A modular structure optimization design method and device

Abstract

This application discloses a modular structure optimization design method and apparatus, relating to the fields of modular structure modeling and modular structuring algorithms. The method includes: constructing a modular structural system based on a target structure and decomposing it into multiple isomorphic subsystems and a residual system; performing finite element analysis on the isomorphic subsystems to construct a hyperelement matrix; transforming the isomorphic subsystems from physical coordinates to modal coordinates and constructing a modular virtual framework based on the hyperelement matrix; initializing the design variables of the modular virtual framework and introducing a Bayesian optimization method to construct a Bayesian surrogate model; training the Bayesian surrogate model until a preset stopping condition is met to obtain optimization parameters; and updating the design variables according to the optimization parameters to obtain the optimized modular structural system. This method achieves efficient optimization design of multi-dimensional modular structures while ensuring the accuracy of the calculation results.

Description

Technical Field

[0001] This application relates to the field of modular structure modeling and modular structuring algorithm technology, and in particular to a modular structure optimization design method and apparatus. Background Technology

[0002] As the complexity of space systems continues to increase, there are higher demands on the reconfiguration design of modular structures. Traditional modular designs in space systems include components such as antennas, solar arrays, and optical mirrors, which are typically composed of two-dimensional modules. Existing three-dimensional modules are mostly composed of cylindrical modules and cubic connecting nodes. Modular structures, with their advantages of high redundancy, robustness, and maintainability, have demonstrated significant potential in meeting the performance and economic requirements of spacecraft structures in complex space environments.

[0003] In the aerospace field, parametric reconfiguration modeling methods for modular structures are used for the rapid design and manufacture of complex products such as aircraft and rockets. Existing optimization designs for modular structures primarily rely on single model dimensionality reduction techniques or efficient optimization algorithms, introducing parameterization at the finite element level to simultaneously optimize the topology of standard modules and the joint configuration of interconnected modules. While traditional methods can perform structural optimization, they have limitations in optimization effectiveness and application scenarios, posing challenges to the accurate evaluation of system dynamic parameters and leading to a dramatic increase in computational costs due to the increased number of modules in the optimization design of modular reconfigurable structures. Maximizing efficiency and robustness in the optimization design of high-dimensional modular structures remains an ongoing exploration. Furthermore, traditional optimization methods are limited in scope, time-consuming, and struggle to guarantee the accuracy of calculation results. Summary of the Invention

[0004] This application provides a modular structure optimization design method and apparatus, solving the problems of complex modular structures and drastically increased computational costs caused by increasing the number of modules in existing modular structure optimization design methods. By using a component modal synthesis method, a virtual framework at the modular structure matrix level is established, enabling rapid parameterized reconfiguration of multi-dimensional modular structures. By using Gaussian processes and / or random forest algorithms as surrogate models for Bayesian optimization, and introducing a material-free computational model with empty positions and a hyperparameter dimensionality reduction method, the computational efficiency and adaptability of the system for solving high-dimensional problems are improved.

[0005] In a first aspect, embodiments of this application provide a modular structure optimization design method, comprising: constructing a modular structural system based on a target structure, and decomposing it into multiple isomorphic subsystems and a residual system; performing finite element analysis on the isomorphic subsystems to construct a super-element matrix; converting the isomorphic subsystems from physical coordinates to modal coordinates, and constructing a modular virtual framework based on the super-element matrix; initializing the design variables of the modular virtual framework, and introducing a Bayesian optimization method to construct a Bayesian surrogate model; training the Bayesian surrogate model until a preset stopping condition is met to obtain optimization parameters; and updating the design variables according to the optimization parameters to obtain an optimized modular structural system.

[0006] In one possible implementation, the step of performing finite element analysis on the isomorphic subsystem and constructing a super-element matrix includes: constructing a finite element model of the isomorphic subsystem; performing finite element analysis on the finite element model of the isomorphic subsystem and extracting local feature matrices of the finite element model of the isomorphic subsystem; wherein the local feature matrices include a stiffness matrix and a mass matrix; setting the degrees of freedom of the nodes in the finite element model of the isomorphic subsystem, and reducing the dimensionality of the local feature matrices based on the degrees of freedom to obtain a reduced local matrix; and assembling the reduced local matrices of the finite element models of each isomorphic subsystem to obtain the super-element matrix.

[0007] In one possible implementation, the construction of the modular virtual framework based on the super-element matrix includes: introducing interface displacement coordination and force balance conditions in the isomorphic subsystems in physical coordinates; determining the finite element model functions and boundaries of each isomorphic subsystem and defining them as independent modules; and establishing connections between the independent modules to construct the modular virtual framework.

[0008] In one possible implementation, the design variables include generalized density variables, module rotation variables, and module type variables.

[0009] In one possible implementation, after the design variables, the method further includes: constructing a comprehensive generalized density design variable; wherein the comprehensive generalized density design variable includes the generalized density variable for the candidate location and the module type variable; and using the comprehensive generalized density design variable to reduce the dimensionality of the design variables.

[0010] In one possible implementation, the dimensionality reduction of the design variables using the comprehensive generalized density design variables includes:

[0011] ,

[0012] ,

[0013] ;

[0014] In the formula, This indicates the search for design variables after dimensionality reduction. This represents the comprehensive generalized density design variables. Represents the module rotation variable. Represents the transpose matrix. Represents a module type variable. Describe the objective function. Indicates the number of candidate positions. This represents the total number of randomly placed candidate modules. This indicates a constraint on the number of modules to be selected. express Dimensional constraints in the axial direction Indicates the first Constraints on the number of candidate modules on the layer and These represent the upper and lower bound constraints of the comprehensive generalized density design variables, respectively. Indicates the first A comprehensive generalized density design variable, Indicates the first One rotational variable, and These represent the upper and lower limits of the module rotation angle, respectively.

[0015] In one possible implementation, the Bayesian surrogate model includes: a Bayesian surrogate model based on Gaussian processes and a Bayesian surrogate model based on random forests.

[0016] Secondly, embodiments of this application provide an apparatus for modular structure optimization design, comprising: a decomposition module for constructing a modular structure system based on a target structure and decomposing it into multiple isomorphic subsystems and a residual system; a reorganization module for performing finite element analysis on the isomorphic subsystems to construct a super-element matrix; converting the isomorphic subsystems from physical coordinates to modal coordinates, and constructing a modular virtual framework based on the super-element matrix; a structure optimization module for initializing the design variables of the modular virtual framework, introducing a Bayesian optimization method to construct a Bayesian surrogate model; training the Bayesian surrogate model until a preset stopping condition is met to obtain optimization parameters; and a reconstruction module for updating the design variables according to the optimization parameters to obtain an optimized modular structure system.

[0017] Thirdly, embodiments of this application provide an apparatus for executing a modular structure optimization design method, the apparatus comprising: a processor; a memory for storing processor-executable instructions; wherein, when the processor executes the executable instructions, it implements the method as described in the first aspect or any possible implementation of the first aspect.

[0018] In a fourth aspect, an embodiment of the present application provides a non-volatile computer-readable storage medium, which includes a device for storing a computer program or instruction, and when the computer program or instruction is executed, the method described in the first aspect or any possible implementation method of the first aspect is implemented.

[0019] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0020] This application employs a modular structure optimization design method and apparatus. By performing finite element analysis on the modular structure system, a modular virtual framework is constructed to characterize its modular structure, enabling rapid multi-dimensional module reconfiguration. Based on the Bayesian optimization method, design variables are initialized, and hyperparameter space dimensionality reduction is performed to reduce the difficulty of constructing the Bayesian surrogate model. Through hyperparameter dimensionality reduction, the Bayesian surrogate model is trained, enabling the redesigned optimized modular aerospace structure system to more efficiently and accurately complete the modular design of large structures. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 A flowchart illustrating a modular structure optimization design method provided in this application embodiment;

[0023] Figure 2 A schematic diagram illustrating the reconstruction of a two-dimensional modular virtual framework provided in an embodiment of this application;

[0024] Figure 3 A schematic diagram of the optimization design space of the Bayesian surrogate model based on Gaussian processes provided in the embodiments of this application;

[0025] Figure 4 A schematic diagram of the initial layout of the Bayesian surrogate model based on Gaussian processes provided in an embodiment of this application;

[0026] Figure 5 A schematic diagram of the optimized layout of the Bayesian surrogate model based on Gaussian processes provided in the embodiments of this application;

[0027] Figure 6 A schematic diagram of the three-dimensional linear optimization design space of the Bayesian surrogate model based on random forest provided in the embodiments of this application;

[0028] Figure 7 A schematic diagram of the three-dimensional linear initial layout of the Bayesian surrogate model based on random forest provided in the embodiments of this application;

[0029] Figure 8 A schematic diagram of a three-dimensional linear optimization layout of a Bayesian surrogate model based on random forest provided in an embodiment of this application;

[0030] Figure 9 A schematic diagram of the three-dimensional nonlinear optimization design space of the Bayesian surrogate model based on random forest provided in the embodiments of this application;

[0031] Figure 10 A schematic diagram of the three-dimensional nonlinear initial layout of the Bayesian surrogate model based on random forest provided in the embodiments of this application;

[0032] Figure 11 A schematic diagram of a three-dimensional nonlinear optimization layout of a Bayesian surrogate model based on random forest provided in an embodiment of this application;

[0033] Figure 12 A schematic diagram illustrating the optimization results of a Bayesian surrogate model based on a Gaussian process, provided in an embodiment of this application;

[0034] Figure 13 A schematic diagram of the three-dimensional linear optimization results of the Bayesian surrogate model based on random forest provided in the embodiments of this application;

[0035] Figure 14 A schematic diagram of the three-dimensional nonlinear optimization results of the Bayesian surrogate model based on random forest provided in the embodiments of this application;

[0036] Figure 15 This is a schematic diagram of a modular structure optimization design device provided in an embodiment of this application. Detailed Implementation

[0037] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0038] The following description of some of the technologies involved in the embodiments of this application is provided to facilitate understanding and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of this application. Similarly, for the sake of clarity and conciseness, some descriptions of well-known functions and structures are omitted from the following description.

[0039] Figure 1 This is a flowchart of a modular structure optimization design method provided in an embodiment of this application, including steps 101 to 106. Figure 1 This is merely one execution order shown in the embodiments of this application and does not represent the only execution order of the modular structure optimization design method. Where the final result can be achieved, Figure 1 The steps shown can be performed in parallel or in reverse order, as detailed below.

[0040] Step 101: Construct a modular structural system based on the target structure, and decompose it into multiple isomorphic subsystems and a residual system. In this embodiment, the target structure is a large-scale modular aerospace structure, and the density of all materials is defined as... The Young's modulus is defined as 71.7 GPa, and the Poisson's ratio as 0.33. A modular aerospace structure system is constructed by combining modular design and parametric modeling methods. The modular aerospace structure system is decomposed into multiple modules with specific functions using finite element analysis, and these modules are rapidly designed and reassembled using parametric modeling techniques, thereby achieving rapid customization and optimization of the target aerospace structure, as detailed below.

[0041] ;

[0042] in, This indicates a modular aerospace structure system. Indicates isomorphic subsystems, Indicates the remaining system, This represents the total number of isomorphic subsystems and the remaining systems.

[0043] Specifically, the modular aerospace structure system is defined as According to the system Decompose the characteristics of the system The internal components of each individual module are highly interdependent and can fully perform specific functions or logic. Due to the low coupling between modules, module independence and system stability are improved. Flexibility.

[0044] Step 102: Perform finite element analysis on the isomorphic subsystem and construct the superelement matrix. In this embodiment, a finite element model of the isomorphic subsystem is constructed based on the component modal synthesis method (CMS). By performing finite element analysis on the finite element model, feature extraction is performed in the isomorphic subsystem to obtain the local feature matrix of each finite element model, thereby analyzing the entire system. The local feature matrix includes the stiffness matrix and the mass matrix.

[0045] Specifically, finite element analysis reduces data storage space by extracting large sparse matrices, and can be performed using finite element software such as Abaqus, Nastran, and Ansys Mechanical. These software programs offer rich libraries of elements and material models, enabling the simulation of various complex engineering problems. They also provide powerful pre- and post-processing capabilities, facilitating subsequent model building, result analysis, and optimization design.

[0046] Based on system The modular construction of isomorphic subsystems uses a finite element model to simulate the actual modular structure with a finite number of polyhedra, reflecting an approximation of the real physical system. This model can be used as a system... The boundary conditions are determined. After defining the superelement and boundary nodes, modal coordinate transformation is performed on the boundary nodes of the isomorphic subsystem. The local characteristic matrix is ​​then reduced to a reduced local matrix by structural degree-of-freedom reduction. The reduced local matrices of the finite element models of each isomorphic subsystem are assembled to obtain the superelement matrix. This process primarily reduces the degrees of freedom of internal component nodes to the boundary nodes, thereby reducing the size of the local matrix. This reduces computational load while maintaining sufficient accuracy and improving computational efficiency. After constructing the superelement matrix, it can be solved using finite element solvers such as Optistruct, FreeFEM, and Elmer.

[0047] Based on the hyperelement matrix, the undamped discretized branch principal modes of the isomorphic subsystem are obtained to... The system performs the following analysis.

[0048] ;

[0049] In the formula, Represents the stiffness matrix. The eigenvalues ​​of the dynamic equations of isomorphic subsystems are represented. Represents the mass matrix, This represents the undamped discretized branched principal mode.

[0050] Among them, the undamped discretized branch principal mode is the system The system exhibits inherent vibration characteristics independent of external excitation. Modal analysis can fully describe its vibration characteristics. The overall dynamic characteristics of the system. The space formed by the modal vectors of each order is defined as the principal space, and its corresponding modal coordinates are the principal coordinates. In modal analysis, the continuous system is considered... Discretization into isomorphic subsystems and residual systems with a finite number of degrees of freedom preserves the system's... The main vibration characteristics are analyzed to facilitate numerical calculations. In each discretized system, each degree of freedom corresponds to a mode, which are arranged in ascending order of frequency. Furthermore, the principal modes are orthogonal in their N-dimensional principal space, meaning the vibrations of different modes are independent, ensuring the system's stability. Low coupling.

[0051] Step 103: Transform the isomorphic subsystem from physical coordinates to modal coordinates, and construct a modular virtual framework based on the hyperelement matrix. For example, constructing the modular virtual framework includes: converting the physical coordinates of the boundary nodes of the isomorphic subsystem into modal coordinates, specifically as follows.

[0052] ;

[0053] Where, Represents the physical coordinates of the isomorphic subsystem. This represents the undamped discretized principal modes. Represents the modal coordinates of the isomorphic subsystem.

[0054] Based on the physical coordinate design interface displacement coordination and force balance conditions, the functions and boundaries of the finite element models of each isomorphic subsystem are determined and defined as independent modules, resulting in the dynamic equations in the global coordinate system, as follows.

[0055] ,

[0056] ;

[0057] Where, and Representing isomorphic subsystems and Interface degrees of freedom in modal coordinate transformation and Representing isomorphic subsystems and Interfacial forces during modal coordinate transformation.

[0058] For example, the dynamic equations of isomorphic subsystem 1 in the global coordinate system of isomorphic subsystem 2 are as follows.

[0059] ;

[0060] Where, This represents the mass matrix of isomorphic subsystem 1. This represents the stiffness matrix of isomorphic subsystem 1. Represents modal coordinates, This represents the displacement of isomorphic subsystem 2.

[0061] The stiffness matrix and mass matrix, composed of modal coordinates, constitute a modular virtual framework. Specifically, based on the hyper-unit node relationships of the isomorphic subsystem, the independent modules of the isomorphic subsystem are connected through matrix transformations to obtain the isomorphic subsystem matrix. The system is then implemented using matrix transformations. The secondary layout yields a recombined modular aerospace structure system, i.e., the new system. Based on the assembly of the isomorphic subsystem matrix, the reconstructed dynamic equations of the isomorphic subsystem are represented by independent modal coordinates, as follows.

[0062] ;

[0063] in, Represents the mass matrix, Represents the stiffness matrix. Represents the reconstructed modal coordinates. Represents independent modal coordinates.

[0064] When the isomorphic subsystem undergoes translational motion, its stiffness and mass matrices remain unchanged, in order to improve the performance of the new system. Design flexibility, incorporating coordinate transformation matrices We rotate the material and process the resulting stiffness and mass matrices as follows.

[0065] ,

[0066] ;

[0067] Where, and indicates The stiffness matrix and mass matrix after rotation, respectively Represents the coordinate transformation matrix. and system representation The original stiffness matrix and mass matrix, Indicates the rotation angle.

[0068] Among them, such as Figure 2 The diagram shown is a schematic representation of the reconstruction of a two-dimensional modular virtual framework. Indicates the first The mass matrix of modal coordinates Indicates the first Stiffness matrix of each modal coordinate, Furthermore, compared with existing technologies, this application can extend the two-dimensional space to the three-dimensional linear and nonlinear design space through the super-element method of component modal synthesis, realizing the rapid reassembly of new systems within a virtual framework, with a wider range of applications and more flexible implementation methods.

[0069] Step 104: Initialize the design variables of the modular virtual framework and introduce Bayesian optimization to construct a Bayesian surrogate model. Specifically, initialize the design variables, which include generalized density design variables, module rotation variables, and module type variables. Within the modular virtual framework, the design variables are as follows.

[0070] ;

[0071] Where, Represents the generalized density design variable. Represents rotation variable, Represents a module type variable. Represents the transpose matrix. Indicates the number of candidate positions. This represents the total number of randomly placed candidate modules.

[0072] The desired reconfiguration pattern is generated by controlling the generalized density design variables at the candidate locations, the rotation variables of the modules at the candidate locations, and the module type variables.

[0073] By introducing the Bayesian optimization method, the optimization objective, design constraints, and number of iterations are set. In this embodiment, the design variables can be a set of hyperparameter combinations, and the final transpose matrix of the optimization result is the hyperparameter combination, i.e., the design variable combination. In high-dimensional space, data sparsity increases, sampling search becomes very time-consuming and difficult, and requires more computational resources. As the complexity of the problem increases, the surrogate modeling of the hyperparameter space, i.e., the design variable space, also becomes more difficult, greatly increasing the uncertainty of the entire system. To avoid the curse of size, the design variable space is reduced in dimensionality according to the characteristics of modular structural layout design. A comprehensive design variable is proposed, which includes generalized density variables of candidate locations and module type variables. To avoid the computational cost brought by high-dimensional space, this embodiment reduces the dimensionality of the design variable space to obtain the design variable combination, as follows.

[0074] ,

[0075] ,

[0076] ;

[0077] Where, This indicates the search for design variables after dimensionality reduction. This represents the comprehensive generalized density design variables. Represents the module rotation variable. Represents the transpose matrix. Represents a module type variable. Describe the objective function. Indicates the number of candidate positions. This represents the total number of randomly placed candidate modules. This indicates a constraint on the number of modules to be selected. express Dimensional constraints in the axial direction Indicates the first Constraints on the number of candidate modules on the layer and These represent the upper and lower bound constraints of the comprehensive generalized density design variables, respectively. Indicates the first A comprehensive generalized density design variable, Indicates the first One rotational variable, and These represent the upper and lower limits of the module rotation angle, respectively.

[0078] This includes design variables describing the selection results of candidate locations and generalized density design variables describing the selected locations of modules. Updated to a comprehensive generalized density design variable, i.e. By reducing dimensionality, the dimension of the design variable space is reduced from the original... Dimensions dropped to This effectively reduces the dimensionality of the design variable space.

[0079] The objective function of the Bayesian optimization method is the system The problem of maximizing intrinsic frequency is transformed into a minimization problem by taking its inverse. In the embodiments of this application, the constructed Bayesian surrogate models are a Bayesian surrogate model based on Gaussian processes and a Bayesian surrogate model based on random forests.

[0080] Step 105: Train the Bayesian surrogate model until the preset stopping condition is met to obtain the optimized parameters. For example, a Gaussian process is selected as the Gaussian surrogate model for Bayesian optimization to perform optimization space design on the two-dimensional modular structure. The maximum number of iterations is set to 1000, and the module size is set to 50mm × 50mm. The design space is as follows: Figure 3 As shown, the initial layout is as follows Figure 4 As shown, a solution approaching the global optimum is obtained within the maximum number of iterations, and the optimized layout is as follows. Figure 5 As shown, the module has a maximum degree of freedom of 4. A prediction following a normal distribution is defined as follows.

[0081] ;in,

[0082] ;

[0083] ;

[0084] ,

[0085] ;

[0086] In the formula, This represents the prediction of a normal distribution using a Bayesian surrogate model based on Gaussian processes. Indicates the target value. Indicates a sample, Represents the sampled dataset. Represents the total number of samples. Indicates a normal distribution. Represents the mean function, Represents the variance function. Represents the sample point set, Indicates the first One sample, This indicates scalar output.

[0087] For example, a probabilistic regression-based random forest algorithm is introduced as a Bayesian optimization random forest surrogate model to optimize the space design of a three-dimensional modular structure. The maximum number of iterations for the three-dimensional linear modular structure is 200, and the module size is set to 50mm × 50mm × 100mm. The design space is as follows: Figure 6 As shown, the initial layout is as follows Figure 7 As shown, the optimized layout is as follows: Figure 8 As shown, the maximum degree of freedom for the module is 8. The maximum number of iterations for the three-dimensional nonlinear modular structure is 100. The outer diameter of the module's outer circle is set to 50mm. The design space is as follows: Figure 9 As shown, the initial layout is as follows Figure 10 As shown, the optimized layout is as follows: Figure 11 As shown, the maximum degree of freedom for the modules is 16. The initial layout contains four types of constrained modules, with three modules of each type, randomly distributed within the design space. The initial natural frequency of the three-dimensional nonlinear modular structure is 64.47 Hz, and the optimized natural frequency is 73.05 Hz, an increase of 13.31%. While the differences in center of gravity, mass, and fundamental frequency among the four types of modules are small, the specific distribution trend of the optimized layout satisfies the general rule that an increase in the moment of inertia of the cross-section leads to an overall increase in the system's center of gravity towards the constrained end, thus improving the structural dynamics.

[0088] A Gaussian distribution is pre-defined to construct the random forest prediction distribution. For example, as shown below... ,in, and You can select the regression values ​​from the random forest regression tree set.

[0089] Specifically, the Bayesian surrogate model based on Gaussian processes and the Bayesian surrogate model based on random forests approach optimization design from the perspective of the problem, escaping local optima in the fewest iterations, and constructing a modular optimization design method.

[0090] Furthermore, the Bayesian surrogate model is optimized using a material-free computational model with empty locations and a hyperparameter dimensionality reduction method. A scaling factor is introduced into the hyperparameter dimensionality reduction method to scale the hypercell matrix and obtain the optimized hypercell matrix.

[0091] Step 106: Update the design variables according to the optimization parameters to obtain the optimized modular structure system.

[0092] In this embodiment, the computational model for empty locations without material is an empty matrix that does not consider empty locations in numerical calculations. Meanwhile, to avoid numerical singularities caused by the empty matrix, a scaling factor is introduced to scale the superelement matrix and update the optimization design variables for empty locations without material. Specifically, the superelement matrix is ​​scaled as follows.

[0093] ,

[0094] ;

[0095] In the formula, and This represents the scaled optimized supercell matrix. and Indicates the new system The initial superunit matrix, This represents the scaling factor, set to 0.1. This represents a constant, which is 3.

[0096] In this approach, by scaling the hyperelement matrix, the design variables for locations without material are brought close to zero. As the design space dimension increases, this embodiment employs a computational model for empty locations without material. In the numerical calculations of structural analysis, the empty matrix at empty locations is not considered, and the design variables are updated and optimized, which can reduce the need for new systems. The error was reduced, computational efficiency was significantly improved, and the new system was completed. The reconstruction.

[0097] For example, the optimization results of the Bayesian surrogate model based on Gaussian processes are as follows: Figure 12 As shown, the optimized new system The layout shortens along the length direction, while conforming to the general rule of improving the dynamic characteristics of the structure. The linear optimization results of the Bayesian surrogate model based on random forest are as follows: Figure 13 As shown, effectively reducing the dimensionality of the high-dimensional design variable space stabilizes the optimization process while accelerating the modeling process. The increased complexity of module connections also enhances the flexibility of structural reorganization, providing greater design space for exploring optimal structural performance. The nonlinear optimization results based on the Bayesian surrogate model using random forest are shown below. Figure 14 As shown, the highly complex and nonlinear design space is optimized more efficiently, presenting a structural layout of highly flexible modules. In the embodiments of this application, the modular structural optimization design under a complex stack in two-dimensional to three-dimensional space is completed, resulting in an optimized modular aerospace structural system. The optimal hyperparameter combination, i.e., the optimal design variables, for the modular structural layout is given more quickly and accurately.

[0098] Although this application provides method operation steps as described in the embodiments or flowcharts, more or fewer operation steps may be included based on routine or non-creative work. The order of steps listed in this embodiment is only one way of executing the steps among many, and does not represent the only execution order. When an actual device or client product executes, the method shown in this embodiment or the accompanying drawings may be executed sequentially or in parallel (for example, in a parallel processor or multi-threaded processing environment).

[0099] like Figure 15 As shown in the figure, this application embodiment also provides a modular structure optimization design device 1500. The device includes: a decomposition module 1501, a recombination module 1502, a structure optimization module 1503, and a reconstruction module 1504, as detailed below.

[0100] Decomposition module 1501 is used to construct a modular structural system based on the target structure and decompose it into multiple isomorphic subsystems and a residual system.

[0101] Recombination module 1502 is used to perform finite element analysis on isomorphic subsystems and construct a super-element matrix; it transforms the isomorphic subsystem from physical coordinates to modal coordinates and constructs a modular virtual framework based on the super-element matrix.

[0102] The structure optimization module 1503 is used to initialize the design variables of the modular virtual framework, introduce the Bayesian optimization method to construct the Bayesian surrogate model, and train the Bayesian surrogate model until the preset stopping condition is met to obtain the optimization parameters.

[0103] Reconstruction module 1504 is used to update design variables based on optimization parameters to obtain an optimized modular structure system.

[0104] The devices or modules described in the above application embodiments can be implemented by computer chips or physical devices, or by products with certain functions. For ease of description, the above devices are described separately by function in various modules. When implementing the embodiments of this application, the functions of each module can be implemented in the same or multiple software and / or hardware. Of course, a module that implements a certain function can also be implemented by combining multiple sub-modules or sub-units.

[0105] The methods, devices, or modules described herein can be implemented in the form of computer-readable program code. The controller can be implemented in any suitable manner. For example, the controller can take the form of a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro)processor, logic gates, switches, an application-specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320. The memory controller can also be implemented as part of the memory control logic. Those skilled in the art will also appreciate that, in addition to implementing the controller in pure computer-readable program code, the controller can also be implemented in the form of logic gates, switches, an application-specific integrated circuit, a programmable logic controller, an embedded microcontroller, etc. by logically programming the method steps. Therefore, such a controller can be considered a hardware component, and the devices included therein for implementing various functions can also be considered as structures within the hardware component. Or even, the means for implementing various functions may be considered to be both a software module for implementing the method and a structure within a hardware component.

[0106] This application also provides an apparatus for a modular structure optimization design method, the apparatus including: a processor; a memory for storing processor-executable instructions; and, when the processor executes the executable instructions, implementing the method described in this application embodiment.

[0107] The embodiments of the present application also provide a non-volatile computer-readable storage medium having a computer program or instruction stored thereon. When the computer program or instruction is executed, the method described in the embodiments of the present application is implemented.

[0108] Furthermore, in the various embodiments of this application, each functional module can be integrated into one processing module, or each module can exist independently, or two or more modules can be integrated into one module.

[0109] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software and necessary hardware. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product, or it can be embodied in the process of data migration. The computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, mobile terminal, server, or network device, etc.) to execute the methods of various embodiments or some parts of the embodiments of this application.

[0110] The various embodiments described in this specification are presented in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on its differences from other embodiments. All or part of this application can be used in numerous general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, mobile communication terminals, multiprocessor systems, microprocessor-based systems, programmable electronic devices, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices, etc.

[0111] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit the present application. Although the present application has been described in detail with reference to the aforementioned embodiments, a person of ordinary skill in the art should understand that the technical solutions described in the aforementioned embodiments can still be modified, or some or all of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the present application.

Claims

1. A modular structure optimization design method, characterized in that, include: Based on the target structure, a modular structural system is constructed and decomposed into multiple isomorphic subsystems and a residual system. Finite element analysis was performed on the isomorphic subsystem to construct a superelement matrix; The isomorphic subsystem is transformed from physical coordinates to modal coordinates, and a modular virtual framework is constructed based on the superunit matrix; Initialize the design variables of the modular virtual framework, and introduce a Bayesian optimization method to construct a Bayesian proxy model; The Bayesian proxy model is trained until a preset stopping condition is met to obtain optimized parameters; The design variables are updated based on the optimization parameters to obtain the optimized modular structure system; The step of performing finite element analysis on the isomorphic subsystem and constructing a superelement matrix includes: Construct a finite element model of the isomorphic subsystem; Finite element analysis is performed on the finite element model of the isomorphic subsystem to extract the local feature matrix of the finite element model of the isomorphic subsystem; wherein, the local feature matrix includes the stiffness matrix and the mass matrix; The degrees of freedom of the nodes in the finite element model of the isomorphic subsystem are set, and the local feature matrix is ​​reduced in dimensionality based on the degrees of freedom to obtain a reduced local matrix; The reduced local matrices of the finite element models of each of the isomorphic subsystems are assembled to obtain the super-element matrix; The construction of a modular virtual framework based on the super-unit matrix includes: In the isomorphic subsystem under physical coordinates, interface displacement compatibility and force equilibrium conditions are introduced; Determine the function and boundary of the finite element model of each isomorphic subsystem and define them as independent modules; Establish connections between the individual modules to construct the modular virtual framework; The target structure involved is a large-scale modular aerospace structure, and the density of all materials is defined as follows: The Young's modulus is defined as 71.7 GPa, and the Poisson's ratio is defined as 0.

33. By combining modular design and parametric modeling methods, a modular aerospace structure system is constructed. The modular aerospace structure system is decomposed into multiple modules with specific functions using finite element analysis, and these modules are rapidly designed and reassembled using parametric modeling technology. This enables rapid customization and optimization of the target aerospace structure, as detailed below: ； in, This indicates a modular aerospace structure system. Representing isomorphic subsystems, Indicates the remaining system, This represents the total number of isomorphic subsystems and the remaining systems.

2. The method according to claim 1, characterized in that, The design variables include generalized density variables, module rotation variables, and module type variables.

3. The method according to claim 2, characterized in that, Following the design variables, it also includes: Construct a comprehensive generalized density design variable; wherein, the comprehensive generalized density design variable includes the generalized density variable for the candidate location and the module type variable; The design variables are reduced in dimensionality using the comprehensive generalized density design variables.

4. The method according to claim 3, characterized in that, The dimensionality reduction of the design variables using the comprehensive generalized density design variables includes: ， ， ； In the formula, This indicates the search for design variables after dimensionality reduction. This represents the comprehensive generalized density design variables. Represents the module rotation variable. Represents the transpose matrix. Represents a module type variable. Describe the objective function. Indicates the number of candidate positions. This represents the total number of randomly placed candidate modules. This indicates a constraint on the number of modules to be selected. express Dimensional constraints in the axial direction Indicates the first Constraints on the number of candidate modules on the layer and These represent the upper and lower bound constraints of the comprehensive generalized density design variables, respectively. Indicates the first A comprehensive generalized density design variable, Indicates the first One rotational variable, and These represent the upper and lower limits of the module rotation angle, respectively.

5. The method according to claim 1, characterized in that, The Bayesian proxy model includes: Bayesian surrogate model based on Gaussian process and Bayesian surrogate model based on random forest.

6. A device for modular structural optimization design, characterized in that, include: Decomposition module is used to build a modular structural system based on the target structure and decompose it into multiple isomorphic subsystems and the remaining system; The recombination module is used to perform finite element analysis on the isomorphic subsystem and construct a super-element matrix; The isomorphic subsystem is transformed from physical coordinates to modal coordinates, and a modular virtual framework is constructed based on the superunit matrix; The structure optimization module is used to initialize the design variables of the modular virtual framework and introduce a Bayesian optimization method to construct a Bayesian proxy model. The Bayesian proxy model is trained until a preset stopping condition is met to obtain optimized parameters; The refactoring module is used to update the design variables according to the optimization parameters to obtain an optimized modular structure system. The step of performing finite element analysis on the isomorphic subsystem and constructing a superelement matrix includes: Construct a finite element model of the isomorphic subsystem; Finite element analysis is performed on the finite element model of the isomorphic subsystem to extract the local feature matrix of the finite element model of the isomorphic subsystem; wherein, the local feature matrix includes the stiffness matrix and the mass matrix; The degrees of freedom of the nodes in the finite element model of the isomorphic subsystem are set, and the local feature matrix is ​​reduced in dimensionality based on the degrees of freedom to obtain a reduced local matrix; The reduced local matrices of the finite element models of each of the isomorphic subsystems are assembled to obtain the super-element matrix; The construction of a modular virtual framework based on the super-unit matrix includes: In the isomorphic subsystem under physical coordinates, interface displacement compatibility and force equilibrium conditions are introduced; Determine the function and boundary of the finite element model of each isomorphic subsystem and define them as independent modules; Establish connections between the individual modules to construct the modular virtual framework; The target structure involved is a large-scale modular aerospace structure, and the density of all materials is defined as follows: The Young's modulus is defined as 71.7 GPa, and the Poisson's ratio is defined as 0.

33. By combining modular design and parametric modeling methods, a modular aerospace structure system is constructed. The modular aerospace structure system is decomposed into multiple modules with specific functions using finite element analysis, and these modules are rapidly designed and reassembled using parametric modeling technology. This enables rapid customization and optimization of the target aerospace structure, as detailed below: ； in, This indicates a modular aerospace structure system. Representing isomorphic subsystems, Indicates the remaining system, This represents the total number of isomorphic subsystems and the remaining systems.

7. An apparatus for performing a modular structural optimization design method, characterized in that, include: processor; Memory used to store processor-executable instructions; When the processor executes the executable instructions, it implements the method as described in any one of claims 1 to 5.

8. A non-volatile computer-readable storage medium, characterized in that, Includes storage of computer programs or instructions that, when executed, cause the method as described in any one of claims 1 to 5 to be implemented.

Images