Honeycomb sandwich composite material structure reliability optimization method considering interval uncertainty

Abstract

The invention discloses a honeycomb sandwich composite material structure reliability optimization method considering interval uncertainty. The method comprises the steps that honeycomb sandwich composite material design variables and uncertainty parameters are selected; the design variables and the uncertainty parameters are sampled; predicting equivalent mechanical parameters of the composite material single-layer plate; predicting honeycomb sandwich equivalent mechanical parameters; constructing a macroscale parameterized model of the honeycomb sandwich composite material structure; establishing a Kriging agent model of the honeycomb sandwich composite material structure; carrying out design variable global sensitivity analysis and uncertainty parameter global sensitivity analysis; constructing a honeycomb sandwich composite material structure reliability optimization model considering interval uncertainty; and performing optimization design on the honeycomb sandwich composite material structure by adopting a double-layer cycle reliability optimization method. According to the method, the multi-target genetic algorithm, the multi-scale modeling method and the agent model method are fused for reliability optimization design, and the robustness and optimization efficiency of the optimization result of the composite material structure can be improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of composite materials, and particularly relates to a multi-scale reliability optimization design method for honeycomb sandwich composite structures considering interval uncertainty. BACKGROUND

[0002] Honeycomb structure is a kind of porous material, compared with traditional solid materials, these structures have high porosity and low mass density, thus have the characteristics of light weight, high stiffness and excellent energy absorption. However, to fully exploit its potential advantages, not only the structure design needs to be optimized to improve the engineering performance, but also the long-term reliability under complex service environment must be ensured. The reliability optimization of honeycomb sandwich composite material not only concerns the reduction of structure mass, but also involves the synergistic improvement of material performance and structure mechanical performance, so as to ensure excellent bearing capacity and service reliability while meeting the requirement of light weight.

[0003] The reliability optimization of honeycomb sandwich composite material essentially depends on the accurate analysis of multi-scale correlation mechanism. Material parameters, fiber volume fraction at micro scale, cell wall thickness ratio at meso scale, and ply angle and geometric size at macro scale, etc. jointly determine the macro mechanical response of the structure. Studies have shown that although the macro homogenization method is efficient at the design stage, the effectiveness of the predicted results is not accurate enough when the material heterogeneity is significant. On the other hand, although the micro scale model can obtain more accurate local mechanical properties, the calculation cost is too high and the data redundancy is serious. Therefore, establishing an accurate relationship between the effective performance of composite materials and the component performance and microstructure parameters has been the focus and core goal of composite material research.

[0004] Although the multi-scale modeling method has made great progress in the mechanical analysis of composite materials, the multi-level structure characteristics of composite materials result in strong coupling effects between micro-meso-macro scale parameters. In addition, the fluctuation of manufacturing process further introduces the uncertainty of composite material parameters, making it difficult for traditional deterministic optimization methods to guarantee the reliability of the structure under extreme load conditions. In recent years, researchers have proposed a series of composite material reliability optimization methods considering uncertainty. However, most of the existing researches focus on uncertainty optimization based on probability theory, but in actual engineering, due to the difficulty in obtaining experimental data, it is difficult to construct an accurate probability distribution of uncertain parameters. Especially under the condition of small sample, the applicability of the uncertainty optimization method based on probability theory is limited. At present, there are few reports on the reliability optimization of honeycomb sandwich composite materials based on non-probabilistic methods. In addition, in the process of multi-scale reliability optimization design of composite materials, the quantification and propagation of multi-scale uncertainty parameters still face challenges. SUMMARY

[0005] The application aims to provide a multi-scale reliability optimization design method for honeycomb sandwich composite structures considering interval uncertainty, which can improve the reliability of the optimization results of honeycomb sandwich composite structures.

[0006] Technical scheme: To achieve the above-mentioned purpose, the application provides a reliability optimization method for honeycomb sandwich composite structures considering interval uncertainty, comprising the following steps:

[0007] S1: selecting honeycomb sandwich composite design variables and uncertainty parameters, determining the design variable value range, i.e., the design variable interval, and quantifying the uncertainty parameters by using interval variables;

[0008] S2: using a hybrid sampling technology combining Latin hypercube and full factorial experimental design to sample the design variables and uncertainty parameters in step S1;

[0009] S3: constructing a representative volume element model of the composite single-layer plate parameterization driven in the honeycomb sandwich composite material to predict the equivalent mechanical parameters of the composite single-layer plate;

[0010] S4: constructing a representative volume element model of the honeycomb sandwich parameterization driven in the honeycomb sandwich composite material to predict the equivalent mechanical parameters of the honeycomb sandwich;

[0011] S5: integrating the equivalent mechanical parameters of the composite single-layer plate and the honeycomb sandwich to construct a macro-scale parameterization model of the honeycomb sandwich composite structure;

[0012] S6: establishing a Kriging surrogate model of the honeycomb sandwich composite structure, wherein the sample set is obtained by the hybrid sampling in step S2, and the mechanical properties of the honeycomb sandwich composite structure are obtained by substituting the macro-scale parameterization model of the honeycomb sandwich composite structure in step S5;

[0013] S7: according to the Kriging surrogate model of the honeycomb sandwich composite structure, performing global sensitivity analysis on the design variables and global sensitivity analysis on the uncertainty parameters, respectively, retaining the parameter-sensitive design variables and uncertainty parameters, and taking the insensitive uncertainty parameters as constants;

[0014] S8: constructing a reliability optimization model of the honeycomb sandwich composite structure considering interval uncertainty for the Kriging surrogate model of the honeycomb sandwich composite structure after global sensitivity analysis;

[0015] S9: according to the reliability optimization model of the honeycomb sandwich composite structure, based on a multi-objective genetic algorithm, using a double-loop reliability optimization method to perform optimization design on the honeycomb sandwich composite structure.

[0016] Further, the sample quantity of the mixed sampling technique in the step S2 is N, the total quantity of design variables and uncertainty parameters is m, and the full-factor 2-level sampling is used in the full-factor experiment design, that is, the sample quantity of the full-factor experiment design is 2 m , and the sample quantity of the Latin hypercube sampling is N-2 m .

[0017] Further, the representative volume element model of the single-layer composite material parameterized driving in the step S3 comprises establishing a representative volume element model of the single-layer composite material, and applying a periodic displacement boundary condition to derive equivalent mechanical parameters of the single-layer composite material.

[0018] Further, the representative volume element model of the honeycomb sandwich parameterized driving in the step S4 comprises establishing a representative volume element model of the honeycomb sandwich, and applying a periodic displacement boundary condition to derive equivalent mechanical parameters of the honeycomb sandwich.

[0019] Further, in the macro-scale parameterized model of the honeycomb sandwich composite structure in the step S5, the upper panel and the lower panel are established by using plate elements, wherein the material parameters of the single-layer composite material are the equivalent mechanical parameters of the single-layer composite material obtained in the step S3, and the honeycomb sandwich is established by using solid elements, wherein the material parameters of the honeycomb sandwich are the equivalent mechanical parameters of the honeycomb sandwich obtained in the step S4.

[0020] Further, in the step S7:

[0021] In the global sensitivity analysis process of the design variables, the uncertainty parameters are fixed values, the design variables are interval variables, and the sensitive design variables are screened out through the global sensitivity analysis.

[0022] In the global sensitivity analysis process of the uncertainty parameters, the design variables are fixed values, the uncertainty parameters are interval variables, and the sensitive uncertainty parameters are screened out through the global sensitivity analysis.

[0023] Further, in the step S8, the reliability optimization model of the honeycomb sandwich composite structure considering the interval uncertainty is expressed as:

[0024]

[0025] wherein, is a set of sensitive design variables, is the number of sensitive design variables, is a set of sensitive uncertainty parameters; is the number of sensitive uncertainty parameters, f i w and f i c are the median and radius of the i-th objective function, respectively, N fis the number of objective functions, α i is the i-th design variable, and α i are the design variables α i The upper and lower boundaries of the design interval, u i is the i-th uncertainty parameter, and u i is the uncertainty parameter u i The upper and lower boundaries of is the response interval of the model, is the constraint function S i The interval form of the allowable value, ξ i is the interval possibility level that the i-th interval constraint should satisfy, β is the weight parameter, P(*) is the possibility operator, The possibility calculation expression is:

[0026]

[0027] Furthermore, the double-layer cycle reliability optimization method for the honeycomb sandwich composite structure in step S9 includes two parts: inner cycle uncertainty analysis and outer cycle optimization. The outer cycle optimization adopts a multi-objective genetic algorithm, and the objective function of its population is obtained by the inner cycle uncertainty analysis.

[0028] The process of inner loop uncertainty analysis includes:

[0029] A1: Determine the number of samples for the uncertainty parameter and use a mixed sampling technique to randomly generate a group of samples;

[0030] A2: Extract a sample from the sample set;

[0031] A3: Establish a parameter-driven representative volume element model of a composite single-layer plate and calculate the equivalent mechanical parameters of the composite single-layer plate;

[0032] A4: Establish a parametrically driven representative volume element model of the honeycomb sandwich composite material and calculate the equivalent mechanical parameters of the honeycomb sandwich core;

[0033] A5: Establish a Kriging proxy model for honeycomb sandwich composite structures to calculate the mechanical properties, quality, and cost of honeycomb sandwich composite structures;

[0034] A6: Check whether the sample has been used up; if not, return to step A2; if yes, go to step A7;

[0035] A7: Using all sample results, calculate the possible range of mechanical properties, mass, and cost of honeycomb sandwich composite structures;

[0036] A8: Calculate the constraint function possible degree level, the median and radius of the objective function;

[0037] A9: Calculate the optimization target based on the penalty function method.

[0038] Advantages: Compared with the prior art, the honeycomb sandwich composite reliability optimization efficiency is low, and the prior art is mostly concentrated in the uncertainty optimization method based on probability theory, the present application considers the small sample condition, adopts interval theory to quantify the uncertainty of parameters, and fuses multi-scale modeling technology and proxy model method, and proposes a multi-scale reliability optimization design method for honeycomb sandwich composite materials with interval uncertainty; The method has the following advantages:

[0039] 1、The present application uses interval theory to quantify the uncertainty of material parameters and structural parameters in honeycomb sandwich composite structure to establish a honeycomb sandwich composite reliability optimization model, thereby improving the reliability of the honeycomb sandwich composite.

[0040] 2、The present application fuses multi-scale modeling technology and proxy model method, which greatly improves the reliability optimization efficiency of the honeycomb sandwich composite.

[0041] 3、The present application reduces the number of design variables and uncertainty parameters by performing global sensitivity analysis on the design variables and uncertainty parameters, thereby further improving the reliability optimization efficiency of the honeycomb sandwich composite.

[0042] 4、The present application proposes a double-layer cyclic reliability optimization method, which adopts outer multi-objective genetic algorithm optimization and inner honeycomb sandwich composite uncertainty analysis nested loop, thereby providing an effective method for honeycomb sandwich composite reliability optimization design. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 The flowchart of the present application method;

[0044] Figure 2 The schematic diagram of the representative volume element model of the composite single-layer plate;

[0045] Figure 3 The schematic diagram of the representative volume element model of the honeycomb sandwich;

[0046] Figure 4 The schematic diagram of the three-point bending of the honeycomb sandwich composite material;

[0047] Figure 5 The result diagram of the Pareto front distribution under different possible degrees in the embodiment. DETAILED DESCRIPTION

[0048] The present application will be further clarified by the following examples, which should not be construed as limiting the scope of the present application. Various modifications of the present application, in addition to those shown and described herein will become apparent to those skilled in the art from the following examples, and it is intended to cover any and all such modifications as fall within the scope of the appended claims.

[0049] Embodiment 1

[0050] As shown in the following, the embodiment provides a multi-scale reliability optimization design method for honeycomb sandwich composite structure considering interval uncertainty, comprising the following steps: Figure 1 S1: selecting design variables and uncertainty parameters of honeycomb sandwich composite material, determining the value range of design variables, i.e. the interval of design variables, and quantifying uncertainty parameters by interval variables;

[0051] S2: using a hybrid sampling technique combining Latin hypercube and full factorial experimental design to sample design variables and uncertainty parameters in step S1;

[0052] The number of samples of the hybrid sampling technique in the embodiment is N, the total number of design variables and uncertainty parameters is m, the full factorial experimental design uses full factorial 2-level sampling, i.e. the number of samples of full factorial experimental design is 2 m , and the number of samples of Latin hypercube sampling is N-2 m .

[0053] The hybrid sampling method ensures the accuracy of the prediction results of the established surrogate model at the interval boundaries by considering the interval boundary values as sample points through full factorial experimental design. The interval boundary values have a high probability of being the maximum and minimum values. If only full factorial experimental design is used, there will be a huge number of samples in order to construct a surrogate model that meets the accuracy requirements in the case of a large number of parameters. The present application combines the two sampling methods to solve this problem.

[0054] S3: constructing a representative volume element model of composite single-layer plate parameterization driving in honeycomb sandwich composite material to predict the equivalent mechanical parameters of the composite single-layer plate;

[0055] The embodiment of constructing a representative volume element model of composite single-layer plate parameterization driving in honeycomb sandwich composite material comprises selecting a square distribution as shown in (a) of

[0056] and applying periodic displacement boundary conditions as shown in (b-g) of Figure 2 to derive the equivalent mechanical parameters of the single-layer plate. Figure 2 S4: constructing a representative volume element model of honeycomb sandwich parameterization driving in honeycomb sandwich composite material to predict the equivalent mechanical parameters of the honeycomb sandwich;

[0057] ​

[0058] The embodiment constructs a representative volume element model of the honeycomb sandwich parameterization driving in the honeycomb sandwich composite material, which includes selecting a honeycomb sandwich representative volume element as Figure 3 As shown in (a) of c h is the honeycomb sandwich thickness, t is the honeycomb cell wall thickness, and l is the honeycomb cell edge length. Because the honeycomb sandwich is of equal wall thickness, the wall thickness of the upper and lower edges of the representative volume element model is t / 2. The finite element model of the honeycomb sandwich representative volume element is established as shown in (b) of Figure 3 To more conveniently apply the periodic displacement boundary conditions, the honeycomb sandwich representative volume element model is filled with a cubic configuration using "elastic air".

[0059] S5: By integrating the equivalent mechanical parameters of the composite single-layer plate and the honeycomb sandwich, a macro-scale parameterization model of the honeycomb sandwich composite structure is constructed;

[0060] In the process of constructing the macro-scale parameterization model of the honeycomb sandwich composite structure, the upper and lower panels are established using plate elements, wherein the single-layer plate material parameters are the equivalent mechanical parameters of the composite single-layer plate obtained in step S3, and the honeycomb sandwich is established using solid elements, wherein the honeycomb sandwich material parameters are the equivalent mechanical parameters of the honeycomb sandwich obtained in step S4.

[0061] S6: A Kriging surrogate model of the honeycomb sandwich composite structure is established to improve the calculation efficiency, wherein the sample set is obtained by hybrid sampling in step S2, and the mechanical properties of the honeycomb sandwich composite structure are obtained by substituting the sample set into the macro-scale parameterization model of the honeycomb sandwich composite structure in step S5.

[0062] S7: According to the Kriging surrogate model of the honeycomb sandwich composite structure, global sensitivity analysis of the design variables and global sensitivity analysis of the uncertainty parameters are respectively performed, the sensitive design variables and uncertainty parameters are retained, and the insensitive uncertainty parameters are taken as fixed values.

[0063] In the process of global sensitivity analysis of the design variables, the uncertainty parameters are taken as fixed values, the design variables are interval variables, and the sensitive design variables are screened out through global sensitivity analysis.

[0064] In the process of global sensitivity analysis of the uncertainty parameters, the design variables are taken as fixed values, the uncertainty parameters are interval variables, and the sensitive uncertainty parameters are screened out through global sensitivity analysis.

[0065] S8: A reliability optimization model of the honeycomb sandwich composite structure considering interval uncertainty is constructed for the Kriging surrogate model of the honeycomb sandwich composite structure after global sensitivity analysis.

[0066]

[0067] wherein, is a set of sensitivity design variables, is the number of sensitivity design variables, is a set of sensitivity uncertainty parameters; is the number of sensitivity uncertainty parameters, i w i c is the median and radius of the ith objective function, respectively, N f is the number of objective functions, α i is the ith design variable, α i is the upper and lower bound of the design interval of the design variable α i i is the ith uncertainty parameter, u i is the upper and lower bound of the uncertainty parameter u i is the response interval of the model, is the constraint function S i is the interval form of the allowable value, ξ i is the interval possibility level that the ith interval constraint should satisfy, β is the weight parameter, and P is the possibility operator, The possibility calculation expression of S is:

[0068]

[0069] S9: According to the reliability optimization model of the honeycomb sandwich composite structure, a double-loop reliability optimization method is used to optimize the design of the honeycomb sandwich composite structure based on the multi-objective genetic algorithm.

[0070] The double-loop reliability optimization method for the honeycomb sandwich composite structure includes inner-loop uncertainty analysis and outer-loop optimization. The multi-objective genetic algorithm is used for outer-loop optimization, and the objective function of the population is obtained by inner-loop uncertainty analysis.

[0071] The process of inner-loop uncertainty analysis includes:

[0072] A1: Determine the sample size of the uncertainty parameter, and randomly generate a set of samples using the mixed sampling technique;

[0073] A2: Extract a sample from the sample set;

[0074] A3: Establish a parameterized driving representative volume element model of the composite single-layer plate, and calculate the equivalent mechanical parameters of the composite single-layer plate;

[0075] ​​​​​A4: Establish a representative volume element model of the honeycomb sandwich parameterized driving in the honeycomb sandwich composite material, and calculate the equivalent mechanical parameters of the honeycomb sandwich;

[0076] A5: Establish a Kriging surrogate model of the honeycomb sandwich composite structure, and calculate the mechanical properties, mass and cost of the honeycomb sandwich composite structure;

[0077] A6: Check whether all the samples are used up; if not, return to step A2; if yes, execute step A7;

[0078] A7: Calculate the possible range of the mechanical properties, mass and cost of the honeycomb sandwich composite structure by using all the sample results;

[0079] A8: Calculate the possible level of the constraint function, the median and the radius of the objective function;

[0080] A9: Calculate the optimization target based on the penalty function method.

[0081] Example 2:

[0082] In order to verify the effectiveness and effect of the method of the application, the method of the application is applied in this embodiment, as follows:

[0083] In this embodiment, the three-point bending of the honeycomb sandwich composite material is taken as an example, and the schematic diagram of the sample is shown in Figure 4 The upper and lower panels are glass fiber / epoxy composite materials, which are symmetrically laid and have a total of 8 layers, and the honeycomb sandwich is aramid paper. The honeycomb sandwich composite structure is completely symmetrical, the left and right ends are free ends, the lower panel is a simply supported boundary condition, and the pressure head is subjected to Z-direction displacement load. The design variables considered in the optimization design include fiber volume fraction, upper panel thickness, lower panel thickness, honeycomb cell wall thickness, honeycomb cell edge length, and honeycomb sandwich thickness, a total of 7 variables. The uncertainty parameters considered include fiber volume fraction, material parameters, and geometric dimensions, a total of 13 variables. The initial values of the design variables and the upper and lower boundaries of the design interval are shown in Table 1. The nominal values of the uncertainty parameters and the upper and lower boundaries of the uncertainty interval are shown in Table 2. The densities and costs of the constituent materials are shown in Table 3.

[0084] Table 1 Initial values and design boundaries of aramid paper honeycomb sandwich composite design variables

[0085]

[0086] Table 2 Nominal values and boundaries of aramid paper honeycomb sandwich composite uncertainty parameters

[0087]

[0088] Table 3 Densities and costs of constituent materials

[0089]

[0090] In order to evaluate the influence of reliability on the optimization results, the present application respectively takes the reliability as 50%, 60%, 70%, 75%, 80%, 85%, 90%, 95% and 99% as the pre-given interval probability level, and the Pareto front of the reliability optimization result is as shown in Figure 5 The results show that the reliability has a significant influence on the optimization results, the Pareto front under different reliability levels forms a layered arrangement from the lower left to the upper right, and almost no intersection, showing the characteristics of progressive change in the solution space with the increase of reliability. Under the lower reliability requirement (such as 50% to 75%), the optimization solution is mainly concentrated in the area with lower cost and quality; and with the increase of reliability requirement (such as 90% to 99%), the optimization solution gradually migrates to the area with higher cost and quality. This phenomenon shows that the higher the reliability, the more conservative the design scheme, and the more resource input required. It is shown that in the actual engineering design, the increase of reliability requirement inevitably leads to the increase of resource input, therefore, in the optimization design stage, the relationship among reliability, quality and cost needs to be comprehensively balanced, and a balanced scheme suitable for the actual engineering requirement is sought.

Claims

1. A reliability optimization method for honeycomb sandwich composite materials considering interval uncertainty, characterized in that: The steps include: S1: Select the design variables and uncertainty parameters of the honeycomb sandwich composite material, determine the value range of the design variables, that is, the design variable interval, and use interval variables to quantify the uncertainty parameters; S2: Sampling the design variables and uncertainty parameters in step S1 using a hybrid sampling technique that combines Latin hypercube and full factorial experimental design; S3: Construct a parameter-driven representative volume element model of a composite single-layer plate in a honeycomb sandwich composite material to predict the equivalent mechanical parameters of the composite single-layer plate; S4: Construct a parameter-driven representative volume element model of the honeycomb core in honeycomb sandwich composites to predict the equivalent mechanical parameters of the honeycomb core; S5: By integrating the equivalent mechanical parameters of composite single-layer plates and honeycomb sandwiches, a macro-scale parametric model of honeycomb sandwich composite structures is constructed; S6: establishing a Kriging proxy model for the honeycomb sandwich composite material structure, wherein the sample set is obtained by mixed sampling in step S2, and is substituted into the macro-scale parameterized model of the honeycomb sandwich composite material structure in step S5 to obtain the mechanical properties of the honeycomb sandwich composite material structure; S7: Based on the Kriging proxy model of honeycomb sandwich composite structure, global sensitivity analysis of design variables and global sensitivity analysis of uncertainty parameters are performed respectively. Sensitive design variables and uncertainty parameters are retained, and insensitive uncertainty parameters are taken as constant values. S8: Based on the Kriging proxy model of honeycomb sandwich composite structure after global sensitivity analysis, a reliability optimization model of honeycomb sandwich composite structure considering interval uncertainty is constructed; S9: According to the reliability optimization model of honeycomb sandwich composite material structure, based on the multi-objective genetic algorithm, a double-layer cyclic reliability optimization method is used to optimize the design of honeycomb sandwich composite material structure.

2. The reliability optimization method of honeycomb sandwich composite material structure considering interval uncertainty according to claim 1, characterized in that: In step S2, the number of samples of the mixed sampling technique is N, the total number of design variables and uncertainty parameters is m, and the full factorial experimental design adopts full factor 2 level sampling, that is, the number of samples of the full factorial experimental design is 2 m , the number of Latin hypercube sampling samples is N-2 m .

3. The reliability optimization method of honeycomb sandwich composite material structure considering interval uncertainty according to claim 1, characterized in that: The parameter-driven representative volume element model of the composite material single-layer plate in step S3 includes establishing a representative volume element model of the single-layer plate and applying periodic displacement boundary conditions to derive equivalent mechanical parameters of the single-layer plate.

4. The reliability optimization method of honeycomb sandwich composite material structure considering interval uncertainty according to claim 1, characterized in that: The parameter-driven representative volume element model of the honeycomb sandwich core in step S4 includes establishing a representative volume element model of the honeycomb sandwich core and applying periodic displacement boundary conditions to derive equivalent mechanical parameters of the honeycomb sandwich core.

5. The reliability optimization method of honeycomb sandwich composite material structure considering interval uncertainty according to claim 1, characterized in that: In the macro-scale parametric model of the honeycomb sandwich composite material structure in step S5, the upper panel and the lower panel are established using plate units, wherein the material parameters of the single-layer plate are the equivalent mechanical parameters of the composite material single-layer plate obtained in step S3, and the honeycomb sandwich is established using solid units, wherein the material parameters of the honeycomb sandwich are the equivalent mechanical parameters of the honeycomb sandwich obtained in step S4.

6. The reliability optimization method of honeycomb sandwich composite material structure considering interval uncertainty according to claim 1, characterized in that: In step S7: In the process of global sensitivity analysis of design variables, the uncertainty parameters are taken as fixed values, the design variables are interval variables, and the sensitive design variables are screened out through global sensitivity analysis; In the process of global sensitivity analysis of uncertainty parameters, the design variables take their fixed values, the uncertainty parameters are interval variables, and the sensitive uncertainty parameters are screened out through global sensitivity analysis.

7. The reliability optimization method of honeycomb sandwich composite material structure considering interval uncertainty according to claim 1, characterized in that: The reliability optimization model of the honeycomb sandwich composite material structure considering interval uncertainty in step S8 is expressed as: in, is the set of sensitivity design variables, is the number of sensitivity design variables, is the set of sensitivity uncertainty parameters; is the number of sensitivity uncertainty parameters, and are the median and radius of the i-th objective function, N f is the number of objective functions, α i is the i-th design variable, and α i are the design variables α i The upper and lower boundaries of the design interval, u i is the i-th uncertainty parameter, and u i is the uncertainty parameter u i The upper and lower boundaries of is the response interval of the model, is the constraint function S i The interval form of the allowable value, ξ i is the interval possibility level that the i-th interval constraint should satisfy, β is the weight parameter, P{*} is the possibility operator, The possibility calculation expression is:

8. The reliability optimization method of honeycomb sandwich composite material structure considering interval uncertainty according to claim 1, characterized in that: The double-layer loop reliability optimization method in step S9 includes two parts: inner loop uncertainty analysis and outer loop optimization; the outer loop optimization adopts a multi-objective genetic algorithm, and the objective function of its population is obtained by the inner loop uncertainty analysis.

9. The reliability optimization method of honeycomb sandwich composite material structure considering interval uncertainty according to claim 8, characterized in that: The process of inner loop uncertainty analysis in step S9 includes: A1: Determine the number of samples for the uncertainty parameter and use a mixed sampling technique to randomly generate a group of samples; A2: Extract a sample from the sample set; A3: Establish a parameter-driven representative volume element model of a composite single-layer plate and calculate the equivalent mechanical parameters of the composite single-layer plate; A4: Establish a parametrically driven representative volume element model of the honeycomb sandwich composite material and calculate the equivalent mechanical parameters of the honeycomb sandwich core; A5: Establish a Kriging proxy model for honeycomb sandwich composite structures to calculate the mechanical properties, quality, and cost of honeycomb sandwich composite structures; A6: Check whether the sample has been used up; if not, return to step A2; if yes, go to step A7; A7: Using all sample results, calculate the possible range of mechanical properties, mass, and cost of honeycomb sandwich composite structures; A8: Calculate the likelihood level of the constraint function, the median value and radius of the objective function; A9: Calculate the optimization objective based on the penalty function method.

Images