Thin-wall reinforced structure design method and system considering manufacturing characteristics

Abstract

The invention discloses a thin-wall reinforced structure design method and system considering manufacturing characteristics, and belongs to the technical field of aerospace structure main bearing component analysis design, and the method comprises the steps: building a fine model and a simplified model, carrying out point sampling on structure parameters in a design space, and carrying out numerical analysis on each sample point to obtain performance indexes of the sample points; constructing a data set according to the performance index data of the sample points, training a mathematical model to represent the relationship between the manufacturing features and key performance indexes, constructing an optimization problem, calling the mathematical model to predict the influence of the manufacturing features on an optimization target and constraints, and solving the optimization problem to obtain an actual design scheme comprehensively considering the manufacturing features; according to the thin-wall reinforced structure design method and system considering the manufacturing characteristics, the influence of the manufacturing characteristics is represented through the mathematical regression model based on data driving, and optimal design capable of accurately considering the bearing capacity and the quality of the structure containing the manufacturing characteristics is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of analysis and design of main load-bearing components of aerospace structures, and particularly to a design method and system for thin-walled stiffened structures considering manufacturing features. Background Art

[0002] Due to advantages such as high specific strength and specific stiffness, stiffened thin-shell structures are key load-bearing components such as fuel tanks of launch vehicles. They mainly bear axial pressure, and typical failure modes include buckling instability and strength failure. In the design of aerospace equipment, it is of great significance to improve the structural load-bearing efficiency and achieve lightweighting, which can reduce the launch cost, increase the payload ratio, and enhance the economy and reliability of aerospace missions. However, the design and optimization of stiffened thin-shell structures face multiple challenges.

[0003] On the one hand, the load-bearing capacity of thin-shell structures is closely related to manufacturing features, such as fine features like chamfers, welds, openings, and geometric defects. These manufacturing features are inevitable during the processing. If not accurately considered during design, significant deviations will occur between the finite element simulation model and the actual structure. Ignoring them may misjudge the structural load-bearing performance, resulting in the underutilization of the structural potential. Conservative design is required, but this will increase the weight redundancy, deviating from the original intention of lightweighting.

[0004] On the other hand, the geometric features of stiffened thin-walled structures are complex, and their finite element simulation models have a large number of degrees of freedom. In the case of high-precision requirements, the model scale is even larger. For example, the number of degrees of freedom of a high-fidelity simulation model of a large-diameter grid-stiffened cylindrical shell structure can reach hundreds of millions or even billions. Moreover, analyzing the load-bearing performance (such as post-buckling behavior) of thin-shell structures under complex loads requires high-time-consuming nonlinear calculations. In the optimization design, a large number of repeated simulation analyses are required, making it difficult to balance efficiency and precision, and increasing the difficulty of design analysis.

[0005] Therefore, there is an urgent need to develop a design method for thin-walled stiffened structures that takes into account both precision and efficiency and manufacturing features. It is necessary to quantify the influence of manufacturing features on the structure and introduce them into the design process, solve the contradiction between precision and efficiency, and also have high applicability and operability to meet the requirements of complex thin-walled cylindrical shell structures in aerospace, improve the lightweight and refinement levels of structural design, and achieve the design goals of high performance and high reliability of aerospace equipment. Summary of the Invention

[0006] The purpose of the present invention is to provide a design method and system for thin-walled stiffened structures considering manufacturing features, solve the above-mentioned design problems of aerospace cylindrical shell structures, and achieve the refined structural design of thin-walled cylindrical shell structures with structural manufacturing features such as chamfers.

[0007] To achieve the above purpose, the present invention provides a design method for thin-walled stiffened structures considering manufacturing features, including the following steps:

[0008] S1. Establish a structural model of the aerospace cylinder structure with fine manufacturing features and a simplified model without features. Conduct experimental design through sampling techniques, and perform sampling of structural parameters at points within the design space.

[0009] S2. Conduct numerical analysis on each sample point to obtain its performance indicators, and construct a data set based on the performance indicator data of the sample points.

[0010] S3. Train a mathematical model to characterize the relationship between manufacturing features and key performance indicators.

[0011] S4. Construct an optimization problem, and call the mathematical model to predict the impact of manufacturing features on the optimization objective and constraints. Solve the optimization problem to obtain a practical design solution that comprehensively considers manufacturing features.

[0012] Preferably, the manufacturing features described in S1 include chamfers, welds, flanges, openings, rivets, geometric defects, material defects, boundary defects, and loading defects; the sampling technique is selected from at least one of random sampling, full factor sampling, Latin hypercube sampling, and optimal Latin hypercube sampling, and the structural parameters include skin size, rib size, and rib distribution parameters.

[0013] Preferably, the specific steps of S2 are as follows:

[0014] S21. Calculate the performance indicators of the fine model and the simplified model using the homogenization method, and the homogenization method is one of the rib smoothing method, equivalent stiffness method, representative volume element method, and progressive homogenization method;

[0015] S22. Construct a multi-dimensional data set with structural design variables and manufacturing feature variables as inputs, and equivalent stiffness coefficients, mass, and load-bearing capacity as outputs.

[0016] Preferably, the specific technical process of the homogenization method is as follows:

[0017] a. Apply periodic displacement boundary conditions to the local solid model to obtain the unit characteristic strain and the corresponding stress:

[0018]

[0019] Among them, σ ij represents stress, (P i ) j represents the i-direction reaction force on the jth outer surface, and S j represents the area of the jth outer surface;

[0020] Obtain the overall average stress and average strain of the local solid model according to the following formula;

[0021]

[0022] Wherein, V represents volume, and ε ij represents strain;

[0023] b. Calculate the equivalent elastic modulus constant of the local solid model according to the following formula:

[0024]

[0025] Wherein, E i represents the equivalent elastic modulus; μ ij represents the Poisson's ratio, and G ij represents the shear modulus.

[0026] Preferably, the performance indicators described in S21 include the equivalent elastic modulus, the equivalent stiffness coefficient, and the structural mass.

[0027] Preferably, the specific steps of S3 are as follows:

[0028] S31. Train a mathematical regression model using a data-driven method, with manufacturing feature parameters and structural dimension parameters as inputs, and the influence of the refined model with manufacturing features on the performance indicators of the simplified model as outputs, to establish a non-linear mapping relationship between manufacturing feature input indicators and performance parameter influences;

[0029] S32. During the training process, use the cross-validation method to evaluate the prediction accuracy of the model, and iteratively optimize the internal parameters in the mathematical regression model to establish a mathematical model that can accurately characterize the relationship between manufacturing features and key performance indicators.

[0030] Preferably, the mathematical models constructed in S32 include radial basis function models, Kriging models, polynomial models, response surface models, and Gaussian process regression models.

[0031] Preferably, the specific steps of S4 are as follows:

[0032] S41. Construct an optimization problem, and select one or more of the structural mass, structural bearing capacity, maximum deformation, maximum stress, and stability coefficient as the optimization objective and optimization constraints according to specific design requirements;

[0033] S42. Carry out optimization using an optimization framework based on a surrogate model, and call the mathematical regression model to simulate the influence of structural manufacturing features during the optimization process;

[0034] S43. Solve the optimization problem using optimization algorithms such as genetic algorithms, ant colony algorithms, particle swarm algorithms, or sequential quadratic programming, and optimization strategies such as surrogate model optimization, continuous distribution class optimization, or hybrid optimization to obtain an actual design solution that comprehensively considers manufacturing features.

[0035] Preferably, the optimization problem formulated in S41 is as follows:

[0036] Find: = [x1, x2, ..., x i , y1, ..., y j ;

[0037] Objective: min(obj);

[0038]

[0039] Wherein, X is the optimized structural parameter, including the parameters x of n simplified models i and the parameters y of m manufacturing features j , the optimization objective is to minimize the custom objective obj, the custom objective obj can be selected according to specific requirements, and the optimization constraints are that the parameters of the simplified model and the manufacturing feature parameters are respectively within their variable design spaces and the custom design constraint sub k meet the requirements. In the above formula, the overline represents the upper limit value of the variable, and the underline represents the lower limit value of the variable.

[0040] A thin-walled stiffened structure design system considering manufacturing features, including:

[0041] An experimental design unit for sampling sample points of a fine model of a space cylinder shell structure containing manufacturing features and a corresponding simplified model within the design space;

[0042] A numerical equivalent analysis unit connected to the experimental design unit for calculating the mechanical property indexes of the model and constructing a corresponding performance index data set;

[0043] An optimization problem solving unit connected to the numerical load-bearing analysis unit for constructing an optimization problem and solving the optimization problem to obtain a design scheme of a space stiffened cylinder shell considering manufacturing features;

[0044] A numerical load-bearing analysis unit for solving the load-bearing capacity of a specific cylinder shell structure, and carrying out and solving an optimization problem.

[0045] Therefore, the present invention adopts the above-mentioned thin-walled stiffened structure design method and system considering manufacturing features, and has the following beneficial effects:

[0046] (1) Considering structural manufacturing features such as chamfers, by sampling and constructing a data set, a mathematical model that can reflect the influence of manufacturing features on the mechanical properties of the overall structure is established. By calling this mathematical model during the optimization process, the influence of complex manufacturing features such as chamfers on the performance of the overall structure can be fully considered in the optimization design, achieving a more refined structural optimization design while ensuring the numerical analysis calculation efficiency, combining the analysis accuracy of the refined numerical model and the analysis efficiency of the mathematical regression model, and solving the problem of insufficient accuracy in the refined and lightweight design of complex thin-walled structures in launch vehicles;

[0047] (2) In the design of aerospace cylindrical shells, the influence of manufacturing features such as chamfers on the structural quality and stiffness is fully considered. The influence coefficient of manufacturing features is quantified by training a mathematical model and comprehensively introduced into the design process. The effect of manufacturing features on the bearing capacity of the overall structure is simulated and considered, realizing the design of thin-walled stiffened structures considering manufacturing features;

[0048] (3) It has high applicability and operability, can meet the refined design requirements of complex aerospace thin-walled cylindrical shell structures, and improve the lightweight and bearing efficiency of the design scheme of stiffened thin-walled structures. Finally, it provides technical guidance and special tools for realizing the high-performance and high-reliability design of aerospace equipment.

[0049] The technical solution of the present invention will be further described in detail below through the accompanying drawings and embodiments. Description of the Drawings

[0050] Figure 1 is a flowchart of a design method for a thin-walled stiffened structure considering manufacturing features according to the present invention;

[0051] Figure 2 is a schematic diagram of a fine model and an equivalent model of a design method for a thin-walled stiffened structure considering manufacturing features according to the present invention;

[0052] Figure 3 is a schematic diagram of the training of a mathematical model of a design method for a thin-walled stiffened structure considering manufacturing features according to the present invention;

[0053] Figure 4 is a schematic diagram of the optimization solution of a design method for a thin-walled stiffened structure considering manufacturing features according to the present invention;

[0054] Figure 5 is a comparison chart of the weights and bearing capacities of the design results obtained by a design method for a thin-walled stiffened structure considering manufacturing features according to the present invention and the traditional method. Detailed Embodiments

[0055] The following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the present invention claimed, but merely represents selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0056] As Figure 1 shown, the present invention provides a design method for a thin-walled stiffened structure considering manufacturing features, including the following steps:

[0057] S1. Establish a structural model of the aerospace cylinder structure with fine manufacturing features and a simplified model without features (modeling can be done using existing commercial finite element software such as ABAQUS, ANSYS, etc., which will not be elaborated here). Among them, for manufacturing features with periodic characteristics (such as chamfers, welds, etc.), the representative volume element method and the progressive homogenization method can be used for unit cell modeling. Taking the chamfer as an example for unit cell modeling, its schematic diagram is as shown in Figure 2 and the fine unit cell model Ω detailed and the simplified unit cell model Ω reduced are obtained. The fine model includes the skin, stiffeners, and manufacturing features (chamfers). The height of the ribs in the fine model is h stiffener , the thickness of the stiffeners is t stiffener , the thickness of the skin is t skin , and the chamfer size is r filler .

[0058] The locally refined solid model includes various models such as the minimum representative unit cell and multiple unit cells. These are all existing models, and only one of them can be selected for specific use.

[0059] Through sampling technology for experimental design, sampling of structural parameters is carried out at sampling points within the design space. Manufacturing features include chamfers, welds, flanges, openings, rivets, geometric defects, material defects, boundary defects, and loading defects; the sampling technology is selected from at least one of random sampling, full factor sampling, Latin hypercube sampling, and optimal Latin hypercube sampling, and the structural parameters include skin size, rib size, and rib distribution parameters.

[0060] S2. Perform numerical analysis on each sample point to obtain its performance indicators, and construct a data set based on the performance indicator data of the sample points.

[0061] S21. Calculate the performance indicators of the fine model and the simplified model using the homogenization method. The homogenization method is one of the rib smoothing method, equivalent stiffness method, representative volume element method, and progressive homogenization method. The performance indicators include equivalent elastic modulus, equivalent stiffness coefficient, and structural mass.

[0062] The specific technical process of the homogenization method is as follows:

[0063] a. Apply periodic displacement boundary conditions to the local solid model to obtain the unit characteristic strain and the corresponding stress:

[0064]

[0065] Among them, σ ij represents stress, (P i ) j represents the i-direction reaction force on the j-th outer surface, and S j represents the area of the j-th outer surface;

[0066] The overall average stress and average strain of the local solid model are obtained according to the following formula:

[0067]

[0068] where V represents volume, and ε ij represents strain;

[0069] b. Calculate the equivalent elastic modulus constant of the local solid model according to the following formula;

[0070]

[0071] where E i represents the equivalent elastic modulus; μ ij represents the Poisson's ratio, and G ij represents the shear modulus.

[0072] The representative volume element method is used to calculate the equivalent stiffness coefficient of the fine model which are the equivalent elastic modulus constants in the X, Y, and Z axis directions respectively; The simplified model includes the skin and stiffeners, and the thickness of the stiffeners in the simplified model is h' stiffner , the thickness of the stiffeners is t' stiffner , and the thickness of the skin is t' skin , and the representative volume element method is used to calculate the equivalent stiffness coefficient of the simplified model which are the equivalent elastic modulus constants in the X, Y, and Z axis directions respectively.

[0073] For manufacturing features without periodicity such as flanges, openings, geometric defects, material defects, boundary defects, and loading defects, numerical analysis methods such as finite element can also be directly used to obtain their impacts on the bearing capacity (such as mass, bearing capacity, etc.).

[0074] S22. Construct a multi-dimensional data set with structural design variables and manufacturing feature variables as inputs and equivalent stiffness coefficients, mass, and bearing capacity as outputs.

[0075] S3. Train a mathematical model to characterize the relationship between manufacturing features and key performance indicators.

[0076] Using the same software as in S1, a shell model is constructed according to the thickness of the stiffeners and the skin of the modified simplified model. The difference between the modeling in S3 and that in S1 is that the modeling in S1 uses solid modeling, while the modeling in S3 uses shell modeling. The shell model is a finite element model, and the calculation cost of the shell model is low, so an efficient and high-fidelity calculation model is obtained, and the calculation efficiency advantage of the shell model is fully utilized.

[0077] S31. Based on the dataset obtained in S2, use the data-driven method to train a mathematical regression model. Take the manufacturing feature parameters and structural dimension parameters as inputs, and take the influence amount of the fine model with manufacturing features on the performance index of the simplified model as the output, and establish a non-linear mapping relationship between the manufacturing feature input index and the performance parameter influence output. The process of model training is as Figure 3 shown.

[0078] S32. During the training process, use the cross-validation method to evaluate the model prediction accuracy, and iteratively optimize the internal parameters in the mathematical regression model to ensure that the established mathematical regression model can accurately reflect the influence law of manufacturing features on performance, and establish a mathematical model that can accurately characterize the relationship between manufacturing features and key performance indicators. The mathematical model includes radial basis function model, Kriging model, polynomial model, response surface model and Gaussian process regression model.

[0079] S4. Construct an optimization problem, and call the mathematical model to predict the influence of manufacturing features on the optimization objective and constraints, and solve the optimization problem to obtain the actual design scheme that comprehensively considers manufacturing features.

[0080] S41. Construct an optimization problem, and select one or several of the structural mass, structural bearing capacity, maximum deformation, maximum stress and stability coefficient as the optimization objective and optimization constraints according to specific design requirements.

[0081] The formulated optimization problem is as follows:

[0082] Find: = [x1, x2,..., x i , y1,..., y j ;

[0083] Objective: min(obj);

[0084]

[0085] Among them, X is the optimization structure parameter, including the parameters x i of n simplified models and the manufacturing feature parameters y j . The optimization objective is to minimize the custom objective obj, and the custom objective obj can be selected according to specific requirements. The optimization constraints are that the parameters of the simplified model and the manufacturing feature parameters are respectively within their variable design spaces and the custom design constraint sub k meets the requirements. The overline in the above formula represents the upper limit value of the variable, and the underline represents the lower limit value of the variable.

[0086] S42. Use the optimization framework based on the surrogate model to carry out optimization to improve the optimization efficiency, and call the mathematical regression model to simulate the influence of structural manufacturing features during the optimization process.

[0087] S43. Solve the optimization problem using an optimization algorithm such as genetic algorithm, ant colony algorithm, particle swarm algorithm or sequential quadratic programming, and an optimization strategy such as surrogate model optimization, continuous distribution optimization, or hybrid optimization to obtain an actual design solution that comprehensively considers manufacturing features.

[0088] Embodiment

[0089] Carry out an optimization design considering manufacturing features for a typical stiffened cylindrical shell structure. The diameter of the cylindrical shell is 1600 mm, the height is 1000 mm, and it is reinforced with a dense and uniform grid on the inner side. Assume the manufacturing process is integral forging combined with CNC milling. The tip of the milling cutter (such as a ball nose milling cutter or a round nose milling cutter) usually has a fillet radius and cannot form an absolute right angle. The cutter will naturally leave a fillet on the edge that matches the tip radius during the cutting path. Therefore, the typical manufacturing features considered in the optimization are the chamfers between ribs and ribs, and between ribs and skins.

[0090] Define the optimization variables as the number of circumferential ribs N C , the number of axial ribs N A , the skin thickness t s , the rib thickness t r , the rib height h r , and a manufacturing feature parameter chamfer radius R fillet . However, considering that the size of the chamfer manufacturing feature is only related to the machining tool and the feed path, and generally will not be adjusted with the change of the product, so here the manufacturing feature parameter chamfer radius R fillet is fixed to a typical value, that is, 4.0 mm. The upper and lower limits of the design variables are determined as shown in the following table:

[0091] Table 1

[0092] <![CDATA[t s > <![CDATA[h r > <![CDATA[t r > <![CDATA[N C > <![CDATA[N A > Upper limit of design variable [mm] 3.0 13.0 4.0 20 150 Lower limit of design variable [mm] 1.0 9.0 2.0 4 80

[0093] The structural material is aluminum alloy, its elastic modulus constant E = 76169 MPa, and Poisson's ratio ν = 0.3. A bilinear elastoplastic model is adopted, with a yield limit of 339.58 MPa, a strength limit of 437.92 MPa, and an elongation of 12.92%. The above material property parameters are all obtained through material tests.

[0094] Use the optimal Latin hypercube sampling method for sampling, and establish a fine model including chamfer features and a simplified model without chamfer features respectively. Since the chamfer has strong periodicity, the representative volume element method is used for unit cell modeling. The input variables involved include: the number of circumferential ribs N C , the number of axial ribs N A , the skin thickness t s , the rib thickness t r , the rib height h rand a manufacturing feature parameter chamfer radius R fillet . The output variables are those of the unit cell which are the equivalent elastic modulus constants in the X, Y, and Z axis directions respectively.

[0095] The representative volume element method is used to calculate the equivalent stiffness coefficients and the equivalent elastic modulus constants

[0096] in the X, Y, and Z axis directions of the fine model and the simplified model. The number of samples in the data set is 6,000 groups, and a mathematical regression model is trained based on the data set. This model is selected as the Gaussian process regression model. After training, the accuracy of the mathematical model is verified, and the prediction error RMSE is all lower than 1%, and the accuracy meets the usage requirements.

[0097] The optimization objective is selected as the structural design load-bearing capacity P design to be maximized, the constraint is that the mass is less than weight0 = 44.5 kg, and all optimization variables are between the upper and lower limits. The following optimization formulation is constructed:

[0098] Find: X = t s , h, t r , N a , N c ;

[0099] Maximize: P design ;

[0100]

[0101] The optimization is carried out using the optimization framework based on the surrogate model, and the mathematical regression model is called during the optimization process to simulate the influence of the structural manufacturing features. The optimization framework is as Figure 4 shown. For comparative verification, in addition to carrying out the optimization design using the method of the present invention, a traditional optimization design without considering the chamfer feature is also carried out.

[0102] The optimization algorithm is selected as the multi-island genetic algorithm, the population size is 50, the number of islands is 2, and the number of generations of evolution is 20. After the optimization, both optimizations reach convergence, and the optimization results are as Figure 5 shown.

[0103] It can be found that the load-bearing capacity of the optimization result obtained by the traditional optimization method is 4396.72 kN, and the load-bearing capacity of the model of the method of the present invention is 5040.85 kN, with an increase of up to 14.650%. And the structural mass of the optimization result of the traditional optimization method is only 41.723 kg, far from reaching the structural mass constraint. From the above analysis, it can be seen that the optimization result obtained by optimizing the design using the method of the present invention has a higher load-bearing capacity compared with the traditional optimization method without considering manufacturing features.

[0104] A thin-walled stiffened structure design system considering manufacturing features, comprising:

[0105] An experimental design unit for sampling sample points for a fine model of a space cylinder shell structure containing manufacturing features and a corresponding simplified model within a design space;

[0106] A numerical equivalent analysis unit connected to the experimental design unit for calculating the mechanical property indexes of the model and constructing a corresponding performance index data set;

[0107] An optimization problem solving unit connected to the numerical load-bearing analysis unit for constructing an optimization problem and solving the optimization problem to obtain a design scheme for a space stiffened cylinder shell considering manufacturing features;

[0108] A numerical load-bearing analysis unit for solving the load-bearing capacity of a specific cylinder shell structure, and carrying out and solving an optimization problem.

[0109] Therefore, the present invention adopts the above-mentioned thin-walled stiffened structure design method and system considering manufacturing features, fully considers the influence of various manufacturing features on the load-bearing capacity of the structure, improves the model analysis accuracy on the premise of ensuring the structural analysis efficiency, has no restrictions on elements such as working condition loads, optimization objectives and constraints, and is applicable to complex mechanical analysis requirements such as buckling processes and defect sensitivities; by quantitatively characterizing the influence of the manufacturing feature structure in the space cylinder shell on stiffness and mass, simulating and considering the effect of manufacturing features on the bearing capacity of the overall structure, the structural design of manufacturing features is realized.

[0110] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that: they can still modify or equivalently replace the technical solutions of the present invention, and these modifications or equivalent replacements cannot make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A design method for thin-walled stiffened structures considering manufacturing features, characterized in that, It includes the following steps: S1. Establish a structural model with fine manufacturing features and a simplified model without features for the space cylinder structure, conduct experimental design through sampling techniques, and perform point sampling on structural parameters within the design space; S2. Conduct numerical analysis on each sample point to obtain its performance indicators, and construct a data set based on the performance indicator data of the sample points; S3. Train a mathematical model to characterize the relationship between manufacturing features and key performance indicators; S4. Construct an optimization problem, and call the mathematical model to predict the impact of manufacturing features on the optimization objective and constraints, and solve the optimization problem to obtain an actual design solution that comprehensively considers manufacturing features.

2. The design method of a thin-walled stiffened structure considering manufacturing features according to claim 1, characterized in that: The manufacturing features described in S1 include chamfers, welds, flanges, openings, rivets, geometric defects, material defects, boundary defects, and loading defects; the sampling techniques are selected from random sampling, full factorial sampling, Latin hypercube sampling, and optimal Latin hypercube sampling, and the structural parameters include skin dimensions, rib dimensions, and rib distribution parameters.

3. A design method for a thin-walled stiffened structure considering manufacturing features according to claim 1, characterized in that, The specific steps of S2 are as follows: S21. Calculate the performance indicators of the fine model and the simplified model using the homogenization method, and the homogenization methods are rib smoothing method, equivalent stiffness method, representative volume element method, and progressive homogenization method; S22. Construct a multi-dimensional data set with structural design variables and manufacturing feature variables as inputs and equivalent stiffness coefficients, mass, and load-bearing capacity as outputs.

4. The design method of a thin-walled stiffened structure considering manufacturing features according to claim 3, characterized in that The specific technical process of the homogenization method is as follows: a. Apply periodic displacement boundary conditions to the local solid model to obtain the unit characteristic strain and the corresponding stress accordingly; Among them, σ ij represents stress, (P i ) j represents the reaction force in the i direction on the j-th outer surface, and S j represents the area of the j-th outer surface; Obtain the overall average stress and average strain of the local solid model according to the following formula; where V represents volume and ε ij represents strain; b. Calculate the equivalent elastic modulus constant of the local solid model according to the following formula; Among them, E i represents the equivalent elastic modulus; μ ij represents the Poisson's ratio, and G ij represents the shear modulus.

5. The design method of a thin-walled stiffened structure considering manufacturing features according to claim 3, characterized in that: The performance indicators described in S21 include equivalent elastic modulus, equivalent stiffness coefficient, and structural mass.

6. A design method for a thin-walled stiffened structure considering manufacturing features according to claim 1, characterized in that, The specific steps of S3 are as follows: S31. Train a mathematical regression model using a data-driven method, take manufacturing feature parameters and structural size parameters as inputs, and take the influence amount of the fine model with manufacturing features on the performance indicators of the simplified model as the output, and establish a non-linear mapping relationship between the manufacturing feature input indicators and the performance parameter influence outputs; S32. During the training process, use the cross-validation method to evaluate the model prediction accuracy, and iteratively optimize the internal parameters in the mathematical regression model to establish a mathematical model that can accurately characterize the relationship between manufacturing features and key performance indicators.

7. A design method for a thin-walled stiffened structure considering manufacturing features according to claim 5, characterized in that: The mathematical models constructed in S32 include radial basis function model, Kriging model, polynomial model, response surface model, and Gaussian process regression model.

8. A design method for a thin-walled stiffened structure considering manufacturing features according to claim 1, characterized in that The specific steps of S4 are as follows: S41. Construct an optimization problem, and select structural mass, structural bearing capacity, maximum deformation, maximum stress, and stability coefficient as the optimization objective and optimization constraints according to specific design requirements; S42. Carry out optimization using an optimization framework based on a surrogate model, and call the mathematical regression model to simulate the influence of structural manufacturing features during the optimization process. S43. Use optimization algorithms such as genetic algorithm, ant colony algorithm, particle swarm algorithm or sequential quadratic programming, and optimization strategies such as surrogate model optimization, continuous distribution optimization, or hybrid optimization to solve the optimization problem and obtain the actual design solution that comprehensively considers manufacturing features.

9. The design method of a thin-walled stiffened structure considering manufacturing features according to claim 8, characterized in that, The optimization problem formulated in S41 is as follows: Find: = [x1, x2,..., x i , y1,..., y j ; Objective: min(obj); Among them, X is the optimized structural parameter, including the parameters x of n simplified models i and the parameters y of m manufacturing features j . The optimization objective is to minimize the custom objective obj, and the custom objective obj can be selected according to specific requirements. The optimization constraints are that the parameters of the simplified model and the manufacturing feature parameters are respectively within their variable design spaces and the custom design constraint sub k meet the requirements. In the above formula, the overline represents the upper limit value of the variable, and the underline represents the lower limit value of the variable 10. A thin-walled stiffened structure design system considering manufacturing features, which is applied to a thin-walled stiffened structure design method according to any one of claims 1-9, and is characterized in that, It includes: An experimental design unit for sampling sample points of the fine model and the corresponding simplified model of the aerospace cylindrical shell structure containing manufacturing features within the design space; A numerical equivalent analysis unit connected to the experimental design unit for calculating the mechanical property indexes of the model and constructing the corresponding performance index data set; An optimization problem solving unit connected to the numerical load-bearing analysis unit for constructing the optimization problem and solving the optimization problem to obtain the design solution of the aerospace stiffened cylindrical shell considering manufacturing features; A numerical load-bearing analysis unit for solving the load-bearing capacity of a specific cylindrical shell structure, and conducting and solving the optimization problem.

Images