A design method of a shell curve stiffened structure and a shell curve stiffened structure

Abstract

The present application relates to the technical field of plate shell structure, and particularly relates to a plate shell curve stiffened structure design method and a plate shell curve stiffened structure. The method is based on given design parameters to select an implicit curve stiffened structure surface and a cross section, then through layout optimization of the curve stiffened structure and optimization of the cross section parameters of the curve stiffened structure by establishing a proxy model, finally the optimal curve stiffened structure design is obtained. In the method, the implicit surface can be given by any general surface form, which has great freedom. The application of the proxy model has higher calculation efficiency and can explicitly handle constraints, reducing the calculation amount. The design method combines the E-AMS method, which can achieve high-precision restoration of the implicit curve while being explicit, and realize accurate modeling of the curve stiffened structure. The plate shell curve stiffened structure provided by the present application is obtained by using the above plate shell curve stiffened structure design method, which has lower design cost, higher efficiency, and better buckling resistance and bearing capacity.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of plate shell structure, and particularly relates to a plate shell curve stiffening structure design method and a plate shell curve stiffening structure. BACKGROUND

[0002] The plate shell structure is one of main load-bearing structural members widely used in engineering structures, and is widely applied in aerospace, energy and wind power, and communication and navigation. Typical failure forms of the plate shell structure are stability failure (buckling) and strength failure, and the stability failure occupies a dominant position. The stiffening design on the plate shell as a method capable of significantly enhancing the buckling resistance of the plate shell is a main content of the plate shell structure design. In the design of the stiffening layout, the new curve layout has a wider design space, better buckling resistance and bearing capacity, and better bearing characteristics than the traditional orthogonal, equilateral triangle and other uniform stiffening structures. Therefore, the curve stiffening design method of the plate shell structure has become an important research direction. However, compared with the traditional stiffening structure, the path representation function of the curve stiffening structure is more complex, which leads to an explosive growth of design variables, and further seriously restricts the layout optimization design of the curve stiffening structure.

[0003] In order to solve the above problems, some solutions have been proposed in the prior art, for example:

[0004] A Chinese patent application with the application number 202010649313.5 and the name of a curve stiffening structure layout intelligent design method based on image feature learning discloses a scheme, which is essentially a method for realizing the curve stiffening layout design based on the b-spline curve combined with the image neural network. The main steps are as follows: 1. Selecting the curve stiffening path function to generate an image set, inputting the auto-encoding network for unsupervised learning and training, and completing the extraction of the curve stiffening image structure features. 2. Establishing an analysis model of the mechanical response of the curve stiffening structure to form a data set for supervised learning and training, and further inputting the convolutional neural network model built by the decoding network model and the full connection layer to complete the learning of the mechanical response of the curve stiffening structure. 3. Based on the convolutional neural network model for predicting the mechanical response of the curve stiffening structure, the evolutionary algorithm is used to complete the optimization design of the curve stiffening structure layout. The most core part is to introduce the neural network technology to reduce the calculation amount in the structure design. However, since the definition of the curve uses the definition form of the spline curve, a single spline curve has more than 8 design parameters, and the parameter amount will increase sharply in the multi-stiffening optimization, which leads to the difficulty in obtaining a good bearing capacity configuration in the multi-stiffening optimization. In addition, the image feature learning method is strongly dependent on the specific form of the problem, which limits the application of the method to other geometric configurations of the plate shell structure (poor universality).

[0005] The Chinese patent application with the application number 202311358627.X and the name "A composite material net-like curve stiffened structure and an optimization design method thereof" provides a solution based on uniform curve stiffened layout by improving the mathematical expression of the stiffener curve to realize the curve stiffened structure. The main steps of this method are: 1. Defining the optimization mathematical model of the curve stiffened structure. 2. Determining the geometric dimensions, composite material design parameters, load and displacement boundary conditions of the net-like curve stiffened structure. 3. Based on the above parameters, an optimization model is constructed, and the obtained optimization design is solved. Similarly, the Chinese patent application with the application number 202311106128.1 and the name "A spiral composite material curve stiffened wallboard multi-working condition stability optimization method" also directly gives the layout form of the curve stiffened structure by using the formula method, and realizes the modeling of the curve stiffened structure by this method. For the application numbers 202311358627.X and 202311106128.1, both methods directly limit the form of the curve, and the configuration freedom is extremely limited. SUMMARY

[0006] The first object of the present application is to provide a plate shell curve stiffened structure design method to solve the problems of low design freedom of plate shell curve stiffened structure and high calculation load due to the complexity of buckling mechanical response.

[0007] The second object of the present application is to provide a plate shell curve stiffened structure based on the plate shell curve stiffened structure design method.

[0008] In order to achieve the above-mentioned first object, in the first aspect, the present application provides a plate shell curve stiffened structure design method, comprising the following steps:

[0009] S100: performing design problem definition, including:

[0010] S101: determining the fixing mode of the four sides of the plate shell and the stress mode;

[0011] S102: determining the maximum mass allowed by the plate shell in the given structure design;

[0012] S103: selecting the implicit surface of the curve stiffened structure and the cross section of the curve stiffened structure;

[0013] S200: curve stiffened structure layout optimization design, including:

[0014] S201: selecting the relative stiffness parameter, fixing the geometric parameters of the selected cross section of the curve stiffened structure in S103, taking the parameters of the implicit surface in S103 as the design parameters, and generating the initial sample set required by the evolutionary algorithm;

[0015] S202: Geometric modeling of all sample points in the initial sample set is performed according to the implicit surface through the E-AMS method, and mechanical response analysis of the initial sample set is performed;

[0016] S203: The bearing efficiency of the initial sample set is evaluated to determine whether the convergence requirement is met, and if the convergence requirement is met, the layout of the curve-stiffened structure is output;

[0017] If the convergence requirement is not met, the evolutionary algorithm iteratively generates a new sample set, geometric modeling of all sample points in the new sample set is performed according to the implicit surface through the E-AMS method, mechanical response analysis of the new sample set is performed, and the bearing efficiency is evaluated to determine whether the convergence requirement is met, and the process is repeated until the convergence requirement is met;

[0018] S300: Cross-section parameter optimization design of the curve-stiffened structure, including:

[0019] S301: Sampling the sample space of the curve-stiffened structure layout output in step S203 to generate a parameter set of samples;

[0020] S302: Modeling the parameter set of samples generated in step S301 using the E-AMS method, and performing finite element buckling calculation using the finite element method to obtain the corresponding buckling load;

[0021] S303: Selecting and training a proxy model;

[0022] S304: Based on the proxy model trained in step S303, calculating the maximum eigenvalue under the equal mass constraint using an optimization algorithm;

[0023] S305: Outputting the final cross-section parameters of the curve-stiffened structure after the result converges.

[0024] Optionally, the training of the proxy model is based on the following buckling modal characteristic equation:

[0025]

[0026] Where λ and λ0 are the buckling eigenvalues of the stiffened and unstiffened plate, respectively, β i is a dimensionless parameter of the appearance of the rib;

[0027] is a dimensionless material constant, where E r , E p are the elastic moduli of the rib and the plate, respectively, and v is the Poisson's ratio of the rib;

[0028] is the area integral of the global buckling modal displacement field on the rib layout;

[0029] is the dimensionless relative stiffness, S r , S p are the area of the stiffened layout and the panel respectively, I z is the cross-sectional moment of inertia of the corresponding curved stiffened structure about the neutral axis of the panel, l1, l2 are the shape parameters of the cross-section, t p is the thickness of the panel.

[0030] Optionally, the surrogate model can be any general surrogate model that takes β{β i} as inputs, as outputs, where the number of β{β i} parameters is equal to the number of cross-sectional parameters of the curved stiffened structure minus 2.

[0031] Optionally, the direct search method is used in step S304 to maximize the eigenvalue under the equal mass constraint based on the following objective function:

[0032] Objective function: where the number of β{β i} parameters is equal to the number of cross-sectional parameters of the curved stiffened structure minus 2.

[0033] Constraint function: t p S p + A r L r = constant, where A r is the cross-sectional area of the curved stiffened structure, and L r is the total length of the stiffeners.

[0034] Optionally, the E-AMS method flow is as follows:

[0035] An initial background mesh is generated for the marching mesh method, and then the first implicit curve is modeled. An epsilon mesh is constructed in the square that captures the implicit curve, and the grid is gradually refined according to the curvature distribution information provided by the epsilon mesh point set distribution to obtain an accurate implicit curve spline.

[0036] Optionally, the parameter expression of the implicit surface in step S103 is {S i (u j )}, where S i represents the i-th implicit surface, u j represents the j-th parameter, and the geometric parameters of the cross-section of the curved stiffened structure are expressed as {t p , l i}, where t p is the thickness of the panel, and the remaining parameters are the cross-sectional shape parameters of the curved stiffened structure.

[0037] The cross-section geometry parameters {t p ,l i} in the fixing step S103 are fixed according to the following formula:

[0038]

[0039] Where S r , S p are the areas of the stiffened structure and the plate shell respectively, I z is the cross-section moment of inertia of the corresponding curved stiffened structure around the neutral axis of the plate, l1, l2 are shape parameters of the cross-section, t p is the thickness of the plate, and C0 is a selected constant.

[0040] Optionally, in step S203, the bearing efficiency of the initial sample set and the new sample set is evaluated according to the following formula:

[0041]

[0042] Where λ, λ0 are the buckling eigenvalues of the plate shell after stiffening and before stiffening respectively.

[0043] Optionally, the evolution algorithm in step S201 is any one of a genetic algorithm, a simulated annealing algorithm, a particle swarm algorithm, an artificial neural network algorithm and an ant colony algorithm.

[0044] Optionally, the mechanical response analysis in step S202 adopts a finite element method, a finite difference method or a boundary element method; and / or

[0045] The optimization algorithm in step S304 adopts a direct search method or a gradient descent method.

[0046] In a second aspect, the present application further provides a plate shell curved stiffened structure obtained by the plate shell curved stiffened structure design method of any implementation manner of the first aspect.

[0047] The above technical solution of the present application has the following advantages:

[0048] The plate shell curve stiffening structure design method provided by the application, by determining the fixing mode and stress mode of the plate shell and the maximum mass allowed in the given structure design, selecting the implicit curved surface and cross section of the curve stiffening structure, then selecting the relative stiffness parameters and the fixed geometric parameters of the selected cross section, taking the parameters of the selected implicit curved surface as the design parameters, generating the initial sample set required by the evolutionary algorithm, performing geometric modeling on all sample points in the initial sample set according to the implicit curved surface through the E-AMS method, and performing mechanical response analysis on the initial sample set and evaluating the bearing efficiency, if the convergence requirement has been reached, the curve stiffening structure layout is output, if the convergence requirement has not been reached, the evolutionary algorithm iteratively generates a new sample set, and the geometric modeling is performed on all sample points in the new sample set according to the implicit curved surface through the E-AMS method, and the mechanical response analysis and bearing efficiency evaluation are performed, until the convergence requirement is reached. The sample space of the output curve stiffening structure layout is sampled to generate a parameter set of the sample, the parameter set of the sample is modeled by using the E-AMS method, and the finite element buckling calculation is performed by using the finite element method, the corresponding buckling load is obtained, the proxy model is selected and trained, and the maximum eigenvalue is calculated under the equal mass constraint based on the trained proxy model, and the final curve stiffening structure cross section parameters are output after the result converges. In the method, the implicit curved surface can be given by any general curved surface form, and has great freedom. Moreover, the proxy model has high calculation efficiency and can explicitly process the constraint, and the calculation amount can be reduced. In addition, the design method combines the E-AMS method, can realize high-precision restoration of the implicit curve while being explicit, and realizes accurate modeling of the curve stiffening structure.

[0049] The plate shell curve stiffening structure design method provided by the application trains a proxy model based on a buckling modal characteristic equation, solves the problem that the buckling mode of the structure cannot be predicted under a certain stiffening layout. Moreover, as part of the result analysis in the layout optimization, the parameters are decoupled, and the calculation cost is reduced.

[0050] The plate shell curve stiffening structure provided by the application is obtained by using the above plate shell curve stiffening structure design method, has lower design cost, higher efficiency, and better buckling bearing capacity and bearing characteristics. BRIEF DESCRIPTION OF DRAWINGS

[0051] The drawings of the application are provided for illustrative purposes only, and the proportions and quantities of the components in the drawings may not be consistent with the actual product.

[0052] Figure 1 is a flowchart of the plate shell curve stiffening structure design method in the embodiment of the application;

[0053] Figure 2is a flow chart of the E-AMS modeling method in the embodiment of the present application;

[0054] Figure 3 is a flow chart of the layout design-section parameter design of the plate shell curve stiffened structure in the embodiment of the present application;

[0055] Figure 4 is a process schematic diagram of the design of the sparse curve stiffened structure under the fixed support boundary condition in the embodiment of the present application;

[0056] Figure 5 is a process schematic diagram of the design of the dense curve stiffened structure under the fixed support boundary condition in the embodiment of the present application;

[0057] Figure 6 is a sectional schematic diagram of the curve stiffened structure in the embodiment of the present application;

[0058] Figure 7 is a sectional schematic diagram of another curve stiffened structure in the embodiment of the present application;

[0059] Figure 8 is a sectional schematic diagram of still another curve stiffened structure in the embodiment of the present application. DETAILED DESCRIPTION

[0060] In order to make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0061] As shown in Figure 1 , the plate shell curve stiffened structure design method provided by the embodiments of the present application comprises the following steps:

[0062] S100: defining the design problem.

[0063] S200: layout optimization design of the curve stiffened structure.

[0064] S300: section parameter optimization design of the curve stiffened structure.

[0065] In step S100, the design problem is defined, which comprises:

[0066] S101: determining the fixing mode of the four sides of the plate shell and the stress mode.

[0067] S102: determining the maximum mass allowed by the plate shell in the given structure design.

[0068] S103: The type of the cross section of the implicit surface of the curve stiffened structure and the curve stiffened structure is selected according to experience. It should be noted that the surface form can be any two-dimensional surface, such as an isoparametric interpolation surface, a Nurbs (non-uniform rational b-spline), a polynomial surface, etc. The cross section can be any cross section, such as a rectangular cross section shown in the figure, Figure 6 a T-shaped cross section shown in the figure, Figure 7 an I-shaped cross section shown in the figure, etc. In the embodiment, the implicit surface level set method is preferably used in defining the curve stiffened structure. Figure 8 After the design problem is defined by step S100, the curve stiffened structure layout optimization design and the curve stiffened structure cross section parameter optimization design are performed on this basis, and then the final curve stiffened structure design is obtained.

[0069] In the embodiment, step S200: curve stiffened structure layout optimization design, includes:

[0070] S201: The relative stiffness parameter is selected, and the geometric parameters of the selected cross section of the curve stiffened structure in step S103 are fixed according to the selected relative stiffness parameter. The parameters of the implicit surface in step S103 are taken as design variables to generate an initial sample set required by an evolutionary algorithm. It should be noted that the term "relative stiffness" refers to the ratio of the stiffness of the rib to the stiffness of the plate shell. In the embodiment, the evolutionary algorithm can use any one of a genetic algorithm, a simulated annealing algorithm, a particle swarm algorithm, an artificial neural network algorithm, an ant colony algorithm, etc.

[0071] S202: All sample points in the initial sample set are geometrically modeled by an E-AMS (Epsilon-net enhanced Adaptive Marching Squares) method according to the implicit surface thereof, and then the initial sample set is subjected to mechanical response analysis. In the embodiment, the mechanical response analysis of the initial sample set can be performed by using a finite element method, a finite difference method, or a boundary element method, etc.

[0072] S203: The carrying efficiency of the initial sample set is evaluated to determine whether the convergence requirement is met. If the convergence requirement is met, the curve stiffened structure layout is output. If the convergence requirement is not met, the evolutionary algorithm iteratively generates a new sample set, all sample points in the new sample set are geometrically modeled by an E-AMS method according to the implicit surface thereof, and the new sample set is subjected to mechanical response analysis and carrying efficiency evaluation to determine whether the convergence requirement is met. The above process is repeated until the convergence requirement is met.

[0073] ​

[0074] S300: Optimization design of the cross-section parameters of the curved stiffened structure, comprising:

[0075] S301: Sampling the sample space of the layout of the curved stiffened structure output by step S203 to generate a parameter set of the sample.

[0076] S302: Modeling the parameter set of the sample generated in step S301 by using the E-AMS method, and performing finite element buckling calculation thereon by using the finite element method to obtain the corresponding buckling load.

[0077] S303: Selecting and training the surrogate model. It should be noted that, by using the design method of the present application, the surrogate model can be any one of the general surrogate models such as Gaussian regression model, artificial neural network, support vector machine, polynomial surface corresponding model, etc.

[0078] S304: Based on the surrogate model trained in step S303, calculating the maximum eigenvalue under the equal mass constraint by using the optimization algorithm. In the present embodiment, the optimization algorithm in step S304 can be selected from the direct search method or the gradient descent method.

[0079] S305: Outputting the final cross-section parameters of the curved stiffened structure after the result converges.

[0080] In an example, the training of the surrogate model is based on the following buckling modal eigenvalue equation:

[0081]

[0082] wherein λ and λ0 are the buckling eigenvalues of the stiffened plate and the unstiffened plate respectively, β i is the dimensionless parameter of the appearance of the stiffener;

[0083] is the dimensionless material constant, wherein E r , E p are the elastic modulus of the stiffener and the plate respectively, and v is the Poisson's ratio of the stiffener;

[0084] is the area integral of the global buckling modal displacement field on the stiffened layout;

[0085] is the dimensionless relative stiffness, S r , S p are the areas of the stiffened layout and the plate respectively, and I z is the cross-section moment of inertia of the cross-section of the curved stiffened structure around the neutral axis of the plate, and l1 and l2 are the shape parameters of the cross-section, and the ellipsis means that there can be more cross-section shape parameters such as l3 and l4 due to different cross-section shapes (see the illustration in Figures 6 to 8 .p h is the thickness of the plate r h is the height of the rib.

[0086] In this example, the training of the surrogate model based on the buckling modal characteristic equation described above solves the problem of being unable to predict the buckling modal of the structure under a certain rib layout. And as part of the result analysis in the layout optimization, the parameters are decoupled, reducing the computational cost.

[0087] Referring to Figure 3 The basic principle of the buckling modal characteristic equation is shown, which is a dimensionless equation. As can be seen from the figure, the modal characteristic equation is a two-variable function, which is composed of two ridge lines. These two ridge lines divide the entire region into three parts, corresponding to three buckling forms that may occur under the same rib layout: local rib buckling, local shell buckling, and global buckling. The intersection of the two ridge lines is the strongest point of the structure's buckling resistance.

[0088] Based on the use of the above function for surrogate model training, the maximum eigenvalue is calculated under the constraint of equal mass using β{β i} as input, as output, where the number of β{β i} parameters is equal to the number of cross-sectional parameters of the curved rib structure minus 2.

[0089] Based on the use of the above function for surrogate model training, the direct search method is selected in step S304 based on the following calculation of the maximum eigenvalue under the constraint of equal mass.

[0090] Where the objective function is: Where the number of β parameters is equal to the number of rib cross-sectional parameters minus 2. The constraint function is: p S p +A r L r = constant, A r is the cross-sectional area of the curved rib structure, and L r is the total length of the rib.

[0091] Referring to Figure 2 In an example, the E-AMS method flow is as follows:

[0092] The initial background grid needed for marching grid method is generated, and then the first implicit curve modeling is performed, at which time the captured curve is often not accurate enough. At this time, an epsilon grid is constructed in the square grid in which the implicit curve is captured, and the distribution of the discrete point set of the epsilon grid at this time represents the curvature distribution on the surface. Then the grid is iterated constantly, and the grid is gradually encrypted at the position where the point set in the epsilon grid is dense. Through this method, an accurate implicit curve spline can be obtained. The E-AMS method process serves as the modeling part of the modeling-computing-analysis process in the optimization design structure, and solves the problem of the inability to accurately realize the explicitization of the geometric spline of the implicit curve stiffening modeling.

[0093] Compared with the Marching squares (marching square method), this general implicit curve modeling method cannot ensure the accuracy of the modeled implicit curve, and is prone to curve distortion problems. At the same time, the epsilon grid technology is one of the common technologies for judging the geometric information of a surface in computational geometry, which can give the geometric information on the surface through a discrete point set, and can further reflect the geometric information of the level set curve on the surface. By combining epsilon grid and marching square method, the marching square method can dynamically judge the required grid density while modeling the implicit surface, and realize accurate geometric modeling.

[0094] In an example, the parameter expression of the implicit surface in step S103 is {S i (u j )}, where S i represents the i-th implicit surface, u j represents the j-th parameter, and the geometric parameter expression of the cross section of the curve stiffening structure is {t p ,l i}, where t p is the thickness of the plate shell, and the remaining parameters are the cross section parameters of the curve stiffening structure. Referring to FIG. 8, l i refers to the parameters needed to determine the cross section geometric shape, which may be different for different cross sections. Figures 6 to 8

[0095] The cross section geometric parameters {t p ,l i} in step S103 are fixed according to the following formula:

[0096]

[0097] where S r , S p are the areas of the stiffening structure and the plate shell, I z ​is the cross-sectional moment of inertia of the corresponding curved stiffened structure about the neutral axis of the plate, l1, l2 are shape parameters of the cross-section, the ellipsis means more cross-sectional shape parameters (see Figures 6 to 8 may exist due to different cross-sectional shapes (see p t is the thickness of the plate, C0 is a selected constant;

[0098] In an example, the carrying efficiency of the initial sample set and the new sample set in step S203 is evaluated according to the following formula:

[0099]

[0100] Wherein, λ, λ0 are respectively the buckling eigenvalue of the plate shell after stiffening and before stiffening.

[0101] It should be noted that the formula in step S203 can be used to evaluate the carrying efficiency of the initial sample set and the carrying efficiency of the new sample set.

[0102] Referring to Figure 4 and Figure 5 are respectively the process diagrams of sparse curved stiffened structure and dense curved stiffened structure design under fixed support boundary conditions using the plate shell curved stiffened structure design method of the embodiment. Figure 4 (a.1) in Figure 5 (b.1) in Figure 4 (a.2) in Figure 5 (b.2) in Figure 4 (a.3) in Figure 5 (b.3) in The maximum value of the equation corresponds to the optimal design point, and it can be seen that the method gives a fast and efficient means for buckling mechanics evaluation of complex curved stiffened structures.

[0103] The embodiment also discloses a plate shell curved stiffened structure obtained based on any one of the plate shell curved stiffened structure design methods.

[0104] The present application does not describe the known or prior art.

[0105] Finally, it should be noted that the above examples are only used to illustrate the technical solutions of the present application, and are not intended to limit the same; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that not every example contains only one independent technical solution, and in the absence of solution conflicts, various technical features mentioned in each example can be combined in any manner to form other embodiments that can be understood by those skilled in the art.

[0106] In addition, modifications can be made to the technical solutions described in the foregoing examples, or equivalent replacements can be made to part of the technical features, without departing from the scope of the present application, so that the essence of the corresponding technical solution does not deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A design method for a plate and shell curved reinforced structure, characterized in that: The following steps are involved: S100: Define the design problem, including: S101: Determine the fixing method and force mode of the plate and shell; S102: Determine the maximum mass allowed for plates and shells in a given structural design; S103: Select the implicit surface and cross section of the curved reinforcement structure; S200: Optimization design of curved reinforced structure layout, including: S201: selecting relative stiffness parameters, fixing the geometric parameters of the cross section of the curved reinforced structure selected in S103, using the parameters of the implicit surface in S103 as design parameters, and generating an initial sample set required by the evolutionary algorithm; S202: performing geometric modeling on all sample points in the initial sample set using the E-AMS method according to implicit surfaces, and performing mechanical response analysis on the initial sample set; S203: Evaluate the load-bearing efficiency of the initial sample set to determine whether the convergence requirement is met. If the convergence requirement is met, output the curved reinforced structure layout. If the convergence requirement is not met, the evolutionary algorithm iteratively generates a new sample set, geometrically models all sample points in the new sample set using the E-AMS method based on implicit surfaces, and performs mechanical response analysis and load-bearing efficiency evaluation on the new sample set to determine whether the convergence requirement is met. This process is repeated until the convergence requirement is met. S300: Optimization design of curved reinforced structure cross-section parameters, including: S301: Sampling the sample space of the curved reinforced structure layout outputted in step S203 to generate a parameter set of the sample; S302: Modeling the parameter set of the sample generated in step S301 using the E-AMS method, and performing finite element buckling calculation on it using the finite element method to obtain the corresponding buckling load; S303: Select and train a proxy model; S304: Based on the proxy model trained in step S303, an optimization algorithm is used to calculate the maximized eigenvalue under equal quality constraints; S305: After the results converge, the final cross-sectional parameters of the curved reinforced structure are output; The surrogate model is trained based on the following buckling mode characteristic equation: in, are the buckling eigenvalues ​​of the plate and shell before and after reinforcement, is the dimensionless parameter of the rib morphology; is the dimensionless material constant, where are the elastic moduli of the reinforcement and the plate and shell, is the Poisson's ratio of the tendon; is the area integral of the global buckling modal displacement field over the reinforcement layout; is the dimensionless relative stiffness, are the areas of reinforcement layout and plate and shell respectively, is the section moment of inertia of the cross section of the corresponding curved reinforced structure around the neutral axis of the plate, is the shape parameter of the cross section, is the thickness of the plate, is the height of the tendon; The E-AMS method process is as follows: Generate the initial background grid required by the marching grid method, and then perform the first implicit curve modeling. By constructing an epsilon grid in the grid that captures the implicit curve, the grid is gradually encrypted according to the curvature distribution information provided by the epsilon grid point set distribution to obtain an accurate implicit curve spline.

2. The design method of plate and shell curved reinforced structure according to claim 1, characterized in that: The proxy model can be any common proxy model, using 、 Two parameters as input, As output, The number of parameters is equal to the number of cross-section parameters of the curved reinforcement structure minus 2.

3. The design method of plate and shell curved reinforced structure according to claim 1, characterized in that: In step S304, a direct search method is selected to calculate the maximum eigenvalue under equal quality constraints as follows; Objective function: in, The number of parameters is equal to the number of cross-sectional parameters of the curved reinforcement structure minus 2; Constraint function: ,in, is the cross-sectional area of ​​the curved reinforced structure, is the total length of the reinforcement.

4. The method for designing a curved reinforced plate and shell structure according to claim 1, wherein: The parameter expression of the implicit surface in step S103 is ,in represents the i-th implicit surface, Represents the jth parameter, and the geometric parameters of the cross section of the curved reinforced structure are expressed as ,in is the thickness of the plate shell, and the remaining parameters are the cross-sectional shape parameters of the curved reinforced structure; The cross-sectional geometric parameters in step S103 are fixed according to the following formula: : in are the areas of the reinforced structure and the plate shell, is the section moment of inertia of the cross section of the corresponding curved reinforced structure around the neutral axis of the plate, is the shape parameter of the cross section, is the thickness of the plate, is the selected constant.

5. The design method of plate and shell curved reinforced structure according to claim 1, characterized in that: In step S203, the carrying efficiency of the initial sample set and the new sample set is evaluated according to the following formula: in, are the buckling eigenvalues ​​of the plate and shell before and after reinforcement, respectively.

6. The design method of plate and shell curved reinforced structure according to claim 1, characterized in that: The evolutionary algorithm in step S201 is any one of a genetic algorithm, a simulated annealing algorithm, a particle swarm algorithm, an artificial neural network algorithm and an ant colony algorithm.

7. The method for designing a plate and shell curved reinforced structure according to claim 1, wherein: The mechanical response analysis in step S202 adopts finite element method, finite difference method or boundary element method; and / or The optimization algorithm in step S304 adopts a direct search method or a gradient descent method.

8. A plate and shell curved reinforced structure, characterized by: The plate and shell curved reinforced structure is obtained by the design method of the plate and shell curved reinforced structure according to any one of claims 1 to 7.

Images