Aircraft structure optimization method, system, device and medium

Abstract

The present invention provides an aircraft structure optimization method, system, device and storage medium, belonging to the technical field of aircraft crashworthiness design. The method comprises the following steps: constructing a finite element model of an aircraft fuselage frame section, analyzing the load transfer path during the aircraft crash impact, establishing an energy absorption performance decomposition model, and constructing a hierarchical crashworthiness level evaluation system for the fuselage frame section based on the energy absorption performance decomposition model; constructing an initial data set, and performing data enhancement on the initial data set using an adversarial neural network algorithm to obtain an enhanced data set; optimizing the initial weights and threshold parameters of a predictive neural network model using a parameter optimization algorithm to obtain an improved predictive neural network model, inputting data in the enhanced data set into the predictive neural network model to obtain an optimal design parameter combination; and optimizing the aircraft structure based on the optimal design parameter combination. The present invention can obtain an aircraft structure with better crashworthiness.

Description

Technical Field

[0001] The present invention belongs to the technical field of aircraft crashworthiness design, and in particular relates to an aircraft structure optimization method, system, equipment and storage medium. Background Art

[0002] Studying the crashworthiness of aircraft is a crucial topic for protecting occupant safety and embodies the core of aircraft passive safety. The performance of aircraft fuselage structures under crash impact involves a complex interplay of material, geometric, and boundary nonlinearities, as well as the coupling between multiple structural design parameters, making independent structural performance predictions difficult.

[0003] Existing research on predicting impact energy dissipation during aircraft fuselage structural impacts has largely relied heavily on commercial finite element simulation programs. However, collecting the energy absorption characteristics of aircraft fuselage structures requires a time-consuming process involving both experimental and simulation calculations, resulting in high computational costs. Consequently, some studies have used ANN neural network algorithms to predict aircraft crash performance. However, this method suffers from insufficient raw sample data, preventing the generation of accurate aircraft structural design parameters and resulting in low crash performance prediction accuracy. Summary of the Invention

[0004] In order to overcome the above-mentioned deficiencies in the prior art, the present invention provides an aircraft structure optimization method, comprising the following steps:

[0005] Construct a finite element model of the aircraft fuselage frame, analyze the load transfer path during an aircraft crash, establish an energy absorption performance decomposition model for the fuselage frame's lower cargo hold and passenger cabin areas, and construct a layered crashworthiness evaluation system based on the energy absorption performance decomposition model.

[0006] Construct an initial dataset of key variables for aircraft structural design and use the Generative Adversarial Network (GAN) algorithm to perform data augmentation on the initial dataset to obtain an enhanced dataset.

[0007] The initial weights and threshold parameters of the prediction neural network model BP are optimized through a parameter optimization algorithm to obtain an improved prediction neural network model. The data in the enhanced data set is input into the improved prediction neural network model. With the goal of maximizing the evaluation indicators in the evaluation system, the improved prediction neural network model is iterated to obtain the optimal design parameter combination; and the aircraft structure is optimized according to the optimal design parameter combination.

[0008] Preferably, the step of constructing a layered crashworthiness evaluation system for fuselage frame sections based on the energy absorption performance decomposition model comprises the following steps:

[0009] Decomposing the kinetic energy dissipation process of the aircraft fuselage frame section in the lower area of ​​the cargo hold and the lower area of ​​the passenger cabin, allocating the energy absorption and the deformation stiffness and residual bearing strength allowed by the structural design to each component, and performing simulation analysis on the finite element model;

[0010] The load-displacement response curve is extracted from the simulation analysis results. The peak load, load efficiency and specific energy absorption rate are collected according to the load-displacement response curve. The peak load, load efficiency and specific energy absorption rate are quantified as comprehensive evaluation indicators to obtain an evaluation system for the layered crashworthiness level of the fuselage frame section.

[0011] Preferably, the construction of the initial data set of key variables for aircraft structural design is specifically as follows: based on nonlinear correlation analysis, key variables in the structural design process are selected, and a sample space of key variables for structural design is established to form an initial data set; the key variables include the structural thickness and lay-up angle of aircraft components.

[0012] Preferably, using a GAN (Generative Adversarial Network) algorithm to perform data augmentation on the initial dataset to obtain an enhanced dataset comprises the following steps:

[0013] The data in the initial data set is input into the adversarial neural network algorithm GAN, and the input data is iteratively processed through the generative model and discriminant model of GAN to expand the data in the initial data set. When the discrimination probability of the generative model and the discriminant model reaches a threshold of 0.5, an enhanced data set is obtained.

[0014] Preferably, the construction of the finite element model of the aircraft fuselage frame section is specifically: establishing a numerical model of the aircraft fuselage frame section, and simplifying the numerical model, simplifying the non-main load-bearing superstructure of the aircraft fuselage frame section into mass nodes and beam units, and building the finite element model of the aircraft fuselage frame section.

[0015] Preferably, before using the adversarial neural network algorithm GAN to perform data enhancement operation on the initial data set, the method also includes normalizing the data in the initial data set.

[0016] The present invention also provides an aircraft structure optimization system, comprising:

[0017] An evaluation system acquisition module is used to build a finite element model of the aircraft fuselage frame, analyze the load transfer path during an aircraft crash, establish an energy absorption performance decomposition model for the fuselage frame's lower cargo hold and passenger cabin areas, and construct a hierarchical crashworthiness evaluation system based on the energy absorption performance decomposition model.

[0018] The dataset enhancement module is used to construct an initial dataset of key variables for aircraft structural design. The GAN algorithm is used to perform data enhancement on the initial dataset to obtain an enhanced dataset.

[0019] The structure optimization module is used to optimize the initial weights and threshold parameters of the prediction neural network model BP through a parameter optimization algorithm to obtain an improved prediction neural network model, input the data in the enhanced data set into the improved prediction neural network model, and iterate the improved prediction neural network model with the goal of maximizing the evaluation indicators in the evaluation system to obtain the optimal design parameter combination; and optimize the aircraft structure according to the optimal design parameter combination.

[0020] The present invention also provides a computer device, comprising a memory and a processor; the memory stores a computer program, and the processor is used to run the computer program in the memory to execute the aircraft structure optimization method.

[0021] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program is suitable for being loaded by a processor to execute the aircraft structure optimization method.

[0022] The aircraft structure optimization method provided by the present invention has the following beneficial effects:

[0023] The present invention enhances the initial data set to obtain an enhanced data set, which can make up for the defect of insufficient original sample data volume, enhance the generalization ability of the prediction model, effectively realize the maximization of the energy absorption of the fuselage structure, thereby effectively improving the crash energy absorption characteristics of the aircraft fuselage structure, and providing support for the anti-crash structure design of the aircraft; the present invention constructs an energy absorption performance decomposition model for the lower area of ​​the cargo hold and the lower area of ​​the passenger cabin of the fuselage frame section, and constructs an evaluation system for the layered crashworthiness level of the fuselage frame section according to the energy absorption performance decomposition model, which can comprehensively consider the structural strength, stiffness and energy absorption and other crashworthiness performance indicators of each sub-area of ​​the aircraft fuselage structure; by constructing an improved prediction neural network model, the data in the enhanced data set are input into the improved prediction neural network model, and with the goal of maximizing the evaluation indicators in the evaluation system, the improved prediction neural network model is iterated to obtain the optimal design parameter combination, so that the improved prediction neural network model can be applied to the prediction of the aircraft crash process, and the optimal design parameter combination for optimizing the aircraft structure is obtained, thereby realizing the optimization of the aircraft structure.

[0024] The present invention combines the finite element method with the neural network model to carry out hybrid prediction, uses less finite element method calculations, and achieves a significant reduction in calculation time and calculation cost, making it possible to carry out multi-parameter optimization of aircraft crashworthiness design in a larger design space, and realizing a multi-level decomposition, rapid prediction and optimization method for aircraft crashworthiness performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] To more clearly illustrate the embodiments of the present invention and its design, the following briefly introduces the drawings required for this embodiment. The drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be derived from these drawings without inventive effort.

[0026] Figure 1 1 is a flow chart of an aircraft structure optimization method according to an embodiment of the present invention;

[0027] Figure 2 It is a simplified finite element model of the aircraft fuselage frame;

[0028] Figure 3 Optimize the flow chart for design parameters;

[0029] Figure 4 This is a comparison chart of the total error.

[0030] Description of reference numerals:

[0031] 1-Connecting truss, 2-Aircraft seat, 3-Cabin floor crossbeam, 4-Cabin floor support column, 5-Cargo hold floor, 6-Cargo hold floor support column, 7-Aircraft skin. DETAILED DESCRIPTION

[0032] In order to enable those skilled in the art to better understand the technical solution of the present invention and to be able to implement it, the present invention is described in detail below with reference to the accompanying drawings and specific embodiments. The following embodiments are only used to more clearly illustrate the technical solution of the present invention and are not intended to limit the scope of protection of the present invention.

[0033] Example

[0034] The present invention provides an aircraft structure optimization method, specifically Figure 1 As shown, the following steps are included:

[0035] Step 1: Build a finite element model of the aircraft fuselage frame, analyze the load transfer path during the aircraft crash impact, establish an energy absorption performance decomposition model for the lower area of ​​the cargo hold and the lower area of ​​the passenger cabin of the fuselage frame, and construct an evaluation system for the layered crashworthiness level of the fuselage frame based on the energy absorption performance decomposition model.

[0036] Specifically, the energy absorption performance decomposition model is constructed through the following steps:

[0037] (1) Structural partitioning: The fuselage frame section of the aircraft is divided into the lower sub-area of ​​the cargo hold (including the cargo hold floor support column structure, cargo hold floor structure, etc.) and the lower sub-area of ​​the passenger cabin (including the cabin floor support column structure, cabin floor crossbeam structure, etc.).

[0038] (2) Energy distribution: Based on the load transfer path, the total energy absorbed by the aircraft during the crash is distributed to the components of each sub-area according to the functional weight, forming the component-level energy absorption target value E i .

[0039] (3) Constraint setting: set the allowable deformation stiffness threshold (δ max,i ) and the residual bearing strength threshold (σ min,i ).

[0040] (4) Verification mechanism: Use numerical simulation methods to verify whether the energy absorbed by the component meets E i ≥E 目标 , and at the same time satisfy δ i ≤δ max,i and σ i ≥σ min,i .

[0041] First, a numerical model of the aircraft fuselage frame is established and simplified. The non-main load-bearing upper structure of the aircraft fuselage frame is simplified into mass nodes and beam units, and a finite element model of the aircraft fuselage frame is constructed. Figure 2 As shown, it includes a connecting truss 1, an aviation seat 2, a cabin floor crossbeam 3, a cabin floor supporting column 4, a cargo floor 5, a cargo floor supporting column 6 and an aircraft skin 7.

[0042] Secondly, the kinetic energy dissipation process of the aircraft fuselage frame section during a crash is decomposed in the lower area of ​​the cargo hold and the lower area of ​​the passenger cabin. The energy absorption and the deformation stiffness and residual bearing strength allowed by the structural design are distributed to each component to obtain an energy absorption performance decomposition model, which is then simulated and analyzed.

[0043] Load-displacement response curves were extracted from the simulation analysis results. Peak load, load efficiency, and specific energy absorption rate were then collected based on these curves. These peak load, load efficiency, and specific energy absorption rate were quantified as comprehensive evaluation indicators, resulting in a hierarchical crashworthiness evaluation system for the fuselage frame. These comprehensive evaluation indicators specifically include the energy absorption rate, specific energy absorption rate, peak load, mean load, load efficiency, deformation stiffness, and structural displacement of the fuselage frame.

[0044] Step 2: Construct an initial dataset of key variables for aircraft structural design, and use the adversarial neural network algorithm GAN to perform data augmentation on the initial dataset to obtain an enhanced dataset.

[0045] First, key variables in the structural design process are selected, and the impact kinetic energy is reasonably allocated based on the Complex Proportional Assessment (COPRAS) method. The sample space of key structural design variables is established to form the initial data set.

[0046] The key variables of the fuselage frame design are selected, and the simulation matrix design of multiple working conditions is carried out based on the Design of Experiments (DOE) method to obtain the original input matrix.

[0047] Perform nonlinear correlation analysis and parameter sensitivity analysis on the original input matrix. Through nonlinear correlation analysis, the relationship between features is mined, and the features with obvious correlation are selected to form the initial data set together with the original input matrix. The data in the initial data set is normalized. The normalization formula is as follows:

[0048] ;

[0049] Where, X normal is the normalized input matrix; X is the original input matrix; X min is the minimum value in the original input matrix; X max is the maximum value in the original input matrix; the original input matrix includes the ply angle, structural thickness and some material parameters of the composite material component.

[0050] Secondly, under small sample conditions, the normalized input matrix is ​​input into the Generative Adversarial Network (GAN) algorithm. The input data is iteratively processed through the GAN's generative model and discriminant model, and the normalized input matrix is ​​fully expanded. When the discrimination probability of the generative model and the discriminant model reaches the 0.5 threshold, a large amount of reliable data is generated from the small data set to obtain an enhanced data set.

[0051] In addition, random noise data is introduced into the GAN model to prevent overfitting of the model and enhance the generalization ability of the model.

[0052] Step 3: Optimize the initial weights and threshold parameters of the prediction neural network model BP through the parameter optimization algorithm to obtain an improved prediction neural network model. Input the data in the enhanced data set into the improved prediction neural network model. With the goal of maximizing the evaluation indicators in the evaluation system, iterate the improved prediction neural network model to obtain the optimal design parameter combination; optimize the aircraft structure based on the optimal design parameter combination.

[0053] The parameter optimization algorithm can be a global optimization algorithm such as Particle Swarm Optimization (PSO), Simulated Annealing (SA), or Greedy. In this embodiment, a Genetic Algorithm (GA) is specifically selected to optimize the initialization parameters of the Backpropagation Neural Network (BP) model. The optimal initial weight parameters are selected based on the fitness function and transferred to the BP model to obtain an improved GA-BP model. The data from the enhanced dataset is then input into the GA-BP model. The BP neural network model is then iterated through numerous iterations to find the optimal solution within the feasible domain. The optimal results and design parameters are then searched for, ultimately achieving a lightweight layout and controllable energy absorption design for the fuselage frame lower structure.

[0054] The specific process is as follows Figure 3 As shown, the input vector in the enhanced data set is input into the prediction neural network model BP after data processing to initialize the BP network and assign initial weights. After initialization encoding through the GA genetic algorithm, chromosome selection, crossover, and mutation operations are performed based on the fitness function. The population is iteratively updated, and the optimal fitness value and average fitness value of each iteration are collected. When the number of runs reaches the maximum number of iterations set by the genetics, it is determined whether the termination condition is met. If the termination condition is not met, the chromosome is returned to be reselected, and the iteration is performed again according to the previous steps until the BP network initialization weight value corresponding to the optimal fitness value (i.e., the optimal weight) is screened out. The obtained optimal weight is input into the prediction neural network model BP, iterative calculation and error analysis are performed, and the weight is updated. Then it is determined whether the training target is terminated. If so, the output vector of the prediction neural network model BP is retained, and the prediction neural network model BP is used to realize data prediction in the prediction space. If the termination condition is not met, the optimal weight is returned to be reselected, and the iteration is performed again according to the previous steps until the preset number of iterations is reached or the error threshold is passed.

[0055] In addition, in this embodiment, the activation function of GA-BP is the SELU algorithm improved based on the RELU algorithm, and the error function of GA-BP is a combination of mean absolute error (MAE), root mean square error (RMSE) and mean absolute error rate (MAPE).

[0056] The present invention also provides an aircraft structure optimization system, comprising an evaluation system acquisition module, a data set enhancement module, a parameter optimization module, and a structure optimization module. The evaluation system acquisition module is used to build a finite element model of an aircraft fuselage frame, analyze the load transfer path during an aircraft crash, establish an energy absorption performance decomposition model for the lower area of ​​the cargo hold and the lower area of ​​the passenger cabin of the fuselage frame, and construct an evaluation system for the hierarchical crashworthiness level of the fuselage frame based on the energy absorption performance decomposition model; the data set enhancement module is used to build an initial data set of key variables for aircraft structural design, and use a GAN (Government Adversarial Network) algorithm to perform data enhancement on the initial data set to obtain an enhanced data set; the structure optimization module is used to optimize the initial weights and threshold parameters of a prediction neural network model (BP) using a parameter optimization algorithm to obtain an improved prediction neural network model, input data in the enhanced data set into the improved prediction neural network model, iterate the improved prediction neural network model with the goal of maximizing the evaluation index in the evaluation system, and obtain an optimal design parameter combination; the aircraft structure is optimized based on the optimal design parameter combination.

[0057] The present invention also provides a computer device, comprising a memory and a processor; the memory stores a computer program, and the processor is used to run the computer program in the memory to execute an aircraft structure optimization method.

[0058] The present invention also provides a computer-readable storage medium storing a computer program suitable for loading by a processor to execute the aircraft structure optimization method.

[0059] Compared with the traditional single BP algorithm prediction, the GA-BP algorithm effectively reduces the initial error of the prediction network by optimizing the initial network weights of BP during the iteration process. The error reduction rate is improved, and it is easier to reach the convergence state. The final error of the GA-BP algorithm is smaller and the prediction effect is better. Figure 4 shown.

[0060] Compared to traditional aircraft structural optimization methods, which rely solely on iterative calculations using the finite element method, the optimization method of the present invention combines the traditional finite element method with multiple neural network models to perform hybrid optimization. This method uses fewer finite element method calculations, significantly reduces computational time and cost, and makes it possible to carry out multi-parameter optimization of aircraft crashworthiness design within a larger design space.

[0061] The above embodiments are only preferred specific implementation methods of the present invention, and the protection scope of the present invention is not limited thereto. Any simple changes or equivalent replacements of the technical solutions that can be obviously obtained by any technician familiar with the field within the technical scope disclosed in the present invention fall within the protection scope of the present invention.

Claims

1. A method for optimizing an aircraft structure, characterized in that: The steps include: Construct a finite element model of the aircraft fuselage frame, analyze the load transfer path during an aircraft crash, establish an energy absorption performance decomposition model for the fuselage frame's lower cargo hold and passenger cabin areas, and construct a layered crashworthiness evaluation system based on the energy absorption performance decomposition model. Construct an initial dataset of key variables for aircraft structural design and use the Generative Adversarial Network (GAN) algorithm to perform data augmentation on the initial dataset to obtain an enhanced dataset. optimizing the initial weights and threshold parameters of the prediction neural network model BP using a parameter optimization algorithm to obtain an improved prediction neural network model, inputting the data in the enhanced data set into the improved prediction neural network model, iterating the improved prediction neural network model with the goal of maximizing the evaluation index in the evaluation system to obtain an optimal design parameter combination; and optimizing the aircraft structure based on the optimal design parameter combination; The evaluation system for the layered crashworthiness level of the fuselage frame section is constructed based on the energy absorption performance decomposition model, comprising the following steps: Decomposing the kinetic energy dissipation process of the aircraft fuselage frame section in the lower area of ​​the cargo hold and the lower area of ​​the passenger cabin, allocating the energy absorption and the deformation stiffness and residual bearing strength allowed by the structural design to each component, and performing simulation analysis on the finite element model; The load-displacement response curve is extracted from the simulation analysis results. The peak load, load efficiency and specific energy absorption rate are collected according to the load-displacement response curve. The peak load, load efficiency and specific energy absorption rate are quantified as comprehensive evaluation indicators to obtain an evaluation system for the layered crashworthiness level of the fuselage frame section.

2. The aircraft structure optimization method according to claim 1, characterized in that: The initial data set for constructing key variables in aircraft structural design is specifically constructed by selecting key variables in the structural design process based on nonlinear correlation analysis, establishing a sample space for the key variables in structural design, and forming an initial data set; the key variables include the structural thickness and layup angle of aircraft components.

3. The aircraft structure optimization method according to claim 2, characterized in that: The adversarial neural network algorithm GAN is used to perform data augmentation on the initial dataset to obtain an enhanced dataset, including the following steps: The data in the initial data set is input into the adversarial neural network algorithm GAN, and the input data is iteratively processed through the generative model and discriminant model of GAN to expand the data in the initial data set. When the discrimination probability of the generative model and the discriminant model reaches a threshold of 0.5, an enhanced data set is obtained.

4. The aircraft structure optimization method according to claim 1, characterized in that: The method of constructing a finite element model of an aircraft fuselage frame section specifically comprises: establishing a numerical model of the aircraft fuselage frame section, simplifying the numerical model, simplifying the non-main load-bearing superstructure of the aircraft fuselage frame section into mass nodes and beam units, and constructing a finite element model of the aircraft fuselage frame section.

5. The aircraft structure optimization method according to claim 1, characterized in that: Before using the adversarial neural network algorithm GAN to perform data enhancement operations on the initial data set, the method also includes normalizing the data in the initial data set.

6. An aircraft structure optimization system, characterized in that: include: An evaluation system acquisition module is used to build a finite element model of the aircraft fuselage frame, analyze the load transfer path during an aircraft crash, establish an energy absorption performance decomposition model for the fuselage frame's lower cargo hold and passenger cabin areas, and construct a hierarchical crashworthiness evaluation system based on the energy absorption performance decomposition model. The dataset enhancement module is used to construct an initial dataset of key variables for aircraft structural design. The GAN algorithm is used to perform data enhancement on the initial dataset to obtain an enhanced dataset. The structure optimization module is used to optimize the initial weights and threshold parameters of the prediction neural network model BP through a parameter optimization algorithm to obtain an improved prediction neural network model, input the data in the enhanced data set into the improved prediction neural network model, and iterate the improved prediction neural network model with the goal of maximizing the evaluation indicators in the evaluation system to obtain the optimal design parameter combination; and optimize the aircraft structure according to the optimal design parameter combination.

7. A computer device, characterized in that: The invention comprises a memory and a processor; the memory stores a computer program, and the processor is used to run the computer program in the memory to execute the aircraft structure optimization method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and the computer program is suitable for being loaded by a processor to execute the aircraft structure optimization method according to any one of claims 1 to 5.

Images