Adjustment Method, System and Electronic Device for Material Topology Structure

Abstract

This application is applicable to the field of computer-aided design and engineering, and provides a method, a system and an electronic device for adjusting the topological structure of a material. The method includes: determining a structure optimization task based on the obtained material topological structure to be optimized and optimization requirement information; decomposing the structure optimization task based on a target analysis method to obtain a plurality of optimization subtasks; allocating the optimization subtasks to a plurality of target neural network processors, so that each target neural network processor solves at least one optimization subtask to obtain a local optimal solution corresponding to the optimization subtask; obtaining a global optimal solution corresponding to the structure optimization task based on the local optimal solution, where the global optimal solution includes a plurality of second manufacturing parameters for determining the target manufacturing parameters at each target position; and adjusting the first manufacturing parameter at each target position in the material topological structure to be optimized to the corresponding target manufacturing parameter to obtain an optimized material topological structure. This solution can improve the optimization efficiency of the material topological structure.

Description

Technical Field

[0001] This application belongs to the field of computer-aided design and engineering, and particularly relates to a method, system, and electronic device for adjusting the topological structure of materials. Background Art

[0002] In the field of industrial engineering design and manufacturing, the topological structure of materials describes the distribution of materials in the design space and is used to guide structural design. By adjusting the material distribution within a given design domain, the optimization of the topological structure of materials can be achieved, and a topological structure of materials with the best performance that meets user requirements can be obtained. However, as the number of design variables increases, the complexity of the solution also increases, and the optimization efficiency of the topological structure of materials decreases. Summary of the Invention

[0003] Embodiments of this application provide a method, system, and electronic device for adjusting the topological structure of materials to solve the problem of reduced optimization efficiency of the topological structure of materials in the prior art.

[0004] The first aspect of the embodiments of this application provides a method for adjusting the topological structure of materials, including:

[0005] Based on the obtained topological structure of the material to be optimized and the optimization requirement information, determining a structure optimization task; the topological structure of the material to be optimized corresponds to structure manufacturing constraint parameters, and a first manufacturing parameter corresponds to each target position of the topological structure of the material to be optimized;

[0006] Based on a target analysis method, decomposing the structure optimization task to obtain a plurality of optimization subtasks;

[0007] Assigning the optimization subtasks to a plurality of target neural network processors, so that each target neural network processor solves at least one of the optimization subtasks to obtain a local optimal solution corresponding to each optimization subtask;

[0008] Based on the local optimal solutions, obtaining a global optimal solution corresponding to the structure optimization task, where the global optimal solution includes a plurality of second manufacturing parameters, and the second manufacturing parameters are used to compare with the structure manufacturing constraint parameters to determine the target manufacturing parameter at each target position; the target manufacturing parameter is the second manufacturing parameter or the structure manufacturing constraint parameter;

[0009] Adjusting the first manufacturing parameter at each target position in the topological structure of the material to be optimized to the target manufacturing parameter corresponding to the target position to obtain an optimized topological structure of the material.

[0010] The second aspect of the embodiments of this application provides a system for adjusting the topological structure of materials, including:

[0011] A determination module, configured to determine a structure optimization task based on the obtained topological structure of the material to be optimized and the optimization requirement information; the topological structure of the material to be optimized corresponds to structure manufacturing constraint parameters, and a first manufacturing parameter corresponds to each target position of the topological structure of the material to be optimized;

[0012] A decomposition module, configured to decompose the structure optimization task based on a target analysis method to obtain a plurality of optimization subtasks;

[0013] An allocation and solution module, configured to allocate the optimization subtasks to a plurality of target neural network processors, so that each of the target neural network processors solves at least one of the optimization subtasks to obtain a local optimal solution corresponding to each of the optimization subtasks;

[0014] A solution module, configured to obtain a global optimal solution corresponding to the structure optimization task based on the local optimal solutions, where the global optimal solution includes a plurality of second manufacturing parameters, and the second manufacturing parameters are used to compare with the structure manufacturing constraint parameters to determine the target manufacturing parameter at each target position; the target manufacturing parameter is the second manufacturing parameter or the structure manufacturing constraint parameter;

[0015] An adjustment module, configured to adjust the first manufacturing parameter at each target position in the topological structure of the material to be optimized to the target manufacturing parameter corresponding to the target position to obtain an optimized topological structure of the material.

[0016] A third aspect of the embodiments of the present application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, where when the processor executes the computer program, the steps of the method described in the first aspect are implemented.

[0017] A fourth aspect of the embodiments of the present application provides a computer program product, where the computer program product includes a computer program, and when the computer program is executed by a processor, the steps of the method described in the first aspect are implemented.

[0018] A fifth aspect of the embodiments of the present application provides a computer-readable storage medium, where the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described in the first aspect are implemented.

[0019] As can be seen from the above, the present application uses a target analysis method to decompose the determined structure optimization task into multiple optimization subtasks. After the task decomposition, the solution complexity of each obtained optimization subtask is greatly reduced. Then, multiple target neural network processors are used to solve the optimization subtasks respectively, and a local optimal solution corresponding to each optimization subtask is obtained. This step realizes parallel solution through the target neural network processors, improving the solution efficiency. Then, a global optimal solution corresponding to the structure optimization task is determined through the local optimal solutions. The second manufacturing parameters in the global optimal solution are compared with the corresponding structure manufacturing constraint parameters to determine the target manufacturing parameters, and the first manufacturing parameters in the material topology structure to be optimized are adjusted to the target manufacturing parameters to realize the adjustment of the material topology structure and obtain an optimized material topology structure. The reduction of the solution complexity and the improvement of the solution efficiency jointly improve the optimization efficiency of the material topology structure. Description of the Drawings

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0021] Figure 1 It is a flowchart of a method for adjusting a material topology structure provided by an embodiment of the present application;

[0022] Figure 2 It is a flowchart of a task linear transformation based on MMA provided by an embodiment of the present application;

[0023] Figure 3 It is a structural diagram of a system for adjusting a material topology structure provided by an embodiment of the present application;

[0024] Figure 4 It is a structural diagram of an electronic device provided by an embodiment of the present application. Detailed Embodiments

[0025] In the following description, specific details such as specific system structures and technologies are proposed for the purpose of illustration rather than limitation, so as to thoroughly understand the embodiments of the present application. However, those skilled in the art should clearly understand that the present application can also be implemented in other embodiments without these specific details. In other cases, the detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present application.

[0026] It should be understood that, as used in this specification and the appended claims, the term "comprising" indicates the presence of the described features, integers, steps, operations, elements, and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0027] It should also be understood that the terminology used in this application's specification is for the purpose of describing particular embodiments only and is not intended to limit this application. As used in this application's specification and the appended claims, unless the context clearly dictates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms.

[0028] It should be further understood that the term "and / or" as used in this application's specification and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0029] As used in this specification and the appended claims, the term "if" can be interpreted, depending on the context, as "when", "once", "in response to determining", or "in response to detecting". Similarly, the phrase "if determined" or "if detected [the described condition or event]" can be interpreted, depending on the context, as meaning "once determined", "in response to determining", "once detected [the described condition or event]", or "in response to detecting [the described condition or event]".

[0030] In a specific implementation, the terminals described in the embodiments of this application include, but are not limited to, other portable devices such as mobile phones, laptop computers, or tablet computers having a touch-sensitive surface (e.g., a touch screen display and / or a touchpad). It should also be understood that in some embodiments, the device is not a portable communication device, but a desktop computer having a touch-sensitive surface (e.g., a touch screen display and / or a touchpad).

[0031] In the following discussion, terminals including a display and a touch-sensitive surface are described. However, it should be understood that a terminal can include one or more other physical user interface devices such as a physical keyboard, a mouse, and / or a joystick.

[0032] The terminal supports various applications, such as one or more of the following: drawing applications, presentation applications, word processing applications, website creation applications, disc burning applications, spreadsheet applications, game applications, telephone applications, video conferencing applications, email applications, instant messaging applications, exercise support applications, photo management applications, digital camera applications, digital video camera applications, web browsing applications, digital music player applications, and / or digital video player applications.

[0033] Various applications that can be executed on the terminal can use at least one common physical user interface device such as a touch-sensitive surface. One or more functions of the touch-sensitive surface and the corresponding information displayed on the terminal can be adjusted and / or changed between applications and / or within the respective applications. In this way, the common physical architecture of the terminal (e.g., the touch-sensitive surface) can support various applications with a user interface that is intuitive and transparent to the user.

[0034] It should be understood that the magnitudes of the sequence numbers of the steps in this embodiment do not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of this application.

[0035] With the development of materials science, computer technology, and optimization algorithms, the optimization of material topological structures has gradually been applied in the fields of engineering design and manufacturing. For example, by adjusting the structural weight of an aircraft to improve fuel efficiency, and by reducing the weight of automotive components to improve the safety and driving performance of the vehicle, and so on. The optimization of material topological structures refers to finding the optimal material distribution within a given design domain to meet the specific performance requirements of the user. That is, under constraint conditions such as a specific volume, mass, or other specific performance requirements, the material distribution of the material topological structure is adjusted to obtain a material topological structure that meets the specific performance requirements of the user and has the best performance.

[0036] When adjusting the material topological structure to achieve its optimization, an increase in design complexity such as an increase in design variables will increase the solution complexity, make the solution more difficult, and it is difficult to efficiently complete the solution and achieve a rapid optimization adjustment of the material topological structure.

[0037] In response to this, this application provides a method, system, and electronic device for adjusting a material topological structure to improve the optimization efficiency of the material topological structure.

[0038] To illustrate the technical solutions described in this application, the following will be described through specific embodiments.

[0039] See Figure 1 , Figure 1It is a flowchart of a method for adjusting the topological structure of a material provided by an embodiment of the present application. As Figure 1 shown, a method for adjusting the topological structure of a material, the method comprising the following steps:

[0040] Step 101, based on the obtained material topological structure to be optimized and optimization requirement information, determine a structure optimization task.

[0041] Wherein, the material topological structure to be optimized corresponds to structure manufacturing constraint parameters, and a first manufacturing parameter corresponds to each target position of the material topological structure to be optimized.

[0042] The material topological structure to be optimized is the material topological structure that needs to be optimized given by the user. The optimization requirement information is the limited configuration information corresponding to the adjustment and optimization of the material topological structure to be optimized input by the user, including a design domain, design variables, an objective function, and constraint conditions, etc. The structure optimization task is the task that needs to be solved, and its purpose is to make the target performance of the material topological structure reach the optimal under given conditions. The structure manufacturing constraint parameters are the constraint parameters that need to be followed for the production and manufacturing of the material topological structure in the actual production process, usually parameter ranges. The target position is the position with manufacturing parameters, located in the design domain. The first manufacturing parameter is the original manufacturing parameter of the material topological structure to be optimized itself.

[0043] Determining the structure optimization task based on the material topological structure to be optimized and the optimization requirement information, clarifying the optimization scope, avoiding waste of computing resources in irrelevant areas, improving the solution efficiency, and further improving the optimization efficiency.

[0044] In some embodiments, an interactive interface is provided to facilitate the user to import the material topological structure to be optimized and input the optimization requirement information through the interactive interface. In addition, the user can also view and obtain information such as the optimized material topological structure after adjustment through the interactive interface.

[0045] In some embodiments, the determining the structure optimization task based on the obtained material topological structure to be optimized and the optimization requirement information includes: extracting the optimization configuration parameters in the optimization requirement information; the optimization configuration parameters include a design domain, design variables, an objective function, and constraint conditions; using the material topological structure to be optimized as the optimization object to obtain the optimization task including the optimization object and the optimization configuration parameters.

[0046] The design domain is the physical area in the material topological structure where the manufacturing parameters are allowed to be adjusted, such as the wing of an aircraft, the engine hood of a car, etc. The design variables are the physical quantities that can be adjusted within the design domain, such as the cross-sectional dimension of a connecting rod, etc. The objective function is an expression for quantifying the structural performance. The constraint conditions are the adjustable ranges of the design variables, such as stress constraints, mass constraints, material properties, etc.

[0047] The topological structure of the material to be optimized is the object of optimization in the structure optimization task, and the optimization configuration parameters composed of the design domain, design variables, objective function, and constraint conditions are the optimization guidance information in the structure optimization task.

[0048] Through the extraction operation, the optimization configuration parameters in the structure optimization task are clarified, which is convenient for accurately and effectively realizing the structure optimization.

[0049] In some embodiments, when the complexity of the task scenario of the structure optimization task is greater than or equal to the set scenario complexity, the neural network processing method is determined as the target analysis method; or, when the complexity of the task scenario of the structure optimization task is less than the set scenario complexity, the finite element analysis method is determined as the target analysis method; or, when a configured analysis method is configured in the optimization requirement information, the configured analysis method is determined as the target analysis method, and the configured analysis method is the neural network processing method or the finite element analysis method.

[0050] The target analysis method is used to analyze and decompose the structure optimization task, simplifying complex tasks. The target analysis method is the neural network processing method or the finite element analysis method (Finite Element Analysis, FEA). The neural network processing method can be the Physics-Informed Neural Networks (PINN) method.

[0051] Determining a suitable target analysis method to achieve task decomposition helps optimize the solution resource configuration and improve the optimization efficiency.

[0052] In some embodiments, the complexity of the task scenario is determined by multi-dimensional factors in the structure optimization task, such as the physical field coupling degree, geometric dimension, number of design variables, etc. According to the comparison result of the task scenario complexity and the set scenario complexity, the target analysis method is determined. For example, when encountering multi-scale physical fields, such as acoustic-solid coupling, fluid-solid-thermal coupling, etc., the neural network processing method, such as the PINN method, is selected.

[0053] In some embodiments, when the complexity of the task scenario of the structure optimization task is less than the set scenario complexity, the neural network processing method can also be determined as the target analysis method.

[0054] In some embodiments, when the user inputs the optimization requirement information, the analysis method can be set. When a configured analysis method is configured in the optimization requirement information input by the user, the configured analysis method can be used as the target analysis method to meet the user's needs.

[0055] In some embodiments, the target analysis method is determined according to the user's solution accuracy.

[0056] In some embodiments, the target analysis method may also be a machine learning method or other hybrid methods, such as the support vector machine method.

[0057] Step 102: Based on the target analysis method, decompose the structure optimization task to obtain multiple optimization subtasks.

[0058] According to the target analysis method, decompose the structure optimization task into multiple optimization subtasks, simplify the complex task, and reduce the solution difficulty.

[0059] In some embodiments, the decomposition of the structure optimization task based on the target analysis method to obtain multiple optimization subtasks includes: when the target analysis method is the finite element analysis method, convert the material topology structure to be optimized into a finite element model; perform mesh division on the finite element model to obtain multiple mesh elements; respectively extract from the design domain and design variables of the optimization configuration parameters the unit design domain and unit design variables that match the unit structure parameters of each mesh element; respectively construct the unit constraint conditions and unit objective functions of each mesh element according to the unit boundary attributes, unit structure parameters, constraint conditions, and objective function of the mesh element; take each mesh element as an optimization sub-object, and take the unit design domain, unit design variables, unit constraint conditions, and unit objective functions of the mesh element as the unit optimization configuration parameters of the mesh element, so as to obtain multiple optimization subtasks including the optimization sub-objects and the unit optimization configuration parameters.

[0060] When the target analysis method is the finite element analysis method, first convert the material topology structure to be optimized into a finite element model, and this finite element model has the loads and boundary conditions of the material topology structure to be optimized. Then, divide the finite element model into multiple mesh elements through mesh division, and discretize the complex structure into finite element meshes to realize the decomposition of the optimization object in the structure optimization task.

[0061] The unit structure parameters are the manufacturing parameters included in the mesh element. The information containing these manufacturing parameters can be respectively extracted from the design domain and design variables of the optimization configuration parameters in combination with the unit structure parameters of the mesh element as the unit design domain and unit design variables.

[0062] The unit boundary attribute is the unit boundary quantity of the mesh element, such as the position of the mesh element in the entire material topology structure. According to the unit boundary attributes, unit structure parameters, and the constraint conditions and objective function in the optimization configuration information of the mesh element itself, respectively construct the unit constraint conditions and unit objective functions corresponding to each mesh element during optimization.

[0063] In some embodiments, an interpolation method is adopted to obtain the unit optimization configuration parameters of each grid unit, and the decomposition of the optimization configuration information is realized.

[0064] When the target analysis method is the finite element analysis method, through decomposition, a plurality of grid units and the unit optimization configuration parameters of each grid unit are obtained, that is, a plurality of optimization subtasks including the optimization sub-objects and the unit optimization configuration parameters are obtained, realizing the refined configuration of local parameters, reducing the task processing difficulty, and facilitating the analysis of each finite element.

[0065] In some embodiments, the user specifies that the target analysis method is the finite element analysis method and sets the number of grids, and then divides the finite element model according to the number of grids.

[0066] In some embodiments, to ensure the efficient solution of the local optimal solution corresponding to each grid unit, during grid division, it is necessary to make each grid unit have approximately the same amount of calculation.

[0067] In some embodiments, the grid units are connected by boundaries. The boundary is the dividing line of the grid units, and the boundary nodes are the nodes located on this dividing line, that is, data. The boundary nodes are shared by at least two grid units. Grid division is realized through the Metis data processing software to minimize the number of boundary nodes, reduce the communication complexity and overhead, and improve the solution efficiency.

[0068] In some embodiments, based on the target analysis method, decomposing the structure optimization task to obtain a plurality of optimization subtasks includes: when the target analysis method is the neural network processing method, inputting the material topology structure to be optimized and the optimization configuration parameters into a first neural network model to obtain a second neural network model; using the set of optimizable parameters of each neuron in the second neural network model as the optimization subtask to obtain a plurality of the optimization subtasks.

[0069] When the target analysis method is the neural network processing method, inputting the material topology structure to be optimized and the optimization configuration parameters into a first neural network model, correspondingly obtaining a second neural network model, and then using the set of optimizable parameters of each neuron in the second neural network model as the optimization subtask to obtain local optimization tasks at the neuron level, that is, obtaining a plurality of the optimization subtasks. Among them, the set of optimizable parameters is a weight matrix, a bias term, an activation function, etc.

[0070] Task decomposition is realized through the neural network model, converting the high-dimensional material topology structure to be optimized into a plurality of low-dimensional optimization subtasks, and reducing the task solution difficulty.

[0071] Step 103: Assign the optimized subtasks to multiple target neural network processors, so that each target neural network processor solves at least one of the optimized subtasks to obtain a local optimal solution corresponding to each optimized subtask.

[0072] After obtaining multiple optimized subtasks, the optimized subtasks are assigned to multiple target neural network processors, and the optimized subtasks are solved by the multiple target neural network processors to achieve parallel solution, improve the solution efficiency, quickly obtain the local optimal solutions of the multiple optimized subtasks respectively, and further improve the optimization efficiency.

[0073] In some embodiments, the assigning the optimized subtasks to multiple target neural network processors includes: determining the load status of multiple neural network processors and the processing resource requirements of the multiple optimized subtasks; and according to the load status and the processing resource requirements, assigning, for each optimized subtask, the target neural network processor whose load status meets the processing resource requirements from the multiple neural network processors.

[0074] Determine the load status of multiple neural network processors (Neural Processing Unit, NPU) and the processing resource requirements of multiple optimized subtasks during task solution, so as to determine the target neural network processors from the multiple neural network processors according to the load status and the processing resource requirements, and achieve the purpose of matching the optimized subtasks to the target neural network processors for solving them.

[0075] In some embodiments, the optimized subtasks can be sequentially assigned to multiple target neural network processors by a polling assignment method to achieve efficient utilization of the solution resources.

[0076] During the solution process of the optimized subtasks, if it is monitored that some target neural network processors are in an idle state for solution, the optimized subtasks can be flexibly scheduled to assign the optimized subtasks to the target neural network processors in the idle state for solution to improve the solution efficiency.

[0077] Taking the finite element analysis method as an example, solve the stiffness matrix or mass matrix of each grid unit.

[0078] During the solution, according to the objective function, schedule the solution code required by the neural network processor during task solution, such as the finite element analysis code.

[0079] To ensure the stability of the solution, suppress the influence of noise, obtain a smoother solution result, analyze the sensitivity of design variables, and use processing methods such as those based on the range of motion or weighted average to smooth the solution result. Taking the finite element analysis method as an example, the sensitivity of each grid cell and the sensitivity of its neighboring grid cells need to be considered to improve the stability of the solution. The retrieval of neighboring grid cells can be performed in parallel on the neural network processor through matrix operations, thus completing the retrieval quickly and accurately.

[0080] Among them, sensitivity analysis is to calculate the gradient of each design variable under the objective function and constraint conditions, which is convenient for numerical adjustment during iteration.

[0081] In some embodiments, gradient calculation is performed through the Compute Unified Device Architecture (CUDA) to achieve high-efficiency processing of a large amount of parallel data.

[0082] When solving each optimization subtask, iterative solution is performed based on the parameters after sensitivity analysis, and the convergence is judged. The solution that meets the convergence condition of the optimization subtask is used as the local optimal solution of the optimization subtask. If not, continue the iteration, and the current solution result can be used as the initial parameter for the next iteration to improve the iteration efficiency.

[0083] In some embodiments, to improve the solution efficiency, reduce the solution difficulty, and obtain a convergent local optimal solution, control each of the target neural network processors to perform linear analysis processing on the assigned optimization subtasks; wherein, in the case where the optimization subtask is a non-linear task, control the target neural network processor to convert the non-linear task into a linear task and then perform the linear analysis processing on the optimization subtask.

[0084] In some embodiments, methods such as the Moving Asymptotes Method (MMA), convex hull method, and penalty function linearization method are used to convert non-linear tasks into linear tasks.

[0085] As Figure 2 shown, Figure 2It is a flowchart of task linear transformation provided by an embodiment of the present application based on MMA. When implementing linear transformation using MMA, first, an approximate linear equation is constructed based on a non-linear task, and then the variable increment of the approximate linear equation is solved, and it is judged whether the variable increment converges in the approximate linear equation. If it does not converge, the variable increment of the approximate linear equation is continuously iteratively solved until the variable increment that converges in the approximate linear equation is obtained. The variables in the non-linear task are adjusted to the converged variable increment, and it is further judged whether the converged variable increment converges in the non-linear task. If it converges, the non-linear task containing the converged variable increment is used as the transformed linear task. If it does not converge, the approximate linear equation is adjusted until the converged variable increment converges in the non-linear task. Through constructing an approximate linear equation and iterative adjustment, MMA continuously promotes the convergence of variable increments, obtains the corresponding linear task, and thereby reduces the solution difficulty of the task.

[0086] When performing linear transformation, parallel processing can also be achieved through a neural network processor to accelerate the processing efficiency.

[0087] In some embodiments, the input and output of data (such as the output of material topology, the output of iterative data, the output of physical field quantities, etc.) are implemented through a central processing unit (CPU), the determination of structure optimization tasks (such as analyzing optimization requirement information, clarifying the design domain and non-design domain), the determination of the target analysis method, the determination of the solution method (such as determining whether to use the analytical method or the numerical method during solution according to the processing resource requirements of the task and the number of grid cells of the grid unit), the decomposition of structure optimization tasks, the allocation of optimization subtasks, and the adjustment of manufacturing parameters of the topology to be optimized, etc. The solution operation and convergence judgment are completed through a neural network processor.

[0088] Among them, the non-design domain participates in the solution of the optimal solution, and the design domain participates in the solution of the optimal solution, sensitivity analysis, manufacturing parameter adjustment, etc.

[0089] Through the central processing unit and multiple target neural network processors, heterogeneous parallel processing is achieved. Heterogeneous parallel processing combines different types of processors in the hardware architecture, effectively processes complex tasks using the specialties of different types of processors, reasonably plans the functions of each processor, avoids resource competition problems during the processing of the central processing unit and the neural network processor, and improves the processing performance and processing efficiency. In structure optimization, heterogeneous parallel processing can improve the accuracy and efficiency of sensitivity analysis, improve the solution efficiency of optimization subtasks, reduce the optimization time, and improve the optimization accuracy.

[0090] The central processing unit and multiple neural network processors have good scalability. Through their combination, different hardware environments and optimization requirements can be adapted.

[0091] In some embodiments, programming languages such as C++ are used to construct the control logic of the central processing unit and the corresponding data interfaces. Task scheduling and management are completed through the Message Passing Interface (MPI) and the Open Multi-Processing (OpenMP) method.

[0092] The neural network processor can efficiently process computationally intensive tasks such as matrix operations and convolution calculations. It has more advantages in parallel solving, can significantly shorten the task solving time, and reduce the solving cost.

[0093] In some embodiments, Mindspore is used to implement the parallel programming processing of the neural network processor, and parallel computing is implemented on multiple neural network processors, so as to utilize the parallel processing ability of the neural network processor and accelerate the solving process.

[0094] In some embodiments, an intermediate layer is set between the central processing unit and multiple neural network processors. The central processing unit monitors the load status of the neural network processors through the intermediate layer and optimizes the flexible allocation of subtasks.

[0095] In some embodiments, a shared memory is set to store shared data, communication data, etc. for optimizing subtasks. For example, boundary node data and data of the NVIDIA Collective Communication Library (NCCL) are stored, avoiding the impact of fragmented storage on the optimization performance, and improving the data access efficiency and data storage efficiency of the neural network processor.

[0096] In some embodiments, during the solution process of the finite element analysis method, data is stored in the form of arrays or sparse matrices.

[0097] Step 104, obtaining a global optimal solution corresponding to the structure optimization task based on the local optimal solution. The global optimal solution includes multiple second manufacturing parameters, and the second manufacturing parameters are used to compare with the structure manufacturing constraint parameters to determine the target manufacturing parameters at each target position.

[0098] Wherein, the target manufacturing parameter is the second manufacturing parameter or the structure manufacturing constraint parameter.

[0099] After local parallel solving, focus on the global task, that is, the structure optimization task. Further obtain the global optimal solution corresponding to the structure optimization task through the local optimal solution, and further obtain the target manufacturing parameters for realizing the optimization adjustment through comparison. While optimizing, ensure the feasibility of the optimization scheme, avoid re-solving caused by the inability to implement the optimization scheme, and improve the optimization efficiency.

[0100] In some embodiments, multiple local optimal solutions are used as the solution results, and it is determined whether the structural optimization task satisfies its convergence condition under this set of local optimal solutions. If it is satisfied, this set of local optimal solutions is used as the global optimal solution. If it is not satisfied, fine-tuning processing is performed based on the local optimal solutions to obtain the global optimal solution corresponding to the structural optimization task.

[0101] Taking the finite element analysis method as an example, the element matrices are assembled into the global matrix, and the boundary conditions are added, and the global optimal solution is obtained by solving the linear equations.

[0102] The global optimal solution includes multiple second manufacturing parameters. The second manufacturing parameters at the same position are compared with the structural manufacturing constraint parameters. If the second manufacturing parameter belongs to the structural manufacturing constraint parameter, the second manufacturing parameter is determined as the target manufacturing parameter. If the second manufacturing parameter does not belong to the structural manufacturing constraint parameter, the parameter in the structural manufacturing constraint parameters that is closest to the second manufacturing parameter is determined as the target manufacturing parameter, so as to determine the target manufacturing parameter at each target position. Combining optimization with the actual production and manufacturing scenario ensures the manufacturability of the material topology structure. The comparison and determination process can be completed in parallel in the neural network processor to improve the comparison and determination efficiency.

[0103] Step 105, adjust the first manufacturing parameter at each target position in the material topology structure to be optimized to the target manufacturing parameter corresponding to the target position to obtain an optimized material topology structure.

[0104] Adjust the first manufacturing parameter at each target position in the material topology structure to be optimized to the target manufacturing parameter corresponding to the target position, realize the conversion from the solution data to the entity structure, and obtain an optimized material topology structure, which is convenient for guiding the user's production and manufacturing.

[0105] The central processing unit performs post-processing on the optimized material topology structure and outputs a visualized optimized material topology structure for the user to view.

[0106] In some embodiments, an analysis report including the manufacturing parameters before and after adjustment, such as stress distribution, mode, etc., may also be output together so that the user can understand the final optimization effect.

[0107] In the embodiments of the present application, a target analysis method is used to decompose the determined structure optimization task into multiple optimization subtasks. After the task decomposition, the solution complexity of each obtained optimization subtask is greatly reduced. Then, multiple target neural network processors are used to solve the optimization subtasks respectively, and a local optimal solution corresponding to each optimization subtask is obtained. This step realizes parallel solution through the target neural network processors, improving the solution efficiency. Then, a global optimal solution corresponding to the structure optimization task is determined through the local optimal solutions, and the second manufacturing parameters in the global optimal solution are compared with the corresponding structure manufacturing constraint parameters to determine the target manufacturing parameters, and the first manufacturing parameters in the material topology structure to be optimized are adjusted to the target manufacturing parameters, realizing the adjustment of the material topology structure and obtaining the optimized material topology structure. The reduction of the solution complexity and the improvement of the solution efficiency together improve the optimization efficiency of the material topology structure.

[0108] See Figure 3 , Figure 3 FIG. is a structural diagram of a system for adjusting a material topology structure provided by an embodiment of the present application. For the sake of illustration, only the parts related to the embodiments of the present application are shown.

[0109] The system 300 for adjusting the material topology structure includes: a determination module 301, a decomposition module 302, an allocation and solution module 303, a solution module 304, and an adjustment module 305.

[0110] The determination module 301 is configured to determine a structure optimization task based on the obtained material topology structure to be optimized and optimization requirement information; the material topology structure to be optimized corresponds to structure manufacturing constraint parameters, and each target position of the material topology structure to be optimized corresponds to a first manufacturing parameter.

[0111] The decomposition module 302 is configured to decompose the structure optimization task based on a target analysis method to obtain multiple optimization subtasks.

[0112] The allocation and solution module 303 is configured to allocate the optimization subtasks to multiple target neural network processors, so that each target neural network processor solves at least one of the optimization subtasks to obtain a local optimal solution corresponding to each optimization subtask.

[0113] The solution module 304 is configured to obtain a global optimal solution corresponding to the structure optimization task based on the local optimal solutions. The global optimal solution includes multiple second manufacturing parameters, and the second manufacturing parameters are used to be compared with the structure manufacturing constraint parameters to determine the target manufacturing parameter at each target position; the target manufacturing parameter is the second manufacturing parameter or the structure manufacturing constraint parameter.

[0114] An adjustment module 305 is configured to adjust the first manufacturing parameter at each of the target positions in the material topology to be optimized to the corresponding target manufacturing parameter at the target position, so as to obtain an optimized material topology.

[0115] In some embodiments, the determining module is specifically configured to:

[0116] Extract the optimization configuration parameters in the optimization requirement information; the optimization configuration parameters include a design domain, design variables, an objective function, and constraint conditions;

[0117] Use the material topology to be optimized as the optimization object to obtain the optimization task including the optimization object and the optimization configuration parameters.

[0118] In some embodiments, the decomposition module is configured to:

[0119] When the target analysis method is the finite element analysis method, convert the material topology to be optimized into a finite element model;

[0120] Perform mesh division on the finite element model to obtain a plurality of mesh elements;

[0121] Respectively extract a unit design domain and unit design variables that match the unit structure parameters of each mesh element from the design domain and the design variables of the optimization configuration parameters;

[0122] According to the unit boundary attributes of the mesh element, the unit structure parameters, the constraint conditions, and the objective function, respectively construct the unit constraint conditions and the unit objective function of the mesh element;

[0123] Use each mesh element as an optimization sub-object, and use the unit design domain, the unit design variables, the unit constraint conditions, and the unit objective function of the mesh element as the unit optimization configuration parameters of the mesh element to obtain a plurality of the optimization sub-tasks including the optimization sub-object and the unit optimization configuration parameters.

[0124] In some embodiments, the decomposition module is further configured to:

[0125] When the target analysis method is the neural network processing method, input the material topology to be optimized and the optimization configuration parameters into a first neural network model to obtain a second neural network model;

[0126] Use the set of optimizable parameters of each neuron in the second neural network model as the optimization sub-task to obtain a plurality of the optimization sub-tasks.

[0127] In some embodiments, the system further includes a target analysis method determination module, configured to:

[0128] When the task scenario complexity of the structure optimization task is greater than or equal to a set scenario complexity, determine the neural network processing method as the target analysis method; or,

[0129] When the task scenario complexity of the structure optimization task is less than the set scenario complexity, determine the finite element analysis method as the target analysis method; or,

[0130] When a configured analysis method is configured in the optimization requirement information, determine the configured analysis method as the target analysis method, where the configured analysis method is the neural network processing method or the finite element analysis method.

[0131] In some embodiments, the allocation module is specifically configured to:

[0132] Determine the load status of multiple neural network processors and the processing resource requirements of multiple optimization subtasks;

[0133] According to the load status and the processing resource requirements, allocate, for each optimization subtask, the target neural network processor whose load status meets the processing resource requirements from multiple neural network processors.

[0134] In some embodiments, the system further includes a linear analysis processing module, configured to:

[0135] Control each target neural network processor to perform linear analysis processing on the allocated optimization subtask;

[0136] Wherein, when the optimization subtask is a non-linear task, control the target neural network processor to convert the non-linear task into a linear task and then perform the linear analysis processing on the optimization subtask.

[0137] The material topology structure adjustment system provided by the embodiments of the present application can implement each process of the embodiments of the above-mentioned material topology structure adjustment method, and can achieve the same technical effects. To avoid repetition, details are not described herein again.

[0138] Figure 4 It is a structural diagram of an electronic device provided by an embodiment of the present application. As shown in this figure, the electronic device 4 of this embodiment includes: at least one processor 40 ( Figure 4 only one is shown), a memory 41, and a computer program 42 stored in the memory 41 and executable on the at least one processor 40. When the processor 40 executes the computer program 42, the steps in any of the above method embodiments are implemented.

[0139] The electronic device 4 may be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The electronic device 4 may include, but is not limited to, a processor 40 and a memory 41. Those skilled in the art can understand that Figure 4 These are merely examples of the electronic device 4 and do not constitute a limitation on the electronic device 4. It may include more or fewer components than shown in the figure, or combine certain components, or different components. For example, the electronic device may further include input / output devices, network access devices, a bus, etc.

[0140] The processor 40 may be a CPU, or may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0141] The memory 41 may be an internal storage unit of the electronic device 4, such as a hard disk or memory of the electronic device 4. The memory 41 may also be an external storage device of the electronic device 4, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc., equipped on the electronic device 4. Further, the memory 41 may also include both the internal storage unit and the external storage device of the electronic device 4. The memory 41 is used to store the computer program and other programs and data required by the electronic device. The memory 41 may also be used to temporarily store data that has been output or is to be output.

[0142] Those skilled in the art can clearly understand that, for the convenience and conciseness of description, only the above division of each functional unit and module is used as an example. In actual applications, the above functions can be allocated to different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiments can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit. In addition, the specific names of each functional unit and module are only for the convenience of mutual distinction and do not limit the protection scope of this application. The specific working processes of the units and modules in the above system can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0143] In the above embodiments, the descriptions of the respective embodiments have their own emphases. For the parts not detailed or recorded in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0144] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or by a combination of computer software and electronic hardware. Whether these functions are executed in the form of hardware or software depends on the specific application and design constraints of the technical solution. Professionals can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0145] In the embodiments provided in this application, it should be understood that the disclosed system / electronic device and method can be implemented in other ways. For example, the system / electronic device embodiments described above are only illustrative. For example, the division of the modules or units is only a logical function division. In actual implementation, there can be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the system or unit can be in electrical, mechanical or other forms.

[0146] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0147] In addition, in each embodiment of the present application, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0148] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, to implement all or part of the processes in the above-mentioned embodiment methods of the present application, it can also be completed by a computer program instructing relevant hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-mentioned various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate forms, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0149] All or part of the processes in the above-mentioned embodiment methods of the present application can also be implemented by a computer program product. When the computer program product runs on an electronic device, the electronic device can be made to execute the steps in the above-mentioned various method embodiments when executed.

[0150] The above-mentioned embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for adjusting the topological structure of a material, characterized in that Including: Determining a structure optimization task based on the obtained topological structure of the material to be optimized and optimization requirement information; The topological structure of the material to be optimized corresponds to structure manufacturing constraint parameters, and a first manufacturing parameter corresponds to each target position of the topological structure of the material to be optimized; The determining a structure optimization task based on the obtained topological structure of the material to be optimized and optimization requirement information includes: extracting optimization configuration parameters from the optimization requirement information; the optimization configuration parameters include a design domain, design variables, an objective function, and constraint conditions; using the topological structure of the material to be optimized as an optimization object to obtain the optimization task including the optimization object and the optimization configuration parameters; Decomposing the structure optimization task based on a target analysis method to obtain a plurality of optimization subtasks; Allocating the optimization subtasks to a plurality of target neural network processors, such that each target neural network processor solves at least one of the optimization subtasks to obtain a local optimal solution corresponding to each optimization subtask; Obtaining a global optimal solution corresponding to the structure optimization task based on the local optimal solutions, the global optimal solution includes a plurality of second manufacturing parameters, and the second manufacturing parameters are used to compare with the structure manufacturing constraint parameters to determine target manufacturing parameters at each target position; the target manufacturing parameter is the second manufacturing parameter or the structure manufacturing constraint parameter; Adjusting the first manufacturing parameter at each target position in the topological structure of the material to be optimized to the target manufacturing parameter corresponding to the target position to obtain an optimized topological structure of the material; Wherein, when the task scenario complexity of the structure optimization task is greater than or equal to a set scenario complexity, determining the neural network processing method as the target analysis method; or, when the task scenario complexity of the structure optimization task is less than the set scenario complexity, determining the finite element analysis method as the target analysis method; or, when a configuration analysis method is configured in the optimization requirement information, determining the configuration analysis method as the target analysis method, and the configuration analysis method is the neural network processing method or the finite element analysis method.

2. The method according to claim 1, wherein The decomposing the structure optimization task based on a target analysis method to obtain a plurality of optimization subtasks includes: When the target analysis method is the finite element analysis method, converting the topological structure of the material to be optimized into a finite element model; Performing mesh division on the finite element model to obtain a plurality of mesh elements; Respectively extracting a unit design domain and unit design variables that match the unit structure parameters of each mesh element from the design domain and the design variables of the optimization configuration parameters; Respectively constructing a unit constraint condition and a unit objective function of each mesh element according to the unit boundary attribute, the unit structure parameters, the constraint conditions, and the objective function of the mesh element; Taking each of the grid cells as an optimization sub-object, and taking the cell design domain, the cell design variables, the cell constraint conditions, and the cell objective function of the grid cell as the cell optimization configuration parameters of the grid cell, a plurality of the optimization sub-tasks including the optimization sub-object and the cell optimization configuration parameters are obtained.

3. The method according to claim 1, wherein Based on the objective analysis method, decomposing the structure optimization task to obtain a plurality of optimization sub-tasks, including: When the objective analysis method is the neural network processing method, inputting the material topology structure to be optimized and the optimization configuration parameters into a first neural network model to obtain a second neural network model; Taking the optimizable parameter set of each neuron in the second neural network model as the optimization sub-task, and obtaining a plurality of the optimization sub-tasks.

4. The method according to claim 1, characterized in that, The step of allocating the optimization sub-tasks to a plurality of target neural network processors includes: Determining the load status of a plurality of neural network processors and the processing resource requirements of a plurality of the optimization sub-tasks; According to the load status and the processing resource requirements, allocating, for each of the optimization sub-tasks, the target neural network processor whose load status meets the processing resource requirements from a plurality of the neural network processors.

5. The method according to claim 1, wherein The method further includes: Controlling each of the target neural network processors to perform linear analysis processing on the allocated optimization sub-tasks; Wherein, when the optimization sub-task is a non-linear task, controlling the target neural network processor to convert the non-linear task into a linear task and then perform the linear analysis processing on the optimization sub-task.

6. An adjustment system for the topological structure of a material, characterized in that, Including: A determination module, configured to determine a structure optimization task based on the obtained material topology structure to be optimized and optimization requirement information; The material topology structure to be optimized corresponds to structure manufacturing constraint parameters, and each target position of the material topology structure to be optimized corresponds to a first manufacturing parameter; the determination module is specifically configured to: extract the optimization configuration parameters in the optimization requirement information; the optimization configuration parameters include a design domain, design variables, an objective function, and constraint conditions; taking the material topology structure to be optimized as an optimization object, and obtaining the optimization task including the optimization object and the optimization configuration parameters; A decomposition module, configured to decompose the structure optimization task based on an objective analysis method to obtain a plurality of optimization sub-tasks; An allocation and solution module, configured to allocate the optimization sub-tasks to a plurality of target neural network processors, so that each of the target neural network processors solves at least one of the optimization sub-tasks to obtain a local optimal solution corresponding to each of the optimization sub-tasks; A solution module, configured to obtain a global optimal solution corresponding to the structure optimization task based on the local optimal solutions, the global optimal solution includes a plurality of second manufacturing parameters, and the second manufacturing parameters are used to compare with the structure manufacturing constraint parameters to determine the target manufacturing parameter at each of the target positions; the target manufacturing parameter is the second manufacturing parameter or the structure manufacturing constraint parameter; An adjustment module, configured to adjust the first manufacturing parameter at each of the target positions in the material topology to be optimized to the corresponding target manufacturing parameter at the target position, so as to obtain an optimized material topology; Wherein, the adjustment system of the material topology further includes a target analysis method determination module, configured to: when the task scenario complexity of the structure optimization task is greater than or equal to a set scenario complexity, determine the neural network processing method as the target analysis method; or, when the task scenario complexity of the structure optimization task is less than the set scenario complexity, determine the finite element analysis method as the target analysis method; or, when a configured analysis method is configured in the optimization requirement information, determine the configured analysis method as the target analysis method, and the configured analysis method is the neural network processing method or the finite element analysis method.

7. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the electronic device implements the method according to any one of claims 1 to 5.

8. A computer program product, characterized in that, It includes a computer program, which when run causes the method according to any one of claims 1 to 5 to be executed.

Images