitoring data can be fully integrated with the finite element model through mathematical equations, model improvement and other methods, so as to carry out refined analysis of the mechanical properties of the substructures in key areas.

[0145] Example 3

[0146] Compared with Example 1 or Example 2, the main structure in this embodiment has no defects, so the substructure material attributes are updated only based on step S73. This method has good practicality and is suitable for modifying various structural models such as concrete and steel structures.

[0147] Example 4

[0148] On the basis of the above-mentioned embodiment, this embodiment provides a method for realizing the bearing capacity safety assessment by using the above-mentioned targeted perception-guided twin substructure interaction method, and utilizes the analysis method of the interaction between the substructure and the main structure in the above-mentioned embodiment to indirectly integrate the monitoring data information into the main structure through the interaction between the substructure and the main structure. The structural safety assessment based on the established main structure finite element model has both accuracy and efficiency. As shown in Figure 6, the sub-assessment method based on the main structure finite element model is specifically embodied in: the theoretical calculated displacement U is obtained through the main structure finite element model. C , the measured displacement U obtained by the actual structural health monitoring system T , comprehensive theoretical calculation of displacement U C and the measured displacement UT, calculate the calibration coefficient The structural safety assessment is carried out based on this verification coefficient. If the verification coefficient η≤1, the structure is judged to be in a safe operating state; if the verification coefficient η>1, the structural bearing capacity is judged to be unsatisfactory.

[0149] Example 5

[0150] Referring to FIG1 , this embodiment provides a targeted perception-guided twin structure interaction system to implement analytical calculations integrating inspection and monitoring data with finite element models. The system includes:

[0151] Finite element information automatic extraction module: used to perform influence line analysis on simplified finite element models. This module mainly provides functions such as batch modification of node loads, batch submission of jobs, and batch extraction of influence line information such as displacement, strain, and rotation. It helps to improve the efficiency of finite element information extraction and realizes the automatic extraction of multi-source monitoring data and finite element influence line information;

[0152] The mathematical equation construction module is used to construct mathematical equations that intrinsically link monitoring data, finite mechanical information, and node loads. The mathematical equation construction module is used to solve the above mathematical equations and obtain the boundary conditions of important areas (substructures). This module integrates singular value analysis, subspace projection, and adaptive sparsity matching tracking algorithms to achieve accurate solution of the physical boundaries of substructures in important areas to be analyzed.

[0153] Boundary Condition Solver: This module is used to construct mathematical equations that intrinsically link monitoring data, finite element information, and nodal loads. This module features automatic data reading, the construction of complete equations, and the automatic construction of underdetermined equations, enabling the first-level integration of monitoring data from non-key areas with the finite element model.

[0154] Twin Substructure Refined Identification Module: This module is used to establish more accurate solid element models for key areas, taking into account common structural defects such as cracks and corrosion. This module features refined solid element modeling, crack and corrosion defects, and substructure finite element model correction, enabling refined identification of key areas.

[0155] Correction Feedback Module: This module is used to obtain the boundary node reactions and material properties of key areas and transmit them back to the main structure. This module features automatic acquisition of node reactions and material properties, automatic calculation of correction forces, and real-time data transmission. The correction forces are then applied to the corresponding nodes of the main structure as equivalent external loads, enabling secondary analysis of the main structure.

[0156] The above modules organically integrate the aforementioned targeted perception-guided twin structure interaction method, with a high degree of automation, and can quickly and efficiently realize bridge analysis and evaluation based on the fusion of inspection and monitoring data.

[0157] Example 6

[0158] This embodiment provides an electronic device, including: one or more processors and a memory, wherein the memory stores at least one program, and the program includes instructions for executing the targeted perception-guided twin structure interaction method as in the aforementioned embodiment.

[0159] Example 7

[0160] This embodiment provides a computer-readable storage medium, including at least one program for execution by at least one processor of an electronic device, the program including instructions for executing the targeted perception-guided twin structure interaction method in the aforementioned embodiment.

[0161] The foregoing description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art will readily conceive of various equivalent modifications or substitutions within the technical scope disclosed herein, and such modifications or substitutions are intended to be encompassed within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be subject to the scope of protection of the claims.

Claims

1. A targeted perception-guided twin structure interaction method, characterized in that: The steps include: Acquire multi-element inspection monitoring data and finite element influence line data of the main structure, wherein the main structure is divided into key areas and non-key areas; Based on the inspection and monitoring data of non-key areas and the finite element influence line data, the adaptive sparsity matching tracking algorithm is used to solve the boundary conditions of the key areas to achieve the first level of data fusion; A refined twin substructure finite element model considering the influence of structural degradation is established for the key areas, and based on the inspection and monitoring data and finite element influence line data of the key areas, the material properties of the refined twin substructure finite element model are corrected to achieve the second level of data fusion; The correction force is calculated based on the boundary conditions of the key area and the material properties, and the correction force is applied to the nodes of the global finite element model as an equivalent external load to complete the interaction between the refined twin structure finite element model and the global finite element model.

2. A targeted perception-guided twin structure interaction method according to claim 1, characterized in that: The process of solving the boundary conditions of the key areas includes: Establish mathematical equations for detection data and finite element influence line data of non-key areas; The mathematical equation is transformed into an NP-hard non-convex combinatorial optimization problem and solved using the adaptive sparsity matching tracking algorithm.

3. A targeted perception-guided twin structure interaction method according to claim 1, characterized in that: The adaptive sparsity matching tracking algorithm comprises the following steps: Constructing an influence line matrix based on finite element influence line data and performing singular value decomposition, projecting the multivariate inspection monitoring data into a subspace formed by the column tensor of the influence line matrix; Based on the projection of the multivariate monitoring data on the subspace formed by the influence line matrix column tensor, the sparsity of the item to be solved is changed through iterative calculation, and the sparsity with the highest solution accuracy after multiple iterations is used as the sparsity of the item to be solved.

4. A targeted perception-guided twin structure interaction method according to claim 3, characterized in that: The adaptive sparsity matching tracking algorithm specifically includes: Step 1, input influence line matrix A and monitoring data Y; Step 2: Perform singular value decomposition on the influence line matrix and project the monitoring data Y onto the influence line matrix A. The subspace spanned by the column vector, that is, y = Proj A (Y); Step 3, initialize r0=y, Λ0=φ, t=1; Step 4, calculate the correlation coefficient u = abs[A T r t-1 ], select the 2K maximum values ​​in u, and form the column number set J0 with the column number j corresponding to the maximum value in A; Step 5, let Λ t = Λ t-1 ∪ J0, A t = A t-1 ∪ a j (j ∈ J0); Step 6, Calculate Step7, A t The corresponding K item is denoted as A tK , the column number corresponding to A is recorded as Λ tK , update the set Λ t =Λ tK ; Step 8, calculate and update the error Step 9, t = t + 1, if t ≤ 2K, return to Step 2 to continue iteration, otherwise go to Step 10; Step 10, update sparsity K = K + ceil (0.02 * size (A, 2)); Step 11, if the sparsity exceeds K = size (A, 2) * 0.5 or the error is less than the preset threshold Threshold, then the equivalent node force F = θ^_tK is output as the boundary condition of the key area, otherwise, Step 4 is executed. Where t is the number of iterations, is an empty set, J0 represents the index obtained in each iteration, ∧ t represents the index set of the tth iteration, and ∧ t The number of elements is L t , a j is the jth column of the influence line matrix A, A t ={a j }(j∈∧ t ) means by index set ∧ t The selected set of columns of the influence line matrix A, θ t For L t ×1 column vector, and the symbol ∪ represents a set union operation.

5. The targeted perception-guided twin structure interaction method according to claim 1, characterized in that: The structural deterioration effects include external crack disease effects and internal corrosion disease effects.

6. A targeted perception-guided twin structure interaction method according to claim 5, characterized in that: The process of constructing the refined twin structure finite element model considering the influence of structural degradation includes: The crack weakening unit method is used to establish the first reduction relationship between the external crack width of the main structure and the weakening unit stiffness, so as to realize the modeling of the influence of external crack diseases; Based on the material parameters inside the steel bar corrosion degradation constitutive model, the second reduction relationship between the steel bar corrosion rate inside the main structure and the structural degradation constitutive model is established to achieve modeling of the impact of internal corrosion diseases; A refined twin structure finite element model is constructed based on the first reduction relationship and the second reduction relationship.

7. A targeted perception-guided twin structure interaction method according to claim 1, characterized in that: The key areas are areas with multiple diseases found during the inspection process or vulnerable areas in mechanical analysis, and the non-key areas are other parts of the main structure except the key areas.

8. The targeted perception-guided twin structure interaction method according to claim 1, characterized in that: The multivariate monitoring data includes node displacement values, node rotation values ​​and strain displacement values, and the material properties include at least one of the constitutive parameters of concrete, the constitutive parameters of steel bars, and the constitutive parameters of steel.

9. An application of the targeted perception-guided twin structure interaction method as claimed in any one of claims 1 to 8, characterized in that: The steps include: The targeted perception-guided twin structure interaction method is used to complete the interaction between the refined twin structure finite element model and the global finite element model; The theoretical displacements of nodes are calculated using the interactively completed global finite element model; Get the measured displacement of the node; The safety of the bearing capacity of the main structure is evaluated based on the theoretical displacement and the measured displacement.

10. A targeted perception-guided twin structure interaction system, characterized in that: include: A finite element information extraction module (1), used for obtaining finite element influence line data; Mathematical equation construction module (2), used to construct mathematical equations for the intrinsic relationship between monitoring data, finite mechanics information and node loads; A boundary condition solving module (3) is used to solve the boundary conditions of the key area based on the finite element influence line data of the non-key area of ​​the main structure and the obtained multi-element inspection monitoring data, for the mathematical equation, through a preset storage medium, wherein the storage medium includes instructions for implementing an adaptive sparsity matching tracking algorithm; A twin substructure fine identification module (4) is used to establish a fine twin substructure finite element model for key areas of the main structure taking into account the influence of structural degradation; A correction feedback module (5) is used to correct the material properties of the refined twin structure finite element model based on the inspection and monitoring data and finite element influence line data of the key area, calculate the correction force based on the boundary conditions of the key area and the material properties, and use the correction force as an equivalent external load to act on the nodes of the global finite element model.

Images