Large-scale structure reanalysis method and system based on separation-reconstruction framework

Abstract

The application discloses a large-scale structure reanalysis method and system based on a separation-reconstruction framework, which comprises the following steps: dividing a truss structure into multiple subdomains and constructing an extraction matrix; constructing a flexibility matrix corresponding to the subdomains according to a total stiffness matrix and the extraction matrix, and modifying the stiffness of the corresponding subdomains according to structure design modification, thereby constructing a dimension-reduced modification equation of the subdomains; iteratively solving the dimension-reduced modification equation of the subdomains according to a stable double-conjugate gradient method to obtain displacement modification amounts of the subdomains; mapping the displacement modification amounts of the subdomains to the whole structure to reconstruct a displacement response vector of the whole structure; calculating an internal and external force residual vector of the displacement response vector, and judging whether the internal and external force residual vector meets a preset convergence condition; and if yes, outputting a structure reanalysis result. The local modification structure reanalysis method based on dimension reduction iteration, in combination with a high-dimensional model representation method and a residual modification mechanism, can improve the reanalysis efficiency of large-scale or whole modified structures.

Description

Technical Field

[0001] This invention relates to a large-scale structural reanalysis method and system based on a separation-reconstruction framework, belonging to the field of high-performance algorithms for structural analysis. Background Technology

[0002] In the field of engineering computation, especially in problems requiring repeated solutions to structural responses, such as structural optimization design and nonlinear analysis, efficient structural re-analysis techniques are of crucial importance. The goal is to quickly and accurately assess structural responses without re-performing a complete finite element analysis when design parameters or external loads change. Despite continuous improvements in computer hardware performance, the ever-increasing demands of engineering computation remain difficult to fully meet. In finite element analysis, the mathematical model typically reduces to solving a system of linear equations; however, in structural optimization, the structural behavior under different parameter settings needs to be evaluated multiple times, leading to repeated solutions to the complete linear equations. As the scale of the structure increases, the computational load of solving these equations grows exponentially, creating a significant efficiency bottleneck and severely hindering the development of engineering applications.

[0003] To avoid repeatedly solving the complete system of equations, researchers have proposed a variety of reanalysis methods, which can be summarized into two categories: direct methods and approximate methods. (1) The Sherman-Morrison-Woodbury (SMW) formula method, as a representative of direct methods, uses the decomposition results of the original stiffness matrix to quickly obtain the displacement response of the modified structure; however, when the stiffness modification range increases, its efficiency drops sharply and it is only suitable for small-scale modifications. (2) Approximate methods approximate the exact solution through iteration, mainly including: a) Single-point approximation method - high computational efficiency, but accuracy is difficult to guarantee when making large modifications; b) Multi-point approximation method - higher accuracy, can handle large modification problems, but computational overhead is large; c) Preprocessed conjugate gradient method (PCG) - high solution accuracy, fast convergence and easy to implement, but its advantages are not obvious for small modifications. (3) The local modification structural reanalysis method that has emerged in recent years integrates system simplification and iterative solution strategies. It expresses the equilibrium equation after local physical property modification as the sum of the original stiffness matrix and the pseudo load introduced by the modification. Based on the spatial distribution of stiffness modification elements, the dimension reduction equation is reconstructed, and then the simplified precondition technique is used for iterative solution, which significantly improves the computational efficiency in the case of local modification.

[0004] However, existing technologies still have significant shortcomings: although the local modification reanalysis method performs well in local modification problems, its applicability and computational efficiency are still insufficient when faced with large modifications or even changes in the overall structure, and further development is urgently needed. Summary of the Invention

[0005] In view of this, the present invention provides a large-scale structural reanalysis method and system based on the separation-reconstruction framework. It is based on a local correction structural reanalysis method of dimensionality reduction iteration, combined with high-dimensional model representation method and residual correction mechanism, so as to improve the reanalysis efficiency of large-scale or overall modified structures.

[0006] The first objective of this invention is to provide a large-scale structural reanalysis method based on a separation-reconstruction framework.

[0007] The second objective of this invention is to provide a large-scale structural reanalysis system based on a separation-reconstruction framework.

[0008] A third objective of this invention is to provide a computer device.

[0009] A fourth objective of this invention is to provide a storage medium.

[0010] The first objective of this invention can be achieved by adopting the following technical solution:

[0011] A large-scale structural reanalysis method based on a separation-reconstruction framework includes: dividing a truss structure into multiple subdomains and constructing an extraction matrix, wherein the extraction matrix is ​​used to transmit displacement and stiffness information between the subdomains and the overall structure; constructing a flexibility matrix corresponding to each subdomain based on the overall stiffness matrix and the extraction matrix, and correcting the stiffness of the corresponding subdomains according to structural design modifications, thereby constructing a reduction maintenance correction equation for each subdomain; iteratively solving the reduction maintenance correction equation for each subdomain using the stable biconjugate gradient method to obtain the displacement correction amount for each subdomain; mapping the displacement correction amounts of each subdomain to the overall structure to reconstruct the displacement response vector of the overall structure; calculating the residual vectors of internal and external forces of the displacement response vector, and determining whether the residual vectors of internal and external forces satisfy a preset convergence condition; if satisfied, outputting the structural reanalysis result.

[0012] The second objective of this invention can be achieved by adopting the following technical solution:

[0013] A large-scale structural reanalysis system based on a separation-reconstruction framework includes: a first construction unit for dividing a truss structure into multiple subdomains and constructing an extraction matrix, wherein the extraction matrix is ​​used to transmit displacement and stiffness information between the subdomains and the overall structure; a second construction unit for constructing a flexibility matrix corresponding to the subdomain based on the total stiffness matrix and the extraction matrix, and correcting the stiffness of the corresponding subdomain based on structural design modifications, thereby constructing a reduction maintenance correction equation for the subdomain; a solution unit for iteratively solving the reduction maintenance correction equation for the subdomain using the stable biconjugate gradient method to obtain the displacement correction amount for each subdomain; a reconstruction unit for mapping the displacement correction amounts of each subdomain to the overall structure, thereby reconstructing the displacement response vector of the overall structure; and an output unit for calculating the residual internal and external force vectors of the displacement response vector, and determining whether the residual internal and external force vectors satisfy a preset convergence condition. If satisfied, the structural reanalysis result is output.

[0014] The third objective of this invention can be achieved by adopting the following technical solution:

[0015] A computer device includes a processor and a memory for storing a processor-executable program, wherein when the processor executes the program stored in the memory, it implements the large-scale structural reanalysis method described above.

[0016] The fourth objective of this invention can be achieved by adopting the following technical solution:

[0017] A storage medium storing a program that, when executed by a processor, implements the aforementioned large-scale structural reanalysis method.

[0018] The embodiments of the present invention have the following advantages over the prior art:

[0019] (1) Applicability: This invention achieves high computational efficiency and solution accuracy in both global and local modification scenarios of structural reanalysis. Within the separation-reconstruction framework, only the subdomains related to the modified region need to be calculated specifically, thereby making full use of the local modification characteristics to complete efficient reanalysis;

[0020] (2) Universality: This invention is not only applicable to structural optimization in civil engineering, but can also be extended to fields such as machine learning and artificial intelligence that require rapid solution of linear equations, and has broad application prospects;

[0021] (3) Distributed parallelism: When the computational load of the model is large, the present invention can distribute several independent sub-problems to multiple computers for parallel solution, and then summarize them to obtain the solution of the original model, which significantly improves the computational efficiency. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0023] Figure 1 This is a simplified flowchart of a large-scale structural reanalysis method based on a separation-reconstruction framework according to an embodiment of the present invention.

[0024] Figure 2 This is a flowchart illustrating a large-scale structural reanalysis method based on a separation-reconstruction framework, according to an embodiment of the present invention.

[0025] Figure 3 This is a schematic diagram of a separation-reconstruction framework according to an embodiment of the present invention.

[0026] Figure 4 This is a schematic diagram of a parametric two-dimensional truss structure according to an embodiment of the present invention.

[0027] Figure 5 As an embodiment of the present invention The truss is used as an example to illustrate the domain partitioning scheme.

[0028] Figure 6 This is a schematic diagram illustrating the computation time ratio between a two-dimensional truss reanalysis method at different scales and a traditional method, according to an embodiment of the present invention.

[0029] Figure 7 This is a schematic diagram illustrating the computation time ratio of the SRM method and the PCG method for a two-dimensional truss under different scales according to an embodiment of the present invention.

[0030] Figure 8 This is a structural block diagram of a large-scale structural reanalysis system based on a separation-reconstruction framework according to an embodiment of the present invention. Detailed Implementation

[0031] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] This invention aims to accelerate structural analysis calculations. Existing methods for reanalyzing structures with local modifications suffer from limited efficiency improvements in scenarios involving large-scale structural modifications. Current technologies offer only minor efficiency gains when handling large-scale structural modifications, failing to meet engineering requirements. For example... Figure 3 As shown, this invention improves computational efficiency by decomposing large-scale modifications into a series of local modifications using a separation-reconstruction framework. This method, based on a separation-reconstruction strategy, decomposes the reanalysis problem of large-scale modified structures into solving a series of locally modified structures. The reanalysis method employs system dimensionality reduction and iterative solutions to efficiently solve the local modification problem. Simultaneously, a reconstruction framework for displacement solutions is established based on a high-dimensional model representation, and a correction mechanism is used to eliminate residuals in the stiffness equations, resulting in high-precision displacement solutions. The total computational cost of reanalyzing several locally modified structures is far less than that of reanalyzing the original overall structure, thus significantly improving solution efficiency. The specific implementation scheme is as follows:

[0033] This invention focuses on large-scale engineering structures (specifically, parametric two-dimensional truss building structures). Figure 4 For example, rapid reanalysis can be performed during the structural design phase of multi-story, large-span factory buildings and super high-rise frame-truss hybrid buildings to meet efficiency requirements in numerical simulation analysis, such as material adjustments and partial modifications.

[0034] In practical engineering design iterations (such as optimizing material selection for truss buildings and adjusting the cross-sectional distribution of local components on floors), traditional singular value decomposition (SVD) methods require a complete recalculation of the structural response for each modification, resulting in a massive computational burden and long processing time. This makes numerical simulation analysis extremely slow and unable to support the high-frequency iterative demands of "multiple scheme comparisons and dynamic parameter adjustments." Especially in material iteration scenarios (such as trying high-performance steel or composite material replacements), it is necessary to repeatedly verify structural stiffness and displacement response. Traditional methods, due to computational efficiency bottlenecks, easily lead to extended design cycles. This large-scale structured optimization algorithm based on a separation-reconstruction framework decomposes the numerical simulation process of structural reanalysis into "rapid correction and reanalysis of local subdomains + global displacement mapping and aggregation reconstruction," reducing the calculation time of a single reanalysis to less than 5% of that of traditional methods, thus adapting to the "high-frequency, fast-response" iterative requirements of the design phase.

[0035] The following diagram, along with actual design scenarios, provides a concrete illustration of the solution:

[0036] This method relies on a structural analysis and calculation system, including:

[0037] 1. Subdomain partitioning module (reference) Figure 5 The number of subdomains is preset based on the number of floors in the structure. m Then, the overall structural model is divided according to the number of subdomains to obtain each extraction matrix T. riThis enables data mapping between subdomains and the overall structure, providing a foundation for subsequent local analysis.

[0038] 2. Stiffness Matrix Calculation and Decomposition Module: Based on finite element theory, the overall stiffness matrix K0 is constructed. The singular value decomposition method (numerically stable and suitable for large-scale sparse matrices) is used to quickly decompose the matrix and output the flexibility matrix. , calculation of initial displacement and subdomain compliance.

[0039] 3. Iterative solution and error correction module: Equipped with a stable biconjugate gradient method solver, it combines HDMR (high-dimensional model representation) theory to realize displacement reconstruction, introduces a residual error correction mechanism to eliminate response error, and adapts to the iterative convergence requirements of structural reanalysis.

[0040] The terminology used in this method includes:

[0041]

[0042]

[0043] like Figure 1 As shown, this embodiment provides a large-scale structural reanalysis method based on a separation-reconstruction framework, which includes:

[0044] S101. Divide the truss structure into multiple subdomains and construct an extraction matrix, which is used to transmit displacement and stiffness information between the subdomains and the overall structure.

[0045] S102. Based on the total stiffness matrix and the extracted matrix, construct the flexibility matrix corresponding to the subdomain, and adjust the stiffness of the corresponding subdomain according to the structural design modifications to construct the reduction correction equation for the subdomain.

[0046] S103. The reduction correction equation of the subdomain is solved iteratively using the stable biconjugate gradient method to obtain the displacement correction of each subdomain.

[0047] S104. Map the displacement corrections of each subdomain to the overall structure and reconstruct the displacement response vector of the overall structure.

[0048] S105. Calculate the residual vectors of internal and external forces of the displacement response vector, and determine whether the residual vectors of internal and external forces satisfy the preset convergence condition. If they do, output the structural reanalysis results.

[0049] Further, based on the total stiffness matrix and the extracted matrix, the flexibility matrix corresponding to the subdomain is constructed, including: constructing the total stiffness matrix K0 and obtaining the flexibility matrix using singular value decomposition. The total stiffness matrix K0 represents the initial stiffness matrix of the structure; using the extracted matrix Tri ,pass Obtain the flexibility matrix of the subdomain , i Indicates the first i Subdomains.

[0050] The extraction matrix described above is obtained by combining the degree-of-freedom numbers of the subdomains with the degree-of-freedom numbers of the overall structure. ri The structure is as follows: Let the total number of global degrees of freedom of the original structure be... n , No. i The number of degrees of freedom of the related units in each subdomain is n ri ≤ n Let the first i The set consists of global degree-of-freedom indices of the relevant units in each subdomain. ;but The k OK( k =1,…, n ri )satisfy:

[0051] ;

[0052] In other words, T ri It is a sparse matrix, with each row containing exactly one 1. Its function is to... n 3D global displacement vector Projected onto its subvector at index L .

[0053] Furthermore, based on the structural design modifications, the stiffness of the corresponding subdomains is adjusted, and the reduction maintenance correction equations for the subdomains are constructed, including: calculating the matrix changes resulting from the subdomain stiffness modifications. , Represents the set of stiffness corrections for each subdomain; utilizes the compliance matrix of the subdomain. Construct the coefficient matrix of the subdomain dimension reduction equation , It is an identity matrix, i.e., a matrix with 1s on the diagonal and 0s on the rest; to obtain the displacement components of the subdomain dimension reduction equation. Construct the maintenance reduction positive equation for the subdomain d ri This represents the subdomain displacement correction.

[0054] Furthermore, the displacement corrections of each subdomain are mapped to the overall structure to reconstruct the displacement response vector of the overall structure, including: calculating the first-order terms based on HDMR theory. and through Reconstruct the global displacement vector of the structure. m Indicates the number of subdomains.

[0055] Further, the residual vectors of internal and external forces of the displacement response vector are calculated as follows:

[0056] ;

[0057] Where R is the external load vector of the structure, and K is the corrected stiffness matrix of the structure. This is the global displacement vector.

[0058] Further, after calculating the residual vectors of internal and external forces of the displacement response vector and determining whether the residual vectors of internal and external forces satisfy the preset convergence condition, the method further includes: if not satisfied, then through... Update the residual displacement vector, return to step S103, continue the loop iteration, further solve the displacement increment until the accuracy requirements are met, and finally output accurate structural reanalysis results.

[0059] Furthermore, the number of subdomains is determined based on the number of floors and the subdomain partitioning strategy.

[0060] like Figure 2 As shown, this embodiment provides a large-scale structural reanalysis method based on a separation-reconstruction framework. This large-scale structural reanalysis method takes a large-span truss office building (number of spans) as an example. Number of floors Taking the example of evaluating the structural response under locally added equipment loads, the parameter values ​​are strongly correlated with the engineering scenario:

[0061] (1) Preprocessing stage (building a basic model for subsequent analysis);

[0062] Step 1.1: Subdomain partitioning and transformation matrix construction.

[0063] In this step, based on the number of subdomains m =50 (meaning every 4 levels are divided into 1 subdomain) =4), satisfying Divide the truss structure and generate the extraction matrix T. ri This matrix is ​​used to convey displacement and stiffness information between the subdomain and the overall structure. For example, it maps the deformation data of members within the subdomain to the global coordinate system, laying the foundation for analyzing the impact of the subdomain on the overall structure.

[0064] Step 1.2: Construction of the overall stiffness matrix and calculation of initial displacement.

[0065] In this step, the overall stiffness matrix K0 is constructed (in this scenario, it is determined by the cross-sectional area of ​​the truss members, A = 20 cm²). 2 And the material's elastic modulus E0 = 20000 kN / cm 2 (The stiffness matrix is ​​obtained by assembling finite element rod elements), and the flexibility matrix is ​​obtained by singular value decomposition. ;in accordance with (R is the external force vector. In this scenario, the resultant force of the original structural dead load and the load of the newly added equipment is included. For example, the horizontal load P=20kN in R acts on the left node.) Calculate the initial displacement solution d0 to provide an initial benchmark for iterative analysis.

[0066] Parameter description: The dimension of K0 is... ( (i.e., structural degrees of freedom); R is... Vectors, load values ​​matched with building usage scenarios (such as standard values ​​for office loads).

[0067] Step 1.3: Construction of Subdomain Compliance Matrix

[0068] In this step, the transformation matrix T is used. ri ,pass Obtain the compliance matrix of each subdomain .

[0069] (2) Dimension reduction-iterative solution stage (fast solution of subdomain response);

[0070] Step 2.1: Subdomain stiffness correction and coefficient matrix construction.

[0071] In this step, the matrix change for subdomain element stiffness modification is calculated. (In this scenario, For the first i Within a subdomain, the equivalent change in member stiffness caused by the additional equipment load on a local floor is determined. For example, if the elastic modulus of a member within this subdomain increases from E0 = 20000 kN / cm² due to an increase in load, then... 2 Adjusted to E=21000kN / cm 2 The subdomain stiffness is updated using the rod element stiffness formula; (This is the set of stiffness corrections for each subdomain); based on this, the coefficient matrix of the subdomain dimension-reduced equation is constructed. By associating subdomain characteristics with dimensionality reduction calculations, efficient mapping of local modifications can be achieved.

[0072] Parameter description (scene binding): The stiffness effect of structural modifications (such as load and material changes) within a subdomain is reflected in this scenario by quantifying the change in the material's elastic modulus. The dimension is matched with the degree of freedom of the subdomain to compress the computational scale of the subdomain.

[0073] Step 2.2: Displacement component update and iterative convergence.

[0074] In this step, the displacement components of the subdomain dimension reduction equation are obtained. Constructing dimensionality reduction equations for subdomains The subdomain displacement correction d is obtained by using the stable double conjugate gradient method. ri and through i From 1 to m Loop (in this scenario) m =50, meaning traversing 50 subdomains), gradually obtaining the response of each subdomain, and finally outputting the displacement d of each subdomain. r1 d r2 、…、d rm .

[0075] (3) Reconstruction of approximate solution stage (global displacement correction and accuracy verification).

[0076] Step 3.1: Displacement reconstruction and residual vector calculation.

[0077] In this step, the composition of the first-order terms is calculated based on HDMR theory. and through Reconstruct the global displacement vector of the structure; simultaneously calculate the residual vectors of internal and external forces. Assess the accuracy of the analysis (e.g., in this scenario, the ratio of the residual vector to the external force vector is required). ).

[0078] Step 3.2: Accuracy assessment and residual displacement update.

[0079] In this step, if If true, output the displacement result. If not satisfied, proceed through Update the residual displacement vector and substitute it into step 2.2 to continue iterating and solving for the displacement increment until the accuracy requirements are met. Finally, output accurate structural reanalysis results.

[0080] Figure 5 This invention demonstrates the subdomain partitioning logic of the Separate-Reconstruction (SRM) strategy in this application: the overall structure is decomposed into several subdomains based on the number of floors, and each subdomain contains... Layer; through The number of associated subdomains m and the floor parameter, in order to Taking a truss as an example, when m=2, When m=3, This presents partitioning schemes for different numbers of subdomains.

[0081] Quantify the efficiency differences between different methods using bar charts and line graphs:

[0082] Figure 6 As the scale of the structure increases, SRM and PCG become more effective compared to traditional methods. The change in the proportion of computation time highlights the efficiency advantage of SRM in large-scale scenarios;

[0083] Figure 7 Comparing the computation time ratios of SRM and PCG, it is verified that the average computation time of SRM is only 35% of that of PCG, and the efficiency advantage remains stable with changes in structure size, intuitively demonstrating the engineering application value of the algorithm.

[0084] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware, and the corresponding program can be stored in a computer-readable storage medium.

[0085] It should be noted that although the method operations of the above embodiments are described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. On the contrary, the order of execution of the described steps may be changed. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0086] like Figure 8 As shown, this embodiment provides a large-scale structural reanalysis system based on a separation-reconstruction framework. This large-scale structural reanalysis system includes:

[0087] The first construction unit 801 is used to divide the truss structure into multiple subdomains and construct an extraction matrix, which is used to transmit displacement and stiffness information between the subdomains and the overall structure.

[0088] The second construction unit 802 is used to construct the flexibility matrix corresponding to the subdomain based on the total stiffness matrix and the extracted matrix, and to correct the stiffness of the corresponding subdomain according to the structural design modification, and to construct the reduction correction equation of the subdomain.

[0089] The solver unit 803 is used to iteratively solve the reduction correction equation of the subdomain according to the stable biconjugate gradient method to obtain the displacement correction of each subdomain.

[0090] The reconstruction unit 804 is used to map the displacement correction of each subdomain to the overall structure and reconstruct the displacement response vector of the overall structure.

[0091] Output unit 805 is used to calculate the residual vectors of internal and external forces of the displacement response vector and determine whether the residual vectors of internal and external forces meet the preset convergence conditions. If they do, the structural reanalysis results are output.

[0092] This embodiment provides a computer device including a processor, a memory, an input device, a display device, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. When the computer programs are executed by the processor, they implement the large-scale structural reanalysis method described in the above embodiment.

[0093] This embodiment provides a storage medium, which is a computer-readable storage medium, storing a computer program. When the computer program is executed by a processor, it implements the large-scale structural reanalysis method for truss structures described in the above embodiment.

[0094] It should be noted that the computer-readable storage medium in this embodiment can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0095] In this embodiment, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this embodiment, the computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying a computer-readable program. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The computer-readable signal medium can also be any computer-readable storage medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The computer program contained on the computer-readable storage medium can be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination thereof.

[0096] The computer-readable storage medium described above can be used to write computer programs for executing this embodiment in one or more programming languages ​​or combinations thereof. These programming languages ​​include object-oriented programming languages—such as Java, Python, and C++—and conventional procedural programming languages—such as C or similar programming languages. The program can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0097] In summary, for structural reanalysis scenarios involving large-scale / global modifications, this invention employs a separation-reconstruction framework to decompose the overall modification problem into several local modification subproblems, which are then solved separately, significantly reducing the scale of the equation solution. Combined with the efficient iteration of the stable biconjugate gradient method, computation time is further compressed. Subsequently, the results of each subproblem are reconstructed to obtain an approximate solution for the overall modification, and a residual correction mechanism is introduced to eliminate truncation errors caused by omitting higher-order terms during reconstruction, ensuring that the displacement solution meets engineering accuracy requirements under large-scale modifications or fine-tuning of local parameters.

[0098] This method is suitable for parametric modification analysis of large-scale structures and is also compatible with complex scenarios such as local damage and nonlinear iteration, covering the needs of multiple fields such as architecture, machinery, and aerospace. It helps engineers shorten the cycle time while improving the reliability of design iteration and performance evaluation, indirectly reducing R&D costs and enhancing the mechanical robustness of products. This invention fully utilizes hardware computing power to improve the efficiency of overall structural reanalysis, providing a new and efficient tool for structural engineering analysis, and has significant theoretical implications.

[0099] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope disclosed in the present invention, based on the technical solution and inventive concept of the present invention, shall fall within the scope of protection of the present invention.

Claims

1. A large-scale structure reanalysis method based on a separation-reconstruction framework, characterized in that, include: The truss structure is divided into multiple subdomains, and an extraction matrix is ​​constructed. The extraction matrix is ​​used to transmit displacement and stiffness information between the subdomains and the overall structure. Based on the total stiffness matrix and the extracted matrix, construct the flexibility matrix corresponding to the subdomain, and adjust the stiffness of the corresponding subdomain according to the structural design modifications to construct the reduction maintenance correction equation for the subdomain. The reduction maintenance equation of the subdomain is solved iteratively using the stable biconjugate gradient method to obtain the displacement correction of each subdomain. The displacement corrections of each subdomain are mapped to the overall structure to reconstruct the displacement response vector of the overall structure. Calculate the residual vectors of internal and external forces of the displacement response vector, and determine whether the residual vectors of internal and external forces satisfy the preset convergence condition. If they do, output the structural reanalysis results.

2. The method of large-scale structural reanalysis of claim 1, wherein, Based on the total stiffness matrix and the extracted matrix, construct the compliance matrix corresponding to the subdomain, including: A total stiffness matrix K0 is constructed, and a flexibility matrix is obtained by singular value decomposition method , the total stiffness matrix K0 represents a structure initial stiffness matrix; Using the extraction matrix T ri , by obtaining the flexibility matrix of the subdomain , i denotes the i subdomain.

3. The large-scale structural reanalysis method according to claim 2, characterized in that, Based on the structural design modifications, the stiffness of the corresponding subdomains is adjusted, and the reduction maintenance correction equations for the subdomains are constructed, including: Calculate the matrix change for subdomain stiffness modification , This represents the set of stiffness corrections for each subdomain; Using the flexibility matrix of the subdomain Construct the coefficient matrix of the subdomain dimension reduction equation , It is the identity matrix; Obtain the displacement components of the subdomain dimension reduction equation Construct the maintenance reduction positive equation for the subdomain d ri This represents the subdomain displacement correction.

4. The large-scale structural reanalysis method according to claim 3, characterized in that, Mapping the displacement corrections of each subdomain to the overall structure, and reconstructing the displacement response vector of the overall structure, includes: Based on HDMR theory, the composition of first-order terms is calculated. and through Reconstruct the global displacement vector of the structure. m Indicates the number of subdomains.

5. The large-scale structural reanalysis method according to claim 4, characterized in that, The residual vectors of internal and external forces in the displacement response vector are calculated as follows: ； Where R is the external load vector of the structure, and K is the corrected stiffness matrix of the structure. This is the global displacement vector.

6. The large-scale structural reanalysis method according to claim 5, characterized in that, After calculating the residual vectors of internal and external forces of the displacement response vector and determining whether the residual vectors of internal and external forces satisfy the preset convergence condition, the method further includes: If not satisfied, then proceed. Update the residual displacement vector, return to the loop iteration operation, further solve for the displacement increment until the accuracy requirements are met, and finally output accurate structural reanalysis results.

7. The large-scale structural reanalysis method according to claim 1, characterized in that, The number of subdomains is determined based on the number of floors and the subdomain partitioning strategy.

8. A large-scale structural reanalysis system based on a separation-reconstruction framework, characterized in that, include: The first construction unit is used to divide the truss structure into multiple subdomains and construct an extraction matrix, which is used to transmit displacement and stiffness information between the subdomains and the overall structure. The second construction unit is used to construct the flexibility matrix corresponding to the subdomain based on the total stiffness matrix and the extracted matrix, and to correct the stiffness of the corresponding subdomain according to the structural design modification, and to construct the reduction correction equation of the subdomain. The solution unit is used to iteratively solve the reduction correction equation of the subdomain according to the stable biconjugate gradient method to obtain the displacement correction of each subdomain. The reconstruction unit is used to map the displacement correction of each subdomain to the overall structure and reconstruct the displacement response vector of the overall structure. The output unit is used to calculate the residual vectors of internal and external forces of the displacement response vector and determine whether the residual vectors of internal and external forces satisfy the preset convergence condition. If they satisfy the condition, the structural reanalysis result is output.

9. A computer device comprising a processor and a memory for storing a processor-executable program, characterized in that, When the processor executes the program stored in the memory, it implements the large-scale structural reanalysis method according to any one of claims 1-7.

10. A storage medium storing a program, characterized in that, When the program is executed by the processor, it implements the large-scale structural reanalysis method according to any one of claims 1-7.

Images