Composite material leading edge structure laying layer optimization design method, system and equipment and storage medium
By combining machine learning surrogate models and optimization algorithms, the problems of high computational cost, easy getting trapped in local optima, and strong dispersion of design variables in the ply design of composite material leading edge structures are solved, and efficient and intelligent ply optimization design is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- AVIC XAC COMMERCIAL AIRCRAFT CO LTD
- Filing Date
- 2025-12-18
- Publication Date
- 2026-04-14
AI Technical Summary
Existing technologies in composite material leading edge structure layup design are computationally expensive, prone to getting trapped in local optima, have highly discrete design variables, and lack intelligence, resulting in long optimization design cycles and poor results.
A proxy model is constructed using machine learning algorithms, and layering design is performed in combination with optimization algorithms. Through finite element analysis verification and database updates, intelligent and automated layering optimization is achieved.
It significantly reduces the number of finite element analyses, shortens the optimization design cycle, improves design quality, and enables rapid, lightweight, and intelligent layup design.
Smart Images
Figure CN121859631A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of aircraft wing design and relates to a method, system, device and storage medium for optimizing the plywood design of composite material leading edge structures. Background Technology
[0002] Thermoplastic composites, due to their advantages such as good toughness, recyclability, and short production cycle, have shown great application potential in high-end equipment fields such as aerospace and wind power. Leading edge structures, as key aerodynamic components, typically withstand complex loads such as external pressure, bird strikes, and vibrations; their performance directly affects the safety and efficiency of the overall equipment.
[0003] Currently, the ply layup design of composite material structures mostly employs experience-based trial-and-error methods or traditional optimization algorithms (such as genetic algorithms and particle swarm optimization). These methods have the following significant drawbacks: 1. High computational cost: Traditional optimization methods require tens of thousands of finite element analyses to evaluate the structural response (such as structural weight, maximum deformation, maximum stress, maximum strain, etc.) of different ply schemes. Each analysis is time-consuming, resulting in the entire optimization design cycle lasting several days or even weeks.
[0004] 2. Prone to getting trapped in local optima: In complex multi-peak design spaces, traditional optimization algorithms are prone to prematurely converging to local optima, making it difficult to obtain the globally optimal layering scheme.
[0005] 3. Highly discrete design variables: The design variables of the plying (such as the plying angle, which is usually a discrete value of 0°, ±45°, 90°, etc.) make the optimization problem highly nonlinear, which is difficult to handle by traditional continuous optimization methods and requires complex discretization processing.
[0006] 4. Lack of intelligence: Existing methods cannot learn from historical design data, and each time a new problem is faced, calculations must be started from scratch, and knowledge cannot be effectively accumulated and reused. Summary of the Invention
[0007] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method, system, device and storage medium for optimizing the ply layup design of composite material leading edge structures, which significantly reduces the number of finite element analysis calls, greatly shortens the optimization design cycle, and effectively improves the quality of optimization results, thereby realizing intelligent and automated ply layup design.
[0008] To achieve the above objectives, the present invention employs the following technical solution: A method for optimizing the layup design of composite material leading edge structures includes the following steps: S1: Establish a parametric geometric model based on the characteristic dimensions of the leading edge structure, define the layup design variables, optimization objectives and constraints, and obtain the finite element model of the composite material leading edge structure; S2: Generate initial ply scheme samples in the ply design variable space, perform finite element analysis on the initial ply scheme samples to obtain structural response data, and form sample pairs with the initial ply scheme samples and structural response data to form an initial training database. S3: Employ machine learning algorithms, using layered design variables as input and structural response data as output, and utilize the initial training database to train and construct a proxy model; S4: Use optimization algorithms to search in the approximate design space formed by the surrogate model to obtain candidate plying schemes that satisfy the optimization objective and constraints. S5: Submit the candidate ply scheme to the finite element analysis program for calculation to obtain the performance data of the new scheme. Add the data pair consisting of the candidate ply scheme and the performance data of the new scheme to the initial training database to obtain the updated training database. S6: Use the updated training database to update or retrain the surrogate model online, repeat S4 to S6 until the convergence condition is met, and output the optimal layup design scheme.
[0009] Optionally, the ply design variables include the total number of plies, the angle of each ply, the ply sequence, and the material type; the angle of each ply is a discrete variable selected from 0°, ±45°, ±45°, and 90°.
[0010] Optional optimization objectives include minimizing structural weight, maximizing a specific stiffness, maximizing a specific strength, or maximizing the first natural frequency; constraints include maximum allowable stress, maximum allowable strain, buckling factor, and manufacturing constraints.
[0011] Optionally, in S2, the Latin hypercube sampling method is used to generate initial ply scheme samples in the ply design variable space; the finite element analysis includes applying loads and boundary conditions to the finite element model of the composite material leading edge structure, performing static analysis or bird strike analysis, and calculating the corresponding structural response.
[0012] Optionally, the construction and training process in S3 includes: dividing the initial training database into a training set and a test set; training the neural network using the training set; testing the neural network using the test set; and obtaining the trained neural network model as a surrogate model after the neural network's prediction accuracy on both the training set and the test set meets the requirements.
[0013] Optionally, the optimization algorithm in S4 adopts a multi-objective genetic algorithm or a sequential quadratic programming method; the optimization algorithm recommends a new layering scheme based on the prediction results of the surrogate model.
[0014] Optionally, the optimal ply design scheme output in S6 includes detailed ply sequence, ply angle, ply thickness, and performance prediction values.
[0015] A composite material leading edge structure layup optimization design system includes: Finite element model construction module: used to establish a parametric geometric model based on the characteristic dimensions of the leading edge structure, define the ply design variables, optimization objectives and constraints, and obtain the finite element model of the composite material leading edge structure; Initial training data acquisition module: used to generate initial ply scheme samples in the ply design variable space, perform finite element analysis on the initial ply scheme samples, obtain structural response data, and form sample pairs with the initial ply scheme samples and structural response data to form the initial training database. Proxy model building module: Used to train and build a proxy model using machine learning algorithms, with the layer design variables as input and the structural response data as output, using an initial training database; Search module: Used to search in the approximate design space formed by the surrogate model using optimization algorithms to obtain candidate plying schemes that meet the optimization objectives and constraints; Training data update module: This module is used to submit candidate ply schemes to the finite element analysis program for calculation, obtain performance data of the new schemes, and add the data pairs consisting of candidate ply schemes and performance data of the new schemes to the initial training database to obtain the updated training database. Iteration module: Used to update or retrain the surrogate model online using the updated training database, repeatedly searching the module and iterating until the convergence condition is met, and outputting the optimal layup design scheme.
[0016] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the composite material leading edge structure layup optimization design method.
[0017] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the composite material leading edge structure layup optimization design method.
[0018] Compared with the prior art, the present invention has the following beneficial effects: This invention replaces most of the time-consuming finite element analysis by introducing machine learning algorithms to construct a high-precision surrogate model. The surrogate model is trained using initial samples to quickly approximate the complex nonlinear relationship between ply design variables and structural response, thereby improving computational efficiency by several orders of magnitude while ensuring prediction accuracy. Simultaneously, by combining optimization algorithms and active learning strategies, a rapid global search is performed in the approximate design space constructed by the surrogate model. The high-potential solutions found are verified by finite element analysis and then fed back to the training database to update the model online. This adaptive iterative mechanism not only enables the model to focus on high-performance design regions and effectively avoid getting trapped in local optima, but also solves the problem of traditional methods struggling to handle discrete variables and manufacturing constraints. Ultimately, it achieves rapid, lightweight, and intelligent design of composite material leading-edge structures while meeting mechanical performance constraints. Attached Figure Description
[0019] Figure 1 This is a flowchart of the composite material leading edge structure layup optimization design method of the present invention; Figure 2 This is a schematic diagram of a typical thermoplastic composite leading edge structure finite element model. Detailed Implementation
[0020] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.
[0021] The following disclosure provides many different embodiments or examples for implementing various structures of the invention. To simplify the disclosure, specific examples of components and arrangements are described below. These are merely examples and are not intended to limit the invention. Furthermore, reference numerals and / or letters may be repeated in different examples; such repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed. In addition, examples of various specific processes and materials are provided in this invention, but those skilled in the art will recognize the application of other processes and / or the use of other materials.
[0022] like Figure 1 As shown, the thermoplastic composite material leading edge structure layup optimization design method of the present invention includes the following steps: S1: Parametric Modeling and Design Space Definition: Parametric geometric modeling is established based on the characteristic dimensions of the leading edge structure.
[0023] Define the ply design variables, including: the total number of plies, the angle of each ply (discrete variable, such as selected from [0°, ±45°, 90°]), the ply sequence, and the optional material type.
[0024] Define optimization objectives, such as minimizing structural weight, maximizing specific stiffness / strength, or maximizing the first-order natural frequency.
[0025] Define constraints, such as maximum allowable stress / strain, buckling factor, and manufacturing constraints (such as symmetrical ply, consecutive identical ply angles not exceeding 4 layers, etc.).
[0026] like Figure 2 As shown, the composite material layers are modeled according to the layup design variables and constraints to obtain the finite element model of the composite material leading edge structure.
[0027] S2: Initial Sample Set Generation and Finite Element Analysis: Apply loads and boundary conditions to the finite element model of the composite material leading edge structure, and obtain the structural response (such as structural weight, maximum deformation, maximum stress, maximum strain, etc.) under the set loading pressure.
[0028] An experimental design method (such as Latin hypercube sampling) is used to generate a set of representative initial layup scheme samples within the defined design variable space.
[0029] For each initial sample, a parametric script drives the finite element analysis software to perform static, modal, or buckling analyses and calculate the corresponding structural response (such as weight, maximum displacement, maximum stress, natural frequency, etc.).
[0030] Each layup scheme (input variable) and its corresponding structural response (output performance) are paired to form a sample pair, thus creating the initial training database.
[0031] S3: Proxy Model Construction and Training: Divide the database into a training set and a test set. Use the training set to train the neural network and use the test set to test the neural network. When the accuracy of the neural network's prediction results on both the training set and the test set meets the requirements, the training is complete, and the trained neural network model is obtained as the proxy model.
[0032] Machine learning algorithms are used to train one or more high-precision surrogate models (or meta-models) with the layered design variables as input and the structural response as output, in order to replace computationally expensive finite element analysis.
[0033] The machine learning algorithm can be selected from: Gaussian process regression, deep neural networks, support vector machines, or combinations thereof. For example, a neural network can be trained to predict maximum stress, and another Gaussian process regression model can be trained to predict first-order natural frequencies.
[0034] The surrogate model is trained and validated using the database generated in step S2 to ensure that its prediction accuracy meets the requirements (e.g., coefficient of determination R² > 0.95).
[0035] S4: Optimization and iteration of the integrated agent model: The trained surrogate model is integrated with an optimization algorithm (such as genetic algorithm, sequential quadratic programming, etc.) to form an efficient optimization loop.
[0036] Optimization process: The optimization algorithm performs a rapid search within the approximate design space constructed by the surrogate model. Since the evaluation speed of the surrogate model is several orders of magnitude faster than finite element analysis, the optimization algorithm can evaluate millions of candidate solutions in an extremely short time.
[0037] In each optimization iteration, the optimization algorithm recommends one or a batch of promising new layup schemes based on the predictions of the surrogate model.
[0038] S5: Finite Element Verification and Database Update The new layup scheme recommended by the optimization algorithm is submitted to a real finite element analysis program for accurate calculation.
[0039] The newly obtained scheme-performance data pairs are added to the initial training database to expand the database.
[0040] The surrogate model is updated or retrained online using the updated database, making its prediction accuracy in the design region of interest increasingly higher.
[0041] S6: Convergence Judgment and Result Output: Using the neural network model obtained in step S5, for a given leading edge structure and convergence conditions, provide an optimized layup design for the composite material layer. This includes detailed layup sequence, angles, and thicknesses, and provides predicted performance values.
[0042] Repeat steps S4 and S5 until the convergence condition is met (e.g., the improvement of the optimal solution is less than a set threshold after several consecutive iterations, or the maximum number of iterations is reached).
[0043] Output the final optimal layup design scheme, including detailed layup sequence, angle and thickness, and provide its performance prediction values.
[0044] The method described in this embodiment has the following significant technical effects: Extremely high computational efficiency: By using machine learning proxy models to replace most of the time-consuming finite element analysis, the computation time for optimization design is reduced from the scale of weeks / days to the scale of hours / minutes, improving efficiency by tens or even hundreds of times.
[0045] Excellent optimization performance: Machine learning surrogate models can more accurately capture the complex nonlinear relationship between design variables and performance. Combined with powerful optimization algorithms, they can help to escape local optima and find global or near-global optimal solutions, thereby obtaining design solutions with better performance or lighter weight.
[0046] Adaptive and intelligent: The active learning strategy adopted in this invention (i.e., continuously updating the model with new data) enables the proxy model to adaptively focus on the high-performance design area, becoming more accurate with optimization, thus reflecting the characteristics of intelligent design.
[0047] Good engineering applicability: This method fully considers the discreteness and manufacturing constraints of composite material layup design and can be directly applied to engineering practice, providing a powerful tool for the rapid, lightweight, and high-performance design of thermoplastic composite leading edge structures.
[0048] Example: Take the lightweight design of the leading edge of a drone wing made of thermoplastic composite material as an example.
[0049] S1: The optimization objective is to minimize the weight of the leading edge structure while meeting the strength (maximum strain < allowable value) and bird strike requirements (impact damage < limit value) constraints under aerodynamic and bird strike loads. The design variable is the ply angle (each layer is selected from [0°, +45°, -45°, 90°]), and manufacturing constraints are applied to ensure that symmetrical plies and continuous unidirectional plies do not exceed 3 layers.
[0050] S2: 200 initial layup schemes were generated using Latin hypercube sampling. Static analysis and bird strike impact analysis were performed using Abaqus finite element software to obtain the weight, maximum strain, and impact damage of each scheme.
[0051] S3: Using Python's Scikit-learn library, construct a deep neural network as a surrogate model. Take the ply angle and number of ply layers as inputs, and the weight, maximum strain, and impact damage as outputs, and train it with 200 samples.
[0052] S4 & S5: The trained neural network surrogate model is integrated with a multi-objective genetic algorithm (such as NSGA-II) for optimization. Each generation of the genetic algorithm generates 50 candidate solutions, which are quickly evaluated by the surrogate model. The five most promising solutions at the Pareto frontier are selected for real-world Abaqus analysis, and the results are added to the database to update the neural network model.
[0053] S6: After 100 iterations, the optimization results converged. The final solution is 15% lighter than the initial empirical design, while fully satisfying all mechanical performance constraints. The entire optimization process was completed within 24 hours, whereas traditional methods would have taken several weeks.
[0054] The following are embodiments of the apparatus of the present invention, which can be used to execute embodiments of the method of the present invention. For details not omitted in the apparatus embodiments, please refer to the embodiments of the method of the present invention.
[0055] In another embodiment of the present invention, a composite material leading edge structure ply optimization design system is provided. This composite material leading edge structure ply optimization design system can be used to implement the above-mentioned composite material leading edge structure ply optimization design method. Specifically, the composite material leading edge structure ply optimization design system includes a finite element model construction module, an initial training data acquisition module, a surrogate model construction module, a search module, a training data update module, and an iteration module.
[0056] The finite element model construction module is used to establish a parametric geometric model based on the characteristic dimensions of the leading edge structure, define the layup design variables, optimization objectives and constraints, and obtain the finite element model of the composite material leading edge structure.
[0057] The initial training data acquisition module is used to generate initial ply scheme samples in the ply design variable space, perform finite element analysis on the initial ply scheme samples to obtain structural response data, and form sample pairs with the initial ply scheme samples and structural response data to form the initial training database.
[0058] The surrogate model building module is used to train and build a surrogate model using machine learning algorithms, with the layered design variables as input and the structural response data as output, using an initial training database.
[0059] The search module is used to search in the approximate design space formed by the surrogate model using optimization algorithms to obtain candidate plying schemes that meet the optimization objectives and constraints.
[0060] The training data update module is used to submit candidate plying schemes to the finite element analysis program for calculation, obtain performance data of the new scheme, and add the data pair consisting of the candidate plying scheme and the performance data of the new scheme to the initial training database to obtain the updated training database.
[0061] The iterative module is used to update or retrain the surrogate model online using the updated training database. The search module is repeated until the convergence condition is met, and the optimal layup design scheme is output.
[0062] In another embodiment of the present invention, a terminal device is provided, comprising a processor and a memory. The memory stores a computer program, the computer program including program instructions, and the processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), or field-programmable gate arrays (FPGAs). Gate Array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., are the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to realize the corresponding method flow or corresponding function; the processor described in this embodiment of the invention can be used for the operation of the composite material leading edge structure ply optimization design method, including: S1: establishing a parametric geometric model based on the characteristic dimensions of the leading edge structure, defining ply design variables, optimization objectives and constraints, and obtaining a finite element model of the composite material leading edge structure; S2: generating an initial ply scheme sample in the ply design variable space, performing finite element analysis on the initial ply scheme sample, obtaining structural response data, and setting the initial ply... S3: Using a machine learning algorithm, with the ply design variables as input and the structural response data as output, a surrogate model is trained and constructed using the initial training database. S4: An optimization algorithm is used to search in the approximate design space formed by the surrogate model to obtain candidate ply schemes that meet the optimization objectives and constraints. S5: The candidate ply schemes are submitted to the finite element analysis program for calculation to obtain the performance data of the new schemes. The data pairs consisting of the candidate ply schemes and the performance data of the new schemes are added to the initial training database to obtain the updated training database. S6: The surrogate model is updated or retrained online using the updated training database. S4 to S6 are repeated until the convergence condition is met, and the optimal ply design scheme is output.
[0063] In another embodiment, the present invention also provides a computer-readable storage medium (Memory), which is a memory device in a terminal device for storing programs and data. It is understood that the computer-readable storage medium here may include both the built-in storage medium in the terminal device and extended storage media supported by the terminal device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, the storage space also stores one or more instructions suitable for loading and execution by a processor, which may be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here may be high-speed RAM or non-volatile memory, such as at least one disk storage device.
[0064] One or more instructions stored in a computer-readable storage medium can be loaded and executed by a processor to implement the corresponding steps of the composite material leading edge structure ply optimization design method in the above embodiments; one or more instructions in the computer-readable storage medium are loaded and executed by the processor in the following steps: S1: Establish a parametric geometric model based on the characteristic dimensions of the leading edge structure, define ply design variables, optimization objectives and constraints, and obtain a finite element model of the composite material leading edge structure; S2: Generate initial ply scheme samples in the ply design variable space, perform finite element analysis on the initial ply scheme samples, obtain structural response data, and form sample pairs with the initial ply scheme samples and structural response data to form an initial training database. S3: Using a machine learning algorithm, with ply design variables as input and structural response data as output, a surrogate model is trained using the initial training database. S4: An optimization algorithm is used to search the approximate design space formed by the surrogate model to obtain candidate ply schemes that satisfy the optimization objective and constraints. S5: The candidate ply schemes are submitted to a finite element analysis program for calculation to obtain performance data of the new schemes. The data pairs consisting of the candidate ply schemes and the performance data of the new schemes are added to the initial training database to obtain an updated training database. S6: The surrogate model is updated or retrained online using the updated training database. S4 to S6 are repeated until the convergence condition is met, and the optimal ply design scheme is output.
[0065] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0066] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0067] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0068] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0069] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.
[0070] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.
[0071] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.
[0072] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.
[0073] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.
[0074] It should be understood that the above description is for illustrative purposes and not for limitation. Many embodiments and applications beyond the provided examples will be apparent to those skilled in the art upon reading the above description. Therefore, the scope of this patent should not be determined by reference to the above description, but rather by reference to the foregoing claims and the full scope of their equivalents. For purposes of completeness, all articles and references, including patent applications and publications, are incorporated herein by reference. The omission of any aspect of the subject matter disclosed herein in the foregoing claims is not intended as a waiver of that subject matter, nor should it be construed as an indication that the applicant has not considered that subject matter as part of the disclosed inventive subject matter.
Claims
1. A method for optimizing the layup design of composite material leading edge structures, characterized in that, Includes the following processes: S1: Establish a parametric geometric model based on the characteristic dimensions of the leading edge structure, define the layup design variables, optimization objectives and constraints, and obtain the finite element model of the composite material leading edge structure; S2: Generate initial ply scheme samples in the ply design variable space, perform finite element analysis on the initial ply scheme samples to obtain structural response data, and form sample pairs with the initial ply scheme samples and structural response data to form an initial training database. S3: Employ machine learning algorithms, using layered design variables as input and structural response data as output, and utilize the initial training database to train and construct a proxy model; S4: Use optimization algorithms to search in the approximate design space formed by the surrogate model to obtain candidate plying schemes that satisfy the optimization objective and constraints. S5: Submit the candidate ply scheme to the finite element analysis program for calculation to obtain the performance data of the new scheme. Add the data pair consisting of the candidate ply scheme and the performance data of the new scheme to the initial training database to obtain the updated training database. S6: Use the updated training database to update or retrain the surrogate model online, repeat S4 to S6 until the convergence condition is met, and output the optimal layup design scheme.
2. The composite material leading edge structure layup optimization design method according to claim 1, characterized in that, The ply design variables include the total number of plies, the angle of each ply, the ply sequence, and the material type; the angle of each ply is a discrete variable selected from 0°, ±45°, ±45°, and 90°.
3. The composite material leading edge structure layup optimization design method according to claim 1, characterized in that, The optimization objectives include minimizing structural weight, maximizing a specific stiffness, maximizing a specific strength, or maximizing the first-order natural frequency; the constraints include maximum allowable stress, maximum allowable strain, buckling factor, and manufacturing constraints.
4. The composite material leading edge structure layup optimization design method according to claim 1, characterized in that, In S2, the Latin hypercube sampling method is used to generate initial ply scheme samples in the ply design variable space; the finite element analysis includes applying loads and boundary conditions to the finite element model of the composite material leading edge structure, performing static analysis or bird strike analysis, and calculating the corresponding structural response.
5. The composite material leading edge structure layup optimization design method according to claim 1, characterized in that, The construction and training process in S3 includes: dividing the initial training database into a training set and a test set; training the neural network using the training set; testing the neural network using the test set; and obtaining the trained neural network model as a surrogate model after the neural network's prediction accuracy on both the training set and the test set meets the requirements.
6. The composite material leading edge structure layup optimization design method according to claim 1, characterized in that, The optimization algorithm in S4 adopts a multi-objective genetic algorithm or a sequential quadratic programming method; the optimization algorithm recommends a new layering scheme based on the prediction results of the surrogate model.
7. The composite material leading edge structure layup optimization design method according to claim 1, characterized in that, The optimal ply design scheme output in S6 includes detailed ply sequence, ply angle, ply thickness, and performance prediction values.
8. A composite material leading edge structure layup optimization design system, characterized in that, include: Finite element model construction module: used to establish a parametric geometric model based on the characteristic dimensions of the leading edge structure, define the ply design variables, optimization objectives and constraints, and obtain the finite element model of the composite material leading edge structure; Initial training data acquisition module: used to generate initial ply scheme samples in the ply design variable space, perform finite element analysis on the initial ply scheme samples, obtain structural response data, and form sample pairs with the initial ply scheme samples and structural response data to form the initial training database. Proxy model building module: Used to train and build a proxy model using machine learning algorithms, with the layer design variables as input and the structural response data as output, using an initial training database; Search module: Used to search in the approximate design space formed by the surrogate model using optimization algorithms to obtain candidate plying schemes that meet the optimization objectives and constraints; Training data update module: This module is used to submit candidate ply schemes to the finite element analysis program for calculation, obtain performance data of the new schemes, and add the data pairs consisting of candidate ply schemes and performance data of the new schemes to the initial training database to obtain the updated training database. Iteration module: Used to update or retrain the surrogate model online using the updated training database, repeatedly searching the module and iterating until the convergence condition is met, and outputting the optimal layup design scheme.
9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the composite material leading edge structure layup optimization design method as described in any one of claims 1 to 7.
10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the composite material leading edge structure layup optimization design method as described in any one of claims 1 to 7.