Light-weight glass container structure optimization design method based on three-dimensional simulation model

Abstract

The invention relates to the technical field of glass container structure optimization, in particular to a lightweight glass container structure optimization design method based on a three-dimensional simulation model. The method comprises the following steps: constructing a continuous curved surface model of the glass container; constructing a glass container geometric variant library based on the continuous curved surface model of the glass container; performing multi-working-condition simulation on the glass container geometric variant library, and performing stress distribution feature extraction to obtain a container stress distribution feature set; constructing a container forming process constraint set; and performing NURBS control point identification on the glass container according to the container forming process constraint set to obtain a constraint container NURBS control point set. On the premise of ensuring the mechanical property and manufacturing feasibility of the glass container, the lightweight design can be realized to the greatest extent.

Description

Technical Field

[0001] The present invention relates to the technical field of optimizing the structure of glass containers, and particularly to a method for optimizing the design of a lightweight glass container structure based on a three-dimensional simulation model. Background Art

[0002] As an important packaging form for food, medicine, and chemical products, the structural design of glass containers needs to consider lightweight, mechanical properties, and manufacturing feasibility. In recent years, with the popularization of computer-aided design (CAD) and finite element analysis (FEA) technologies, the traditional experience-driven manual design mode has gradually been replaced by parametric modeling and simulation verification. However, there are still significant bottlenecks in the existing methods when dealing with the coupled effects of glass material properties and complex working conditions. Glass is a typical brittle material, and its fracture behavior is extremely sensitive to local stress concentration. In lightweight design, although traditional uniform wall thinning or simple topology optimization methods can reduce the material usage to a certain extent, they are difficult to effectively suppress stress peaks. This results in the fact that in actual use, glass containers are prone to rupture due to excessive local stress, seriously affecting their reliability and safety. In addition, the forming process of glass containers also has strict requirements on geometric features such as wall thickness gradient and curvature transition. Existing simulation-driven multi-objective optimization models often lack dynamic integration of these manufacturability indicators, resulting in a large deviation between the simulation results and the performance of the actually manufactured products. Summary of the Invention

[0003] Based on this, it is necessary for the present invention to provide a method and system for optimizing the design of a lightweight glass container structure based on a three-dimensional simulation model to solve at least one of the above technical problems.

[0004] To achieve the above object, a method for optimizing the design of a lightweight glass container structure based on a three-dimensional simulation model includes the following steps: Step S1: Construct a continuous surface model of the glass container; construct a geometric variant library of the glass container based on the continuous surface model of the glass container; perform multi-condition simulations on the geometric variant library of the glass container, and extract stress distribution characteristics to obtain a set of container stress distribution characteristics; Step S2: Construct a set of constraints for the container forming process; identify the NURBS control points of the glass container according to the set of constraints for the container forming process to obtain a set of NURBS control points of the constrained container; perform gradient-conserving deformation regulation on the set of NURBS control points of the constrained container to obtain a container wall thickness gradient constraint model; generate a distribution map of the stress-sensitive area of the container based on the container wall thickness gradient constraint model; Step S3: Construct the objective function for minimizing the mass of the glass container, the objective function for maximizing the strength, and the manufacturing feasibility function; construct a forming multi-objective optimization model based on the objective function for minimizing the mass, the objective function for maximizing the strength, and the manufacturing feasibility function; solve the forming multi-objective optimization model to obtain the design parameter set of the glass container; Step S4: Construct a parametric 3D model of the glass container according to the design parameter set of the glass container; conduct a forming performance evaluation on the parametric 3D model of the glass container to obtain the forming simulation performance data; conduct a parameter perturbation analysis on the design parameter set of the glass container to obtain the robustness index evaluation result; Step S5: Generate design feedback adjustment suggestions based on the forming simulation performance data and the robustness index evaluation result; generate a lightweight glass container structure design scheme according to the design feedback adjustment suggestions.

[0005] Through constructing a continuous surface model and a parametric modeling function set of the glass container, and generating a geometric variant library based on this for multi-condition simulation, the present invention can accurately extract the stress distribution feature vector set, thereby providing a detailed and comprehensive stress data basis for subsequent optimization design, effectively solving the problems of difficult suppression of stress peaks and insufficient structural reliability in the prior art due to the lack of in-depth analysis of the coupling effect between the glass material properties and complex working conditions. By constructing a set of forming process constraints for the container, and accordingly conducting NURBS control point identification and gradient-conserved deformation regulation, a wall thickness gradient constraint model and a stress-sensitive area distribution map are generated, realizing the accurate control of the forming process constraints of the glass container, effectively making up for the defect that the existing simulation-driven multi-objective optimization model lacks the dynamic integration of manufacturability indicators, and significantly reducing the deviation between the simulation results and the actual product performance. By constructing a forming multi-objective optimization model and solving it to obtain the design parameter set, comprehensively considering mass minimization, strength maximization, and manufacturing feasibility, it is possible to maximize the lightweight design on the premise of ensuring the mechanical properties and manufacturing feasibility of the glass container, effectively balancing the relationship between lightweight, performance, and manufacturability. By generating design feedback adjustment suggestions based on the forming simulation performance data and the robustness index evaluation result, and accordingly adjusting the structure of the parametric 3D model of the glass container, the design scheme is further optimized, the robustness and reliability of the design are improved, and the stability and durability of the glass container in actual use are ensured. Brief Description of the Drawings

[0006] Other features, objectives, and advantages of the present invention will become more obvious by reading the following detailed description with reference to the accompanying drawings: Figure 1 The step flow diagram of the lightweight glass container structure optimization design method based on a 3D simulation model in an embodiment is shown.

[0007] Figure 2Shows the detailed step - by - step schematic diagram of step S17 of an embodiment. Detailed implementation manner

[0008] The technical method of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0009] In addition, the accompanying drawings are only schematic diagrams of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, so repeated descriptions of them will be omitted. Some of the block diagrams shown in the figures are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor methods and / or microcontroller methods.

[0010] It should be understood that although terms such as "first", "second", etc. may be used here to describe each unit, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, the first unit can be called the second unit, and similarly the second unit can be called the first unit. The term "and / or" used here includes any and all combinations of one or more of the listed related items.

[0011] To achieve the above - mentioned purpose, please refer to Figures 1 to 2 , the present invention provides a lightweight glass container structure optimization design method based on a three - dimensional simulation model, including the following steps: Step S1: Construct a continuous surface model of the glass container; construct a geometric variant library of the glass container based on the continuous surface model of the glass container; perform multi - condition simulations on the geometric variant library of the glass container and extract stress distribution characteristics to obtain a container stress distribution characteristic set; Step S2: Construct a set of container forming process constraints; identify NURBS control points of the glass container according to the set of container forming process constraints to obtain a set of NURBS control points of the constrained container; perform gradient - conservation deformation regulation on the set of NURBS control points of the constrained container to obtain a container wall - thickness gradient constraint model; generate a distribution map of the stress - sensitive area of the container based on the container wall - thickness gradient constraint model; Step S3: Construct the objective function for minimizing the mass, the objective function for maximizing the strength, and the manufacturing feasibility function of the glass container; construct a forming multi-objective optimization model based on the objective function for minimizing the mass, the objective function for maximizing the strength, and the manufacturing feasibility function; solve the forming multi-objective optimization model to obtain the glass container design parameter set; Step S4: Construct a parametric 3D model of the glass container according to the glass container design parameter set; conduct a forming performance evaluation on the parametric 3D model of the glass container to obtain forming simulation performance data; conduct a parameter perturbation analysis on the glass container design parameter set to obtain the robustness index evaluation result; Step S5: Generate design feedback adjustment suggestions based on the forming simulation performance data and the robustness index evaluation result; generate a lightweight glass container structural design scheme according to the design feedback adjustment suggestions.

[0012] In this embodiment, a continuous surface model of a glass container is constructed using Rhinoceros software in combination with the Grasshopper plug-in. By importing the two-dimensional sketch or point cloud data of the glass container, a smooth three-dimensional surface model is created using the modeling function of Rhinoceros. Based on this model, a set of parametric modeling functions is defined, including Bezier curve functions for controlling the bottle body contour, radial thickness functions for describing the wall thickness distribution, and spline interpolation functions for defining the transition region. Using these functions, a geometric variant library containing multiple design variants is generated, covering parameter variations such as different bottle body heights, bottleneck wall thicknesses, and bottle bottom curvatures. The models in the geometric variant library are subjected to multi-condition simulations using ANSYS Mechanical, including conditions such as vertical compression, lateral extrusion, internal pressure, and thermal shock. The stress distribution characteristics of each variant are extracted through finite element analysis to generate a set of stress distribution characteristic vectors. At the same time, in combination with the production process data provided by the glass container manufacturer, MATLAB is used to construct a set of container forming process constraints, covering glass forming fluidity, mold demolding conditions, and thermal gradient constraints. Based on these constraint conditions, NURBS control point identification is performed on the glass container in Rhinoceros to obtain a set of constrained container NURBS control points, and a container wall thickness gradient constraint model is generated through gradient-conserving deformation regulation, and further a stress-sensitive region distribution map is generated. MATLAB is used to construct the objective functions of minimizing the mass of the glass container, maximizing the strength, and the manufacturing feasibility function. By integrating these objective functions, a multi-objective optimization model for forming is constructed, and the optimization toolbox of MATLAB (such as fmincon or gamultiobj) is used for solution to obtain an optimized set of glass container design parameters. Based on these parameters, a parametric three-dimensional model of the glass container is constructed in Rhinoceros, and computational fluid dynamics simulations are performed using ANSYS Fluent, combined with structural analysis using ANSYS Mechanical to obtain forming simulation performance data. At the same time, MATLAB is used to perform parametric perturbation analysis on the set of design parameters to evaluate the robustness of the design scheme and generate the evaluation results of the robustness index. Finally, according to the forming simulation performance data and the evaluation results of the robustness index, Excel and MATLAB are used to generate design feedback adjustment suggestions. These suggestions include adjusting the bottleneck wall thickness, optimizing the bottle body height, etc. According to these suggestions, the parametric three-dimensional model of the glass container is structurally adjusted in Rhinoceros and Grasshopper, and finally a lightweight and excellent-performance glass container structural design scheme is obtained.

[0013] Preferably, step S1 includes the following steps: Step S11: Obtain a list of common glass containers; collect the surface topography of each common glass container according to the list of common glass containers to obtain a glass container point cloud data set; Step S12: filtering out noise points on the glass container point cloud dataset to obtain a smooth glass container point cloud dataset; Step S13: performing NURBS surface fitting reconstruction based on the smoothed glass container point cloud data set to obtain a continuous surface model of the glass container; Step S14: Structural partitioning of the continuous curved surface model of the glass container is performed to obtain curved surface blocks of the bottleneck area, the bottle shoulder area, the bottle body area and the bottle bottom area; Step S15: extracting characteristic dimensions of the curved surface blocks of the bottleneck area, the bottle shoulder area, the bottle body area and the bottle bottom area to obtain a set of geometric characteristic parameters of the glass container; Step S16: constructing a glass container modeling function set according to the glass container geometric feature parameter set, wherein the glass container modeling function set includes a bottleneck area thermal-mechanical coupling gradient function, a shoulder area rheological stress transfer function, a body area wall thickness gradient constraint function, and a bottom area contact stress attenuation function. The specific construction process of the glass container modeling function set is as follows: Identify the thermal stress concentration area in the bottleneck area of ​​the glass container and obtain the thermal stress distribution data of the bottleneck area; Layout the mold cooling channel according to the thermal stress distribution data of the bottleneck area to obtain the bottleneck-mold cooling channel mapping data; Based on the bottleneck-mold cooling channel mapping data, thermal-mechanical coupling gradient modeling is performed to obtain the thermal-mechanical coupling gradient function of the bottleneck area; The transition curvature of the shoulder area of ​​the glass container is identified, and the pressure field of the blow molding process is inverted to obtain the transition curvature field of the shoulder area and the pressure gradient parameters of the shoulder blow molding; According to the transition curvature field of the bottle shoulder area and the pressure gradient parameters of the bottle shoulder blow molding, the curved surface rheological characteristics of the glass container are constrained, and the rheological stress transfer function of the bottle shoulder area is obtained; Decompose the stacking load direction vector field of the bottle body area of ​​the glass container to obtain the load direction vector diagram of the bottle body area; Locate the mold parting line of the bottle body area of ​​the glass container to obtain the bottle body-mold parting line space topology data; Based on the load direction vector diagram of the bottle body area and the spatial topological data of the bottle body-mold parting line, the wall thickness-parting line dynamic coupling modeling is carried out to obtain the wall thickness gradient constraint function of the bottle body area; Conduct mold ejector pin distribution contact stress identification on the bottom area of ​​the glass container to obtain ejector pin distribution constraint data in the bottom area; Stress attenuation gradient modeling is performed based on the ejector pin distribution constraint data in the bottom area of ​​the bottle to obtain the contact stress attenuation function in the bottom area of ​​the bottle; Step S17: constructing a glass container geometric variant library based on the glass container modeling function set; performing multi-condition simulation on the glass container geometric variant library, and extracting stress distribution features to obtain a container stress distribution feature set.

[0014] In this embodiment, a list of common glass wine bottles, beverage bottles and medicine bottles on the market is obtained. The surface morphology of each common glass container is collected using a three-dimensional scanner. Taking a glass wine bottle as an example, place it on the working platform of the scanner, start the scanner and set a suitable scanning resolution (such as 0.1mm). The scanner uses laser or structured light technology to perform an all-round scan of the surface of the wine bottle to collect high-precision point cloud data. These point cloud data are stored in a standard point cloud file format (such as .ply or .pcd) to form a glass container point cloud data set. After obtaining the glass container point cloud data set, the data is filtered out of noise using point cloud processing software (such as CloudCompare). Taking the point cloud data of a glass beverage bottle as an example, open the software and import the point cloud file. Select the "Filter" function module in the software, set the filtering algorithm to "Voxel Grid Filter", and set the voxel size to 0.5mm to remove discrete noise caused by the accuracy limitation of the scanning equipment or environmental interference. After processing by this method, a smooth glass beverage bottle point cloud data set is obtained. Based on the smoothed point cloud data set of glass containers, the NURBS surface fitting reconstruction tool (such as the NURBS modeling function in the Rhinoceros software) is used to construct a continuous surface model of the glass container. Taking the glass medicine bottle as an example, the smoothed point cloud data is imported into the Rhinoceros software, and the "Create Surface from Point Cloud" function is selected. The fitting accuracy is set to 0.05mm, and the software automatically fits a smooth and continuous NURBS surface model through calculation. This model can accurately reflect the geometric shape of the medicine bottle, including the surface features of the bottleneck, bottle body and bottle bottom. After obtaining the continuous surface model of the glass container, use 3D modeling software (such as SolidWorks) to partition its structure. Taking the glass wine bottle as an example, the continuous surface model is imported into the SolidWorks software. Using the "Split Surface" function of the software, according to the geometric features of the wine bottle, the surface blocks of the bottleneck area, bottle shoulder area, bottle body area and bottle bottom area are manually or automatically identified and divided. In the division process, the position and direction of the dividing line are set to ensure that each area is divided accurately. After completing the structural partitioning of the glass container, use 3D measurement software (such as GeomagicControl) to extract the characteristic dimensions of each area. Take the glass beverage bottle as an example, import the partitioned surface model into the software. Select the "Measure" function in the software to accurately measure the characteristic dimensions such as the diameter of the bottleneck area, the curvature of the bottle shoulder area, the height of the bottle body area, and the thickness of the bottle bottom area. The measurement results are displayed in the software interface in numerical form and recorded to form a set of geometric characteristic parameters of the glass container.

[0015] ANSYS Mechanical is used to identify the thermal stress concentration area in the bottleneck area of ​​the glass container. Taking a glass wine bottle as an example, first import the geometric feature parameters in ANSYS Mechanical and set the material properties (such as the thermal expansion coefficient and elastic modulus of the glass). Then, define the thermal boundary conditions to simulate the thermal stress distribution caused by temperature changes in the glass during the molding process. Through finite element analysis, the thermal stress distribution cloud map of the bottleneck area is obtained, and the stress concentration area is clearly identified. For example, it is found that the peak value of the thermal stress at the transition between the bottleneck and the bottle body reaches 50MPa, while the thermal stress in other areas is generally below 30MPa. Based on the thermal stress distribution data of the bottleneck area, SolidWorks is used to layout the mold cooling channel. Taking a glass medicine bottle as an example, open the geometric model of the bottleneck area in SolidWorks and import the thermal stress distribution data. According to the location of the stress concentration area, design the layout of the cooling channel to ensure that the coolant can flow through the parts with higher thermal stress first. For example, a spiral cooling channel with a diameter of 5mm is designed at the transition between the bottleneck and the bottle body, and the channel spacing is 10mm. The inlet and outlet positions of the cooling channel are also optimized to ensure the flow efficiency of the coolant. According to the bottleneck-mold cooling channel mapping data, MATLAB is used to perform thermomechanical coupling gradient modeling. Taking a glass beverage bottle as an example, the cooling channel layout data and thermal stress distribution data are imported into MATLAB. Using the numerical analysis function of MATLAB, a thermomechanical coupling gradient model is established. The model considers the thermal conduction effect of the cooling channel and the thermal expansion characteristics of the glass material. By solving the heat conduction equation and the stress balance equation, the thermomechanical coupling gradient function of the bottleneck area is obtained. For example, the model predicts that near the cooling channel, the thermal stress gradually decays from 50MPa to 30MPa, and the decay gradient is 2MPa / mm. The function is saved as a MATLAB function file. ANSYS Mechanical is used to identify the transition curvature of the bottle shoulder area and invert the pressure field of the blow molding process. Taking a glass wine bottle as an example, the geometric model of the bottle shoulder area is imported into ANSYS Mechanical, and the boundary conditions of the blow molding process (such as blowing pressure, blowing temperature, etc.) are set. Through finite element analysis, the transition curvature field and pressure field distribution of the bottle shoulder area are obtained. For example, it is found that the maximum transition curvature of the bottle shoulder area is 0.05mm⁻¹, and the corresponding blow molding pressure gradient is 0.5MPa / mm. According to the transition curvature field of the bottle shoulder area and the blow molding pressure gradient parameters, MATLAB is used to perform surface rheological property constraint modeling. Taking a glass medicine bottle as an example, the transition curvature field and pressure gradient parameters are imported into MATLAB. Using the symbolic calculation function of MATLAB, the rheological stress transfer function of the bottle shoulder area is established. This function takes into account the flow characteristics and pressure transfer effect of the glass during the blow molding process, and obtains the rheological stress distribution of the bottle shoulder area by solving the rheological equation.For example, the model predicts that in the shoulder area, the rheological stress gradually increases from 1.2MPa at the blow molding port to 1.5MPa in the middle of the shoulder. ANSYS Mechanical is used to decompose the stacking load direction vector field in the bottle area. Taking a glass beverage bottle as an example, the geometric model of the bottle area is imported into ANSYS Mechanical, and the boundary conditions of the stacking load (such as stacking pressure, contact conditions, etc.) are set. Through finite element analysis, the load direction vector field distribution in the bottle area is obtained. For example, it is found that the load direction in the bottle area is mainly along the axial direction of the bottle body, but there is a certain lateral component at the transition between the bottle body and the bottom of the bottle. Use SolidWorks to locate the mold parting line in the bottle area. Taking a glass medicine bottle as an example, the geometric model of the bottle area is opened in SolidWorks, and the load direction vector field data is imported. According to the load direction and mold design requirements, the parting line is designed in the bottle area. For example, the parting line is arranged along the axial direction of the bottle body, but it is optimized at the transition between the bottle body and the bottleneck to avoid stress concentration. According to the load direction vector diagram of the bottle body area and the spatial topological data of the bottle body-mold parting line, MATLAB is used to perform wall thickness gradient constraint modeling. Taking a glass wine bottle as an example, the load direction vector diagram and parting line data are imported into MATLAB. Using the numerical analysis function of MATLAB, a wall thickness gradient constraint function is established. This function considers the influence of the load direction and the parting line position on the wall thickness distribution, and obtains the wall thickness gradient constraint function of the bottle body area by solving the wall thickness distribution equation. For example, the model predicts that in the bottle body area, the wall thickness gradually increases from 2mm at the bottleneck to 3mm at the bottom of the bottle, with a gradient of 0.1mm / mm. ANSYS Mechanical is used to identify the contact stress of the mold ejector pin distribution in the bottle bottom area. Taking a glass beverage bottle as an example, the geometric model of the bottle bottom area is imported into ANSYS Mechanical, and the distribution and contact conditions of the mold ejector pin are set. The contact stress distribution in the bottle bottom area is obtained through finite element analysis. For example, it is found that the maximum contact stress in the ejector pin contact area is 80MPa, while the contact stress in other areas is generally below 50MPa. According to the ejector pin distribution constraint data in the bottle bottom area, MATLAB is used to perform stress attenuation gradient modeling. Taking a glass medicine bottle as an example, the ejector pin distribution constraint data is imported into MATLAB. Using the numerical analysis function of MATLAB, a contact stress attenuation function of the bottom area of ​​the bottle is established. This function takes into account the ejector pin distribution and the attenuation characteristics of the contact stress. By solving the stress attenuation equation, the contact stress attenuation function of the bottom area of ​​the bottle is obtained. For example, the model predicts that in the ejector pin contact area, the contact stress gradually decays from 80MPa to 50MPa, and the attenuation gradient is 3MPa / mm. Through the combination of these functions, a complete set of glass container modeling functions is constructed.

[0016] The present invention provides high-quality basic data for subsequent modeling by collecting the surface topography of common glass containers and generating a point cloud data set, and then filtering out noise points and smoothing the data, ensuring the accuracy and reliability of the model. Through NURBS surface fitting and reconstruction, a continuous and smooth surface model of the glass container can be generated. By partitioning the surface model into structures and extracting the geometric feature parameters of each region, the details of the model are further refined, making the subsequent parametric modeling more targeted and adaptable. The constructed parametric modeling function set covers key features such as the bottle body contour, wall thickness distribution, and transition regions, and can flexibly adjust the geometric shape and structural characteristics of the glass container, providing a powerful tool for generating diverse geometric variants. Through multi-condition simulation and stress distribution feature extraction of the geometric variant library constructed based on the parametric modeling function set, a container stress distribution feature set is generated, which not only provides rich data support for optimization design but also comprehensively evaluates the mechanical performance of different design schemes in actual use.

[0017] Preferably, step S17 includes the following steps: Step S171: Define variables for the glass container modeling function set and construct the dependency constraint relationships between geometric parameters to obtain the geometric parameter association map of the glass container; Step S172: Based on the geometric parameter association map of the glass container, rank the importance of the parameters in each region of the geometric feature parameter set of the glass container to obtain the key parameter ranking table of the glass container; Step S173: Determine the core parameter control set of the glass container according to the key parameter ranking table of the glass container; Step S174: Obtain the value range of each parameter in the core parameter control set of the glass container and generate a parameter value range table; Step S175: Based on the parameter value range table, construct design variants for each parameter in the core parameter control set of the glass container to obtain the parameter change sequence of the glass container; Step S176: Construct a geometric variant library of the glass container based on the parameter change sequence of the glass container, where the geometric variant library of the glass container contains no less than 100 design variants; Step S177: Conduct multi-condition simulation on the geometric variant library of the glass container and extract stress distribution features to obtain a container stress distribution feature set.

[0018] In this embodiment, when defining variables for the glass container modeling function set, the parametric modeling tool Grasshopper is used. Taking a glass beverage bottle as an example, first, variables in the Bezier curve function that controls the bottle body contour are defined, including the positions of the control points, weights, and the order of the curve. Next, variables in the radial thickness function that describes the wall thickness distribution are defined, such as the specific values of the wall thickness of the bottle body, bottleneck, and bottle bottom. In addition, variables in the spline interpolation function for the transition region are defined, including the starting point, ending point, and smoothness of the transition curve of the transition region. Through the "parametric component" function of Grasshopper, the dependency constraint relationships between geometric parameters are further constructed. For example, when the wall thickness in the bottleneck region increases, the wall thickness in the bottle body region will decrease accordingly to keep the overall mass unchanged. Finally, through the visualization function of Grasshopper, a geometric parameter correlation map of the glass container is generated, clearly showing the mutual relationships between the parameters. Based on the geometric parameter correlation map of the glass container, the MATLAB software is used to rank the importance of the parameters in each region of the glass container geometric feature parameter set. Taking a glass wine bottle as an example, first, the geometric parameter correlation map is imported into MATLAB, and its data processing function is used to perform a sensitivity analysis on the parameters in each region. By calculating the influence degree of each parameter on the overall model quality, strength, and manufacturing feasibility, it is determined that the wall thickness parameter in the bottleneck region, the height parameter in the bottle body region, and the curvature parameter in the bottle bottom region are the key parameters. Then, according to the importance degree of these parameters, a key parameter ranking table of the glass container is generated. For example, the importance of the bottleneck wall thickness is ranked first because it has the greatest impact on the overall strength and stress distribution of the glass container. According to the key parameter ranking table of the glass container, the core parameter control set of the glass container is determined. Taking a glass medicine bottle as an example, referring to the key parameter ranking table, the three most important parameters, namely the bottleneck wall thickness, bottle body height, and bottle bottom curvature, are selected as the core parameter control set. These parameters have the greatest impact on the structural performance and manufacturing feasibility of the medicine bottle, so they are preferentially considered for the construction of subsequent design variants. The names of these core parameters and their corresponding ranking positions are recorded in an Excel table. To obtain the value ranges of each parameter in the glass container core parameter control set, industry standards and actual production data are referred to. Taking a glass beverage bottle as an example, by consulting the standard specifications of the glass container manufacturing industry, the value range of the bottleneck wall thickness is determined to be from 1.0 mm to 2.0 mm, the value range of the bottle body height is from 150 mm to 200 mm, and the value range of the bottle bottom curvature is from 10 mm to 20 mm. These value ranges are comprehensively determined based on the physical properties of the glass material, the limitations of the forming process, and the actual usage requirements of the product. These value ranges are sorted into a parameter value range table and stored in an Excel file. According to the parameter value range table, Grasshopper is used to construct design variants for each parameter in the glass container core parameter control set.Taking a glass wine bottle as an example, first, set the value range of the bottleneck wall thickness from 1.2 mm to 1.8 mm with a step size of 0.2 mm in Grasshopper; the value range of the bottle body height from 180 mm to 220 mm with a step size of 10 mm; and the value range of the bottle bottom curvature from 12 mm to 18 mm with a step size of 2 mm. Then, through the "parametric drive" function of Grasshopper, automatically generate design variants under all potential parameter combinations. For example, when the bottleneck wall thickness is 1.2 mm, the bottle body height is 180 mm, and the bottle bottom curvature is 12 mm, generate a design variant; when the bottleneck wall thickness is 1.4 mm, the bottle body height is 190 mm, and the bottle bottom curvature is 14 mm, generate another design variant. Finally, obtain a sequence of design variants containing multiple parameter combinations. Based on the glass container parameter change sequence, use Rhinoceros and Grasshopper software to construct a glass container geometric variant library. Taking a glass medicine bottle as an example, import the parameter change sequence into Grasshopper, and use its powerful parametric modeling function to generate corresponding three-dimensional geometric models according to each parameter combination. By adjusting the variable values in the parametric modeling function set, generate 150 different geometric variants of glass medicine bottles. These variants differ in terms of bottleneck wall thickness, bottle body height, and bottle bottom curvature, covering different design possibilities. Export these geometric variants to a standard three-dimensional model file format and store them in a dedicated folder to form a glass container geometric variant library containing no less than 100 design variants. When performing multi-condition simulations on the glass container geometric variant library, use the finite element analysis software ANSYS Mechanical. Taking a glass beverage bottle as an example, first import each model in the geometric variant library into ANSYS Mechanical and configure multiple working conditions according to the actual use scenarios of the beverage bottle. For example, in the vertical top compression condition, apply a load of 150 N and constrain the bottle bottom; in the lateral extrusion condition, apply a load of 80 N and constrain the bottleneck; in the internal pressure condition, apply an internal pressure of 0.6 MPa; in the thermal shock condition, the temperature change range is from -10°C to 50°C; in the vibration transportation condition, the acceleration is 1.5 g and the frequency is 12 Hz; in the drop impact condition, the drop height is 1.2 m. Perform finite element solutions for each geometric variant separately to obtain stress and strain data under multiple working conditions. Through the post-processing function of ANSYS Mechanical, extract the maximum principal stress point, maximum shear stress point, and the region with the largest stress gradient of the beverage bottle under each working condition, and organize these data into a container stress distribution characteristic set.

[0019] Through constructing a correlation map of geometric parameters, the present invention clarifies the dependency relationships among various parameters, provides a clear logical framework for subsequent optimization design, and avoids blindness in the process of parameter adjustment. By determining the core parameter control set based on the key parameter ranking table, it further focuses on the parameters that have the greatest impact on the design, improving the pertinence and efficiency of design optimization. By generating a parameter value range table and constructing design variants, it can systematically explore design solutions under different parameter combinations, greatly enriching the design space and providing more possibilities for finding the optimal design solution. By constructing a geometric variant library containing no less than 100 design variants and conducting multi-condition simulations and stress distribution feature extraction, it not only provides comprehensive performance evaluation data for design optimization, but also can deeply analyze the mechanical performance of different design solutions based on the stress distribution feature vector set, thereby achieving precise optimization.

[0020] Preferably, step S2 includes the following steps: Step S21: Collect the glass forming process constraint parameter set; normalize the glass forming process constraint parameter set to obtain a standard manufacturing process parameter library; Step S22: Identify the glass temperature-viscosity relationship according to the standard manufacturing process parameter library to obtain glass temperature-viscosity relationship data; Step S23: Evaluate the material flow characteristics based on the glass temperature-viscosity relationship data to obtain the glass forming fluidity constraint conditions; Step S24: Identify the demolding constraint conditions for the production mold of typical glass containers to obtain the mold demolding constraint conditions; Step S25: Obtain the glass physical property data; identify the thermal stress distribution law according to the glass physical property data to obtain the glass thermal gradient constraint conditions; Step S26: Integrate the glass forming fluidity constraint conditions, the mold demolding constraint conditions and the glass thermal gradient constraint conditions to obtain the container forming process constraint set; Step S27: Identify the NURBS control points of the glass container according to the container forming process constraint set to obtain the constrained container NURBS control point set; Step S28: Conduct gradient-conserving deformation regulation on the constrained container NURBS control point set to obtain the container wall thickness gradient constraint model; generate a container stress-sensitive area distribution map based on the container wall thickness gradient constraint model.

[0021] In this embodiment, 5 glass container manufacturers can be contacted, including Company A, Company B, Company C, Company D, and Company E. Detailed production process data are collected from each company, including key parameters such as glass melting temperature, forming speed, mold temperature, and blowing pressure. These data are provided in the form of an Excel table, covering the production processes of different types of glass containers (such as wine bottles, beverage bottles, and medicine bottles). Python scripts are used to normalize these data, scaling all parameter values to the range of 0 to 1 to eliminate the influence of different dimensions and numerical ranges. Finally, a standard manufacturing process parameter library is generated. Based on the standard manufacturing process parameter library, MATLAB software is used to identify the glass temperature-viscosity relationship. Taking the production process data provided by Company A as an example, the glass melting temperature and the corresponding viscosity data are extracted. In MATLAB, the curve fitting toolbox is used to fit these data, and the Arrhenius equation is selected as the fitting model. By adjusting the model parameters, the glass temperature-viscosity relationship curve is obtained, and the fitting results are exported as a data file. According to the glass temperature-viscosity relationship data, ANSYS Fluent software is used to evaluate the material flow characteristics during the glass forming process. Taking the glass wine bottle produced by Company B as an example, a three-dimensional model of glass forming is established in ANSYS Fluent, and the functional relationships between the glass melting temperature and viscosity varying with temperature are set. By simulating the flow process of the glass in the mold, the fluidity of the glass under different temperature and pressure conditions is evaluated. The simulation results show that at a forming temperature of 1100 °C, the fluidity of the glass is the best and it can uniformly fill the mold. According to the simulation results, the glass forming fluidity constraint conditions are determined, including the minimum flow velocity and the maximum viscosity range, and these conditions are recorded in the process constraint parameter library. Reverse engineering parameter extraction is carried out on the typical glass container production mold. Taking the glass beverage bottle mold produced by Company C as an example, the mold is scanned with high precision using a three-dimensional scanner to obtain the point cloud data of the mold. Geomagic Control software is used to process the point cloud data and extract the key parameters of the mold, including the size, shape, demolding slope, and draft angle of the mold. By analyzing these parameters, the mold demolding constraint conditions are determined. For example, the demolding slope should be greater than 2° to ensure that the glass container can be smoothly demolded from the mold. Glass physical property data are obtained from Company D, including the thermal expansion coefficient, elastic modulus, and thermal conductivity of the glass, etc. MATLAB software is used to analyze these data to identify the thermal stress distribution law of the glass at different temperatures. By calculating the temperature gradient during the glass forming process, the glass thermal gradient constraint conditions are determined. For example, during the glass forming process, the temperature gradient should be controlled to not exceed 10 °C per centimeter to avoid glass breakage caused by excessive thermal stress. The glass forming fluidity constraint conditions, the mold demolding constraint conditions, and the glass thermal gradient constraint conditions are integrated to generate a container forming process constraint set.Taking the glass medicine bottle produced by Company E as an example, the above three constraints are integrated into a unified process constraint model. These constraints are organized using Excel tables to ensure that each constraint has a clear parameter range and calculation formula. For example, the fluidity constraint requires that the viscosity of the glass be kept below 10^5Pa·s during the molding process, the demoulding constraint requires that the draft angle of the mold be greater than 3°, and the thermal gradient constraint requires that the temperature gradient does not exceed 8℃ / cm. According to the container molding process constraint set, the Rhinoceros software is used to identify the NURBS control points of the glass container. Taking the glass medicine bottle as an example, the three-dimensional model of the medicine bottle is first imported into Rhinoceros. Using the "NURBS control point extraction" function of Rhinoceros, the control point grid is evenly arranged on the surface of the medicine bottle. The control point density of the bottle body and bottom area is set to one control point per 10mm×10mm, and the control point density of the bottleneck and transition area is encrypted to one control point per 5mm×5mm. By adjusting the weight coefficient and node vector of the control point, a complete NURBS control point set is obtained. Then, according to the constraints in the process constraint set, these control points are constrained to ensure that the design of the medicine bottle meets the molding process requirements. Finally, the constrained container NURBS control point set is obtained. The gradient conservation deformation control of the constrained container NURBS control point set is performed to generate the container wall thickness gradient constraint model and stress sensitive area distribution map. Taking the glass beverage bottle as an example, the Grasshopper plug-in is used to perform deformation control on the NURBS control point set. First, the parameter sensitivity of the control points is evaluated to determine the degree of influence of each control point on the wall thickness and stress distribution. According to the glass molding fluidity constraints and thermal gradient constraints, the control point motion chain transfer rules are constructed. Dynamic damping constraints are set in Grasshopper to ensure that the wall thickness gradient and stress distribution meet the process requirements during the deformation process. Through gradient conservation deformation control, the wall thickness gradient constraint model is obtained, and the stress sensitive area distribution map is generated based on the model. The distribution map clearly shows the stress concentration of the beverage bottle in different areas.

[0022] The present invention collects and normalizes the production process data of multiple manufacturers to build a standard manufacturing process parameter library. Through a comprehensive analysis of the glass temperature-viscosity relationship, material flow characteristics, mold demolding conditions, and thermal stress distribution laws, a molding process constraint set covering glass molding flow, mold demolding, and thermal gradient constraints is generated, which fully covers the key process requirements in the glass container manufacturing process. Further, through NURBS control point identification and gradient conservation deformation regulation, a wall thickness gradient constraint model and a stress sensitive area distribution map are established, achieving precise control of the geometric features of the glass container, ensuring that the design scheme can effectively adapt to the strict requirements of the actual manufacturing process while meeting the lightweight requirements, significantly reducing the deviation between the simulation design and the actual production, and improving the practicality and reliability of the design.

[0023] Preferably, step S27 includes the following steps: Step S271: constructing glass forming parameterized process boundary conditions based on the container forming process constraint set; Step S272: construct constraint violation penalty rules for the glass forming parameterized process boundary conditions to obtain process feasibility evaluation rules; Step S273: performing non-uniform rational B-spline control mesh division on the glass container mesh model to obtain a NURBS control mesh of the glass container, wherein the non-uniform rational B-spline control mesh division is specifically as follows: evenly arranging a control point grid on the surface of the glass container, with a control point density of one control point per 10 mm×10 mm area in the bottle body and bottle bottom regions, and a control point density of one control point per 5 mm×5 mm area in the bottleneck and transition regions; Step S274: setting weight coefficients and node vectors for the glass container NURBS control grid to obtain NURBS control points of the complete glass container; Step S275: Constraints are set on the NURBS control points of the complete glass container based on the process feasibility evaluation rules to obtain a constrained container NURBS control point set.

[0024] In this embodiment, MATLAB software is used. Taking a glass wine bottle as an example, key process parameters are first extracted from the container molding process constraint set, including the glass molding temperature range (1050°C to 1200°C), the mold demoulding angle (minimum 3°), the wall thickness gradient (maximum 0.5mm / cm) and the thermal gradient (maximum 10°C / cm). In MATLAB, the upper and lower limits of these parameters are defined and parameterized. For example, the molding temperature is defined as a variable T_form, and its value range is 1050°C to 1200°C. Through the above steps, a parameterized process boundary condition model containing all key process parameters is constructed. Based on MATLAB software, the constraint violation penalty rule is constructed for the parameterized process boundary conditions of glass molding. Taking a glass beverage bottle as an example, the penalty function of each process parameter is first defined. For example, for the molding temperature T_form, when it exceeds the range of 1050°C to 1200°C, the penalty function will linearly increase the penalty value according to the size of the range exceeded. Specifically, if the temperature is lower than 1050°C, the penalty value is (1050-T_form)*10; if the temperature is higher than 1200°C, the penalty value is (T_form-1200)*10. Similarly, for the mold demolding angle, if the angle is less than 3°, the penalty value is (3-demolding angle)*5. Through these penalty rules, a complete process feasibility evaluation rule is obtained. The glass container mesh model is meshed with non-uniform rational B-spline (NURBS) control meshes using Rhinoceros software. Taking the glass medicine bottle as an example, the three-dimensional model of the medicine bottle is first imported into Rhinoceros. In the software, select the "Meshing" function, set the control point density of the bottle body and bottom area to one control point per 10mm×10mm, and the control point density of the bottleneck and transition area is encrypted to one control point per 5mm×5mm. Through the above steps, the control point grid is evenly arranged on the surface of the medicine bottle, and a non-uniform rational B-spline control grid is generated. This control grid can accurately capture the geometric features of the medicine bottle, especially in the bottleneck and transition areas, where the encrypted control points can better reflect the geometric changes in these areas. Use Rhinoceros software to set the weight coefficient and node vector of the glass container NURBS control grid. Taking the glass wine bottle as an example, first select the "NURBS control point editing" function in Rhinoceros. For each control point, set the weight coefficient according to its importance in the geometric model. For example, the weight coefficient of the control point in the bottleneck area is set to 1.5 to ensure the geometric accuracy of the bottleneck; the weight coefficient of the control point in the bottle body area is set to 1.0. At the same time, the node vector is set according to the topological structure of the geometric model to ensure that the position of the control point in the geometric model is accurate. Through these settings, the complete glass container NURBS control points are obtained. Constraints are set for the NURBS control points of the complete glass container based on the process feasibility evaluation rules.Taking the glass beverage bottle as an example, the process feasibility assessment rules are first imported into the Rhinoceros software. In the software, the process feasibility assessment rules are applied to constrain the control points according to the weight coefficient and node vector of each control point. For example, for the molding temperature constraint, the temperature value of each control point must be set between 1050℃ and 1200℃. If the temperature value of a control point exceeds this range, the software will automatically adjust the position or weight of the control point to meet the process requirements. Through the above steps, the constrained container NURBS control point set is obtained.

[0025] By defining process boundary conditions and constraint violation penalty rules, the present invention can effectively evaluate the process feasibility of the design scheme, ensure that the design parameters are within the allowable range of the actual manufacturing process, and thus improve the manufacturability of the design. At the same time, the use of non-uniform rational B-spline control grid division, especially the encryption of control points in the bottleneck and transition areas, can more accurately capture the geometric characteristics and stress distribution of the glass container. By setting the weight coefficient and node vector of the NURBS control grid, and setting constraints based on the process feasibility evaluation rules, the final set of NURBS control points of the constraint container can effectively guide the shape adjustment and optimization of the glass container, ensuring that the design scheme achieves an optimal balance between lightweight and structural performance while meeting the process constraints.

[0026] Preferably, step S28 comprises the following steps: Step S281: evaluating the control point parameter sensitivity of the constraint container NURBS control point set to obtain control point parameter sensitivity distribution data; Step S282: constructing a displacement influence factor table based on the control point parameter sensitivity distribution data to generate a control point displacement influence factor table; Step S283: constructing a control point motion chain transfer rule according to glass forming fluidity constraints; Step S284: injecting dynamic damping constraints into the control point motion chain transfer rule according to the stress sensitive area distribution data to obtain a glass container deformation control strategy; Step S285: deforming and adjusting the NURBS control points of the complete glass container according to the glass container deformation control strategy to obtain a deformable mesh control point set; Step S286: constructing a wall thickness gradient restriction model according to the deformable grid control point set; Step S287: discretizing the wall thickness gradient restriction model to obtain a wall thickness gradient restriction discrete model; Step S288: using the wall thickness gradient constraint discrete model to perform gradient conservation deformation control on the deformable grid control point set to obtain a container wall thickness gradient constraint model; Step S289: Calculate the local stress response coefficient of the container wall thickness gradient constraint model according to the container stress distribution feature set to generate stress sensitive area distribution data; and generate a container stress sensitive area distribution map according to the stress sensitive area distribution data.

[0027] In this embodiment, taking a glass medicine bottle as an example, the constraint container NURBS control point set is first imported into MATLAB. Using the symbolic calculation function of MATLAB, the parameters of each control point (such as position, weight and node vector) are subjected to sensitivity analysis. Specifically, the influence degree of each control point parameter on the wall thickness and stress distribution of the glass container is calculated. For example, through differential analysis, it is found that the sensitivity of the control point position change in the bottleneck area to the wall thickness is 0.05mm / unit change, while the sensitivity in the bottle body area is 0.02mm / unit change. These sensitivity data are organized into control point parameter sensitivity distribution data and stored as a matrix file. Based on the control point parameter sensitivity distribution data, Excel is used to construct a control point displacement influence factor table. Taking a glass wine bottle as an example, the sensitivity data is imported into an Excel table. In the table, displacement influence factors are defined for each control point, which reflect the influence degree of the control point displacement on the overall geometry and stress distribution of the glass container. For example, the control point displacement influence factor in the bottleneck area is set to 1.2, indicating that the displacement in this area has a greater impact on the overall structure; while the displacement influence factor in the bottle bottom area is set to 0.8. Through the above steps, a detailed control point displacement influence factor table is generated. According to the fluidity constraints of glass forming, the control point motion chain transfer rules are constructed using ANSYS Mechanical software. Taking glass beverage bottles as an example, the flow characteristics of glass during the forming process are simulated in ANSYS Mechanical. According to the simulation results, the motion transfer rules between control points are defined to ensure that the fluidity of glass meets the process requirements during the deformation process. For example, when the control points in the bottleneck area are displaced, the control points in the adjacent bottle body area will be displaced in a certain proportion (such as 0.7) to maintain the continuity and uniformity of glass flow. These rules are integrated into the deformation analysis module of ANSYS Mechanical. According to the distribution data of stress-sensitive areas, ANSYS Mechanical is used to inject dynamic damping constraints into the control point motion chain transfer rules. Taking glass medicine bottles as an example, the stress-sensitive area distribution data is first imported into ANSYS Mechanical, which identifies the stress concentration of medicine bottles in different areas. According to the distribution of stress-sensitive areas, dynamic damping constraints are injected into the control point motion chain transfer rules. For example, in the stress-sensitive bottleneck area, a larger damping coefficient (such as 1.5) is set to slow down the movement speed of the control points and avoid excessive deformation in the stress concentration area. Through the above steps, the deformation control strategy of the glass container is obtained. According to the deformation control strategy of the glass container, the Rhinoceros software is used to adjust the deformation of the NURBS control points of the complete glass container. Taking the glass wine bottle as an example, the deformation control strategy is imported into Rhinoceros. In the software, the displacement of each control point is accurately adjusted according to the displacement influence factor and dynamic damping constraint of the control point.For example, for the control points in the bottleneck area, according to the deformation control strategy, it is moved inward by 0.5mm, and according to the kinematic chain transfer rule, the control points in the adjacent bottle body area are moved inward by 0.35mm. Through the above steps, a deformable mesh control point set is obtained. Based on the deformable mesh control point set, a wall thickness gradient restriction model is constructed using MATLAB. Taking the glass beverage bottle as an example, the deformable mesh control point set is first imported into MATLAB. Using the numerical analysis function of MATLAB, the wall thickness gradient corresponding to each control point is calculated, and the wall thickness gradient restriction condition is set according to the process requirements (such as the maximum wall thickness gradient does not exceed 0.5mm / cm). For example, for the control points in the bottleneck area, the calculated wall thickness gradient is 0.4mm / cm, which meets the process requirements; while for some control points in the bottle body area, the calculated wall thickness gradient is 0.6mm / cm, which exceeds the limit value, so it needs to be adjusted. Through the above steps, a wall thickness gradient restriction model is obtained. Taking the glass medicine bottle as an example, the wall thickness gradient restriction model is first imported into ANSYSMechanical. In the software, the geometric model of the medicine bottle is divided into finite element grids, and each grid unit corresponds to a wall thickness gradient value. Through discretization processing, a wall thickness gradient restricted discrete model is obtained. For example, the bottleneck area of ​​the medicine bottle is divided into 10 grid units, and the wall thickness gradient values ​​of each unit are 0.35mm / cm, 0.38mm / cm, etc. These discrete values ​​can accurately reflect the distribution of wall thickness gradients in different areas of the medicine bottle. Using the wall thickness gradient restricted discrete model, Rhinoceros and Grasshopper plug-ins are used to perform gradient conservation deformation control on the deformable grid control point set. Taking the glass wine bottle as an example, the wall thickness gradient restricted discrete model is first imported into Grasshopper. In Grasshopper, the corresponding control points are fine-tuned according to the wall thickness gradient value of each grid unit. For example, if the wall thickness gradient of a grid unit exceeds the limit value, the position of the control point corresponding to the unit will be appropriately adjusted to reduce the wall thickness gradient. Through the above steps, the container wall thickness gradient constraint model is obtained. According to the container stress distribution feature set, MATLAB is used to calculate the local stress response coefficient of the container wall thickness gradient constraint model. Taking a glass beverage bottle as an example, the stress distribution feature vector set is first imported into MATLAB. Using the numerical analysis function of MATLAB, the local stress response coefficient corresponding to each control point is calculated. These coefficients reflect the degree of influence of the change in the position of the control point on the local stress distribution. For example, the local stress response coefficient of the control point in the bottleneck area is 0.08, which means that a change of 0.1mm in the position of the control point will cause a local stress change of 0.008MPa. Based on these coefficients, the stress sensitive area distribution data is generated, and the stress sensitive area distribution map is generated using the visualization tool of MATLAB.

[0028] The present invention can clarify the degree of influence of each control point parameter on the overall structure by performing sensitivity evaluation on the NURBS control point set of the constraint container. Through the displacement influence factor table and control point motion chain transfer rule constructed based on sensitivity data, dynamic regulation of the deformation process of the glass container is further realized, ensuring that the structure is optimized and adjusted under the premise of meeting the requirements of glass molding fluidity and stress distribution. Through dynamic damping constraint injection and gradient conservation deformation regulation, a wall thickness gradient restriction model and stress sensitive area distribution map are constructed, which can effectively identify and optimize stress concentration areas, ensuring that the wall thickness distribution of the glass container is reasonable and the stress distribution is uniform during the lightweight design process, thereby significantly improving the structural strength and reliability of the glass container.

[0029] Preferably, step S3 comprises the following steps: Step S31: performing material density distribution on the glass container mesh model to obtain glass container volume element density data; Step S32: calculating the volume of the grid unit of the glass container grid model based on the glass container voxel density data to obtain a glass container unit mass distribution map; performing global integration on the glass container unit mass distribution map to obtain a total mass value of the glass container; Step S33: Modeling the relationship between mass and design parameters based on the glass container parameterized modeling function and the total mass value of the glass container to obtain a mass minimization objective function; Step S34: extracting stress peak points from the container stress distribution feature set to obtain a key stress evaluation point set for the glass container; performing Weibull distribution fitting on the key stress evaluation point set for the glass container to obtain failure probability distribution data for the glass container; Step S35: calculating the overall structural failure probability based on the failure probability distribution data of the glass container to obtain the overall structural failure probability of the glass container; Step S36: performing an inverse transformation on the failure probability of the entire structure of the glass container to obtain a strength maximization objective function; Step S37: Evaluate the manufacturing process complexity according to the process feasibility evaluation rules to obtain the glass container process difficulty index; Step S38: cost mapping the glass container process difficulty index to obtain glass container manufacturing cost prediction data; constructing a manufacturing feasibility function based on the glass container manufacturing cost prediction data and the glass container process difficulty index; Step S39: constructing a molding multi-objective optimization model based on the mass minimization objective function, the strength maximization objective function and the manufacturing feasibility function; solving the molding multi-objective optimization model to obtain a glass container design parameter set.

[0030] It is particularly important that step S39 further includes the following steps: Step S391: Perform multi-objective weight design on the mass minimization objective function, the strength maximization objective function and the manufacturing feasibility function to obtain an initial weight configuration scheme; Step S392: constructing a comprehensive performance evaluation function of a glass container based on the initial weight configuration scheme; Step S393: Acquire multi-objective optimization iteration historical data; Step S394: performing target conflict identification based on multi-target optimization iteration historical data to obtain glass container manufacturing target conflict data; Step S395: constructing a target weight adjustment strategy based on the glass container manufacturing target conflict data; Step S396: integrating the glass container comprehensive performance evaluation function and the target weight adjustment strategy to obtain a molding multi-objective optimization model; Step S397: Solve the molding multi-objective optimization model to obtain a glass container design parameter set.

[0031] In this embodiment, taking a glass wine bottle as an example, first import the mesh model of the wine bottle into ANSYS Mechanical. In the material property setting module, select the glass material and enter its density value (such as 2.5g / cm³). Then, assign the density value to each unit of the mesh model to ensure that each unit has the same material density. Through the above steps, the volume element density data of the glass container is obtained. Based on the volume element density data of the glass container, ANSYS Mechanical is used to calculate the volume of the mesh unit. Taking a glass beverage bottle as an example, select the "grid unit volume calculation" function in ANSYS Mechanical, the software automatically calculates the volume of each grid unit, and combines the volume element density data to obtain the mass of each unit. Use the post-processing function of ANSYS Mechanical to globally integrate the mass of all units to obtain the total mass value of the glass beverage bottle. For example, the calculation results show that the total mass of the beverage bottle is 300 grams. Through the above steps, the unit mass distribution diagram and total mass value of the glass container are obtained. Import the parametric modeling function and the total mass value of the glass container into MATLAB. Use the symbolic calculation function of MATLAB to establish a relationship model between mass and design parameters. For example, suppose the relationship between the mass M of a medicine bottle and the thickness of the neck, body, and bottom of the bottle is: ; in, , and are the volumes of the bottleneck, bottle body and bottle bottom respectively. MATLAB is used to extract the stress peak points of the container stress distribution feature set. Taking a glass wine bottle as an example, the stress distribution feature vector set is first imported into MATLAB. Using the data analysis function of MATLAB, the stress peak points are identified. For example, through analysis, it is found that the maximum principal stress in the bottleneck area is 50MPa and the maximum shear stress is 30MPa; the maximum principal stress in the bottle body area is 40MPa and the maximum shear stress is 20MPa. These stress peak points are extracted to form a set of key stress evaluation points for glass containers. Then, the curve fitting toolbox of MATLAB is used to fit the Weibull distribution to these key stress evaluation points to obtain the failure probability distribution data of the glass container. Through the above steps, the failure probability distribution of the glass container is obtained. Based on the failure probability distribution data of the glass container, MATLAB is used to calculate the failure probability of the overall structure. Taking a glass beverage bottle as an example, the failure probability distribution data is first imported into MATLAB. Using the numerical integration function of MATLAB, the failure probability of the entire structure is calculated. For example, assuming that the failure probability distribution function is: 1-; where is the stress, is the characteristic stress, and m is the Weibull modulus. Through numerical integration, the overall structural failure probability of the glass beverage bottle is 0.05%. Through the above steps, the overall structural failure probability of the glass container is obtained. According to the process feasibility evaluation rules, MATLAB is used to evaluate the complexity of the manufacturing process. Taking the glass wine bottle as an example, the process feasibility evaluation rules are imported into MATLAB. Using the data analysis function of MATLAB, the design parameters of the glass wine bottle (such as wall thickness gradient, mold demolding angle, etc.) are evaluated. For example, according to the process requirements, the wall thickness gradient does not exceed 0.5mm / cm, and the mold demolding angle is not less than 3°. Through the evaluation, the process difficulty index of the glass wine bottle is 0.7 (the full score is 1, and the lower the value, the greater the process difficulty). Through the above steps, the process difficulty index of the glass container is obtained. The process difficulty index is imported into MATLAB. Using the data analysis function of MATLAB, the process difficulty index is mapped to the manufacturing cost prediction data based on historical data and empirical formulas. For example, assume that the relationship between the manufacturing cost C and the process difficulty index I is C=a×I+b, where a and b are constants. Through cost mapping, the manufacturing cost prediction data of the glass beverage bottle is obtained. Based on the manufacturing cost prediction data and the process difficulty index, a manufacturing feasibility function is constructed. Use MATLAB to assign weights to the mass minimization objective function, the strength maximization objective function, and the manufacturing feasibility function. Taking the glass wine bottle as an example, the initial weight configuration scheme of the three objective functions is first defined in MATLAB. Assume that the weight of the mass minimization objective function is 0.4, the weight of the strength maximization objective function is 0.4, and the weight of the manufacturing feasibility function is 0.2. These weight values ​​are assigned according to the importance of the design goals and actual needs. Through the above steps, the initial weight configuration scheme is obtained.Based on the initial weight configuration scheme, MATLAB is used to construct a comprehensive performance evaluation function for glass containers. Taking glass beverage bottles as an example, the initial weight configuration scheme is imported into MATLAB. Using the symbolic calculation function of MATLAB, the three objective functions are weighted and summed according to the weights to obtain a comprehensive performance evaluation function. For example, assuming that the comprehensive performance evaluation function F is: ;in, , and They are the objective function of minimizing mass, the objective function of maximizing strength, and the manufacturing feasibility function. Through the above steps, the comprehensive performance evaluation function of the glass container is obtained. To obtain the historical data of multi-objective optimization iterations, use MATLAB's multi-objective optimization toolbox (such as gamultiobj). Taking the glass medicine bottle as an example, run the multi-objective optimization function in MATLAB and record the objective function value and design parameter value of each iteration. For example, in the iterative process of the genetic algorithm, record the objective function value and corresponding wall thickness parameter of each generation of the population.

[0032] The present invention obtains an accurate mass minimization objective function by performing material density distribution and mass distribution calculation on the glass container grid model, thus providing a quantitative basis for lightweight design. By extracting stress peak points and analyzing failure probabilities based on the stress distribution feature vector set, a strength maximization objective function is further obtained, ensuring that the glass container has sufficient structural strength while being lightweight. Through manufacturing process complexity evaluation and cost mapping, a manufacturing feasibility function is constructed, which closely integrates the design with the actual manufacturing process, ensuring that the design solution has good manufacturability while meeting performance requirements. Finally, by solving the multi-objective optimization model, the optimized glass container design parameter set is obtained, achieving the best balance between quality, strength and manufacturing feasibility.

[0033] Preferably, step S4 comprises the following steps: Step S41: constructing a three-dimensional detailed model of the glass container based on the glass container design parameter set; performing a CAD model integrity check on the three-dimensional detailed model of the glass container to obtain a model quality assessment report; Step S42: performing surface repair on the three-dimensional detailed model of the glass container according to the model quality assessment report to obtain a repaired three-dimensional model of the glass container; Step S43: performing tetrahedral and hexahedral mixed mesh division on the repaired glass container three-dimensional model to obtain a divided glass container three-dimensional model; performing material parameter configuration on the divided glass container three-dimensional model to obtain a parameterized glass container three-dimensional model; Step S44: simulating the filling process of the parameterized glass container three-dimensional model to obtain fluid dynamic load distribution data; Step S45: performing structural force response analysis on the parameterized glass container three-dimensional model according to the fluid dynamic load distribution data to obtain a coupled stress distribution field; performing stacking load simulation on the parameterized glass container three-dimensional model to obtain static bearing capacity data; Step S46: performing temperature gradient thermal stress calculation on the parameterized glass container three-dimensional model to obtain thermal stability evaluation data; performing internal pressure bearing capacity calculation on the parameterized glass container three-dimensional model to obtain internal pressure resistance performance data; Step S47: performing modal decomposition according to the fluid dynamic load distribution data and the coupled stress distribution field to obtain dynamic characteristic data of the glass container; Step S48: performing a drop impact dynamic simulation and evaluation on the parameterized glass container three-dimensional model to obtain impact resistance performance data; recording the static load-bearing capacity data, thermal stability evaluation data, internal pressure resistance performance data and impact resistance performance data as molding simulation performance data; Step S49: performing parameter perturbation analysis on the glass container design parameter set according to the dynamic characteristic data of the glass container to obtain a robustness index evaluation result.

[0034] It is particularly important that step S49 further includes the following steps: Step S491: identifying the main geometric parameters of the glass container design parameter set according to the dynamic characteristic data of the glass container to obtain a key geometric parameter set; Step S492: performing Monte Carlo random perturbation on the glass container design parameter set according to the key geometric parameter set to obtain a glass container parameter perturbation sample set; Step S493: constructing a glass container parameter perturbation three-dimensional model set based on the glass container parameter perturbation sample set; Step S494: quantify the statistical robustness of the glass container parameter perturbation three-dimensional model set to obtain a robustness index evaluation result.

[0035] In this embodiment, Rhinoceros software is used to build a three-dimensional detailed model based on the glass container design parameter set. Taking a glass wine bottle as an example, an accurate three-dimensional model is created in Rhinoceros according to design parameters (such as bottleneck wall thickness 1.2mm, bottle height 180mm, bottle bottom curvature 15mm). After completion, SolidWorks is used to check the CAD model integrity of the three-dimensional detailed model. The inspection content includes geometric continuity, surface gaps and topological structure. The inspection results show that there are a small number of surface gap problems in the model, and a model quality assessment report is generated based on this. According to the model quality assessment report, GeomagicWrap software is used to repair the surface of the three-dimensional detailed model of the glass container. Taking a glass beverage bottle as an example, after importing the model, the "surface repair" tool of GeomagicWrap is used to automatically fill the detected surface gaps and smooth the surface connections. After the repair is completed, the repaired three-dimensional model of the glass container is exported. ANSYSMeshing is used to perform tetrahedral and hexahedral mixed meshing on the repaired three-dimensional model of the glass container. Taking a glass medicine bottle as an example, the meshing parameters are set in ANSYSMeshing. Hexahedral meshes are selected in the bottle body and bottom area to improve the calculation accuracy, while tetrahedral meshes are used in the bottleneck and transition area to adapt to complex geometric shapes. After the meshing is completed, the material parameters of the meshed model are configured using ANSYSMaterialDatabase. The elastic modulus of the glass is set to 70GPa, the Poisson's ratio is set to 0.2, and other parameters are set to obtain a parameterized glass container 3D model. ANSYSFluent is used to perform fluid dynamics simulation on the parameterized glass container 3D model. Taking a glass wine bottle as an example, the fluid properties (such as water density and viscosity) are set in ANSYSFluent, and the boundary conditions of the filling process (such as inlet flow and outlet pressure) are defined. Through simulation calculations, the flow state and load distribution data of the fluid in the bottle are obtained, including the pressure distribution and velocity field of the fluid on the bottle wall. According to the fluid dynamic load distribution data, ANSYSMechanical is used to perform structural force response analysis on the parameterized glass container 3D model. Taking glass beverage bottles as an example, the fluid dynamic load data is imported into ANSYS Mechanical, and the material properties and boundary conditions (such as the fixed constraint of the bottle bottom) are set. Through finite element analysis, the coupled stress distribution field is obtained to show the stress concentration in the bottle body and bottleneck area. At the same time, the stacking load simulation is performed in ANSYS Mechanical, the stacking pressure is set to 500N, and the static bearing capacity data is calculated to ensure the safety of the glass containers when stacked. ANSYS Mechanical is used to calculate the temperature gradient thermal stress of the parameterized glass container 3D model.Taking a glass medicine bottle as an example, a temperature gradient boundary condition is set in ANSYS Mechanical (such as a linear change in the bottle body temperature from 20°C to 60°C), and material properties such as the thermal expansion coefficient and elastic modulus of the glass are considered. Through simulation calculation, thermal stability evaluation data, including thermal stress distribution and maximum thermal stress value, are obtained. At the same time, the internal pressure bearing capacity is calculated, and the internal pressure is set to 0.6MPa to obtain the internal pressure resistance performance data. According to the fluid dynamic load distribution data and the coupled stress distribution field, MATLAB is used for modal decomposition. Taking a glass wine bottle as an example, the fluid dynamic load and stress distribution data obtained by simulation are imported into MATLAB, and the modal decomposition is performed using MATLAB's signal processing toolbox. The dynamic characteristics data of the glass container are obtained through analysis, including the modal frequencies and vibration shapes of each order. ANSYS Mechanical is used to perform a drop impact dynamic simulation on the parameterized three-dimensional model of the glass container. Taking a glass beverage bottle as an example, the drop height is set to 1.2m in ANSYS Mechanical, the ground is a rigid plane, and the bottom of the bottle is the initial contact point. Through dynamic simulation calculation, the impact resistance data is obtained, including the stress peak, deformation and energy absorption during the impact process. The static load-bearing capacity data, thermal stability evaluation data, internal pressure resistance data and impact resistance data are integrated into the molding simulation performance data. The dynamic characteristics data of glass containers are analyzed using MATLAB. Taking glass wine bottles as an example, the dynamic characteristics data are imported into MATLAB, and the geometric parameters that have a greater impact on the dynamic characteristics, such as bottleneck length, bottle body diameter and bottle bottom thickness, are identified using MATLAB's data analysis function. These parameters are classified as a set of key geometric parameters. According to the key geometric parameter set, MATLAB is used to perform Monte Carlo random perturbation on the glass container design parameter set. Taking glass beverage bottles as an example, the perturbation range (such as ±5%) and perturbation number (such as 1000 times) of key geometric parameters (such as bottleneck length and bottle body diameter) are set in MATLAB. The random perturbation sample set is generated by the Monte Carlo method, and each sample represents a set of design parameters after random changes. Based on the glass container parameter perturbation sample set, Rhinoceros is used to construct a three-dimensional model set of glass container parameter perturbation. Taking a glass medicine bottle as an example, the perturbation sample set is imported into Rhinoceros, and its parametric modeling function is used to generate a corresponding three-dimensional model according to each set of perturbation parameters. Finally, a model set containing multiple perturbation models is obtained. MATLAB is used to analyze the three-dimensional model set of glass container parameter perturbations. Taking a glass wine bottle as an example, the perturbation model set is imported into MATLAB, and the performance indicators of each model (such as stress peak and deformation) are quantitatively analyzed using MATLAB's statistical analysis tools. By calculating the statistical quantities such as the mean, standard deviation and coefficient of variation of the performance indicators, the robustness index evaluation results are obtained.For example, if the standard deviation of the stress peak is small, it indicates that the design is more robust to parameter perturbations.

[0036] The present invention ensures the high quality and accuracy of the model by performing CAD integrity check and surface repair on the three-dimensional detailed model. By adopting tetrahedron and hexahedron mixed grid division and configuring material parameters, an accurate parameterized model is generated, which improves the accuracy and efficiency of simulation analysis. Through a series of multi-condition simulations such as filling process simulation, structural force response analysis, stacking load simulation, thermal stress calculation, internal pressure bearing capacity calculation and drop impact simulation, comprehensive performance data of glass containers in different usage scenarios are obtained, including fluid dynamic load distribution, coupled stress distribution field, static bearing capacity, thermal stability, internal pressure resistance and impact resistance. By performing parameter perturbation analysis based on dynamic characteristic data, the robustness index evaluation result is obtained, which further verifies the stability of the design scheme in the face of manufacturing errors and changes in the use environment.

[0037] Preferably, step S5 comprises the following steps: Step S51: extracting statistical indicators from the forming simulation performance data to obtain a forming performance indicator set; Step S52: Perform reliability evaluation on the robustness index evaluation result to obtain a design reliability analysis report; Step S53: performing a comprehensive performance score on the glass container according to the molding performance index set and the design reliability analysis report to obtain a design scheme evaluation index; Step S54: obtaining an initial molding design target; quantifying the difference between the design scheme evaluation index and the initial molding design target to obtain glass container performance gap data; Step S55: performing Sobol global sensitivity calculation on the glass container design parameter set based on the molding simulation performance data to obtain a key parameter impact data table; Step S56: sorting the adjustment parameters according to the glass container performance gap data and the key parameter impact data table to obtain a parameter adjustment priority list; generating design feedback adjustment suggestions based on the parameter adjustment priority list; Step S57: Generate a lightweight glass container structure design solution based on the design feedback adjustment suggestions.

[0038] In this embodiment, MATLAB is used to extract statistical indicators from the molding simulation performance data. Taking a glass wine bottle as an example, the static load-bearing capacity data, thermal stability evaluation data, internal pressure resistance data, and impact resistance data obtained by simulation are imported into MATLAB. Using MATLAB's statistical analysis tools, statistical indicators such as the mean, standard deviation, maximum value, and minimum value of these performance data are calculated. For example, the mean of the static load-bearing capacity is 550N and the standard deviation is 20N; the maximum value of the impact resistance is 1200J and the minimum value is 1000J. These statistical indicators are organized into a table to form a molding performance indicator set. MATLAB and Minitab software are used. Taking a glass beverage bottle as an example, the robustness index evaluation results are imported into MATLAB, and a preliminary analysis is performed using MATLAB's reliability analysis toolbox. Minitab software is used to perform hypothesis testing to verify the statistical significance of the robustness index. For example, through hypothesis testing, it is found that the confidence interval of the robustness index is [0.85, 0.95], indicating that the design scheme has a high reliability. A design reliability analysis report is generated based on these analysis results, and the process and conclusions of the reliability evaluation are recorded in detail. According to the molding performance index set and the design reliability analysis report, Excel is used to perform a comprehensive performance score on the glass container. Taking the glass medicine bottle as an example, weights are assigned to each performance index and reliability index in Excel. For example, the weight of static load capacity is 0.25, the weight of thermal stability is 0.20, the weight of internal pressure resistance is 0.20, the weight of impact resistance is 0.20, and the weight of reliability is 0.15. Then, the score is calculated according to the actual value and target value of each indicator, and all scores are weighted and summed to obtain the design scheme evaluation index. For example, the comprehensive performance score of a design scheme is 85 points (out of 100 points), indicating that the overall performance of the design scheme is good. Using MATLAB and Excel. Taking the glass wine bottle as an example, the initial molding design goals are first defined in MATLAB, including the target static load capacity of 600N, the target thermal stability index of 0.9, the target internal pressure resistance of 1.2MPa, etc. The design scheme evaluation index is compared with these initial design goals to calculate the difference. For example, the static load-bearing capacity of a design is 550N, which is 50N different from the target value of 600N; the thermal stability index is 0.88, which is 0.02 different from the target value of 0.9. These differences are quantified as performance gap data and recorded in an Excel spreadsheet. Based on the molding simulation performance data, SALib (a Python-based global sensitivity analysis library) is used to perform Sobol global sensitivity calculations on the glass container design parameter set. Taking the glass beverage bottle as an example, first install and import the SALib library in Python. Then, the design parameters (such as bottleneck wall thickness, bottle height, bottle bottom curvature) and the corresponding performance data (such as static load-bearing capacity, impact resistance) are input into SALib.By running the Sobol sensitivity analysis, the key parameter impact data table is obtained, which shows that the bottleneck wall thickness has the most significant impact on the static load-bearing capacity, with a sensitivity index of 0.45; the bottle height has the second largest impact on thermal stability, with a sensitivity index of 0.30. According to the glass container performance gap data and the key parameter impact data table, Excel is used to sort the adjustment parameters. Taking the glass medicine bottle as an example, the performance gap data and the key parameter impact data are integrated in Excel. For example, for a static load-bearing capacity gap of 50N, the sensitivity index of the bottleneck wall thickness is 0.45, so the bottleneck wall thickness is listed as the primary adjustment parameter; for a thermal stability gap of 0.02, the sensitivity index of the bottle height is 0.30, so the bottle height is listed as the second priority adjustment parameter. Based on these rankings, a parameter adjustment priority list is generated, and design feedback adjustment suggestions are generated based on this list, such as suggesting to increase the bottleneck wall thickness by 0.1mm to improve the static load-bearing capacity. According to the design feedback adjustment suggestions, Rhinoceros and Grasshopper are used to adjust the structure of the parametric glass container 3D model. Taking the glass wine bottle as an example, we first open the parametric model in Rhinoceros. According to the adjustment suggestions, we use the Grasshopper plug-in to modify the bottleneck wall thickness parameter from 1.2mm to 1.3mm; and adjust the bottle height parameter from 180mm to 182mm. After the modification is completed, the 3D model is regenerated, and the necessary geometric checks and optimizations are performed. Finally, the lightweight glass container structure design is obtained.

[0039] The present invention can comprehensively quantify the performance and reliability level of the design solution by extracting statistical indicators and evaluating the reliability of the robustness index of the molding simulation performance data. Through the comprehensive score generated based on performance indicators and reliability analysis, combined with the quantitative analysis of the differences in the initial design goals, the improvement direction and performance improvement space of the design solution are further clarified. Through the Sobol global sensitivity analysis, the key parameters that have the greatest impact on performance are accurately identified, making the design adjustment more targeted and efficient. Finally, the three-dimensional model is structurally adjusted according to the feedback adjustment suggestions, which realizes the continuous optimization of the design solution and ensures that the glass container has excellent mechanical properties and reliability while meeting the lightweight goal.

[0040] Preferably, step S57 includes the following steps: Step S571: updating the constraint conditions of the molding multi-objective optimization model according to the design feedback adjustment suggestions to obtain an optimization boundary correction scheme; Step S572: verifying the mathematical feasibility of the optimization boundary correction scheme to obtain an optimization constraint condition update set; constructing a constraint iteration optimization strategy based on the optimization constraint condition update set; Step S573: recording the glass container design parameter set, molding simulation performance data, design feedback adjustment suggestions and constraint iteration optimization strategy as glass container design case feature data; Step S574: extracting design rules and experience knowledge from the glass container design case feature data to obtain a glass container design knowledge entry set; Step S575: semantically annotating and building relationships for the glass container design knowledge item set to obtain knowledge graph construction data; Step S576: constructing a glass container design knowledge base based on the knowledge graph construction data; performing knowledge verification and evaluation on the glass container design knowledge base to obtain a knowledge base quality evaluation report; Step S577: supplementing and optimizing the glass container design knowledge base according to the knowledge base quality evaluation report to obtain a perfect glass container design knowledge base; Step S578: performing knowledge-driven evaluation on the parameterized glass container three-dimensional model by improving the glass container design knowledge base to obtain a design rationality verification result; Step S579: Make final adjustments to the parameterized glass container three-dimensional model based on the design rationality verification results and the comprehensive performance score to obtain a lightweight glass container structure design solution.

[0041] In this embodiment, according to the design feedback adjustment suggestion, MATLAB is used to update the constraint conditions of the molding multi-objective optimization model. Taking a glass wine bottle as an example, an existing multi-objective optimization model is opened in MATLAB, and the constraint conditions of the model are adjusted according to the feedback suggestions (such as increasing the bottleneck wall thickness to improve the strength). For example, the minimum value of the bottleneck wall thickness is updated from 1.2mm to 1.3mm. By modifying the constraint conditions of the model, an optimization boundary correction scheme is obtained, which provides new boundary conditions for subsequent optimization calculations. In order to verify the mathematical feasibility of the optimization boundary correction scheme, the optimization toolbox of MATLAB is used for verification. Taking a glass beverage bottle as an example, the updated constraint conditions are imported into MATLAB, and the multi-objective optimization model is run. By checking the convergence and feasibility of the model, it is confirmed whether the new constraint conditions meet the requirements of mathematical optimization. For example, if the multi-objective optimization model can find a feasible solution under the new constraint conditions, it is considered that the constraint conditions are feasible. Based on these verification results, an optimization constraint condition update set is obtained, and a constraint iteration optimization strategy is constructed to provide strategy support for subsequent optimization iterations. Excel is used to structure the glass container design parameter set, molding simulation performance data, design feedback adjustment suggestions and constraint iteration optimization strategy. Taking a glass medicine bottle as an example, a worksheet was created in Excel to record design parameters (such as bottleneck wall thickness, bottle height), simulation performance data (such as static load-bearing capacity, impact resistance), feedback adjustment suggestions (such as increasing bottleneck wall thickness) and optimization strategies (such as iteration step and direction). Through the above steps, the characteristic data of glass container design cases are obtained, which provides a data basis for the subsequent design rules and empirical knowledge extraction. In order to extract design rules and empirical knowledge, MATLAB and Excel are used to analyze the characteristic data of glass container design cases. Taking a glass wine bottle as an example, cluster analysis is performed on the data in MATLAB to identify the key parameter combinations that affect performance. For example, it is found that the combination of bottleneck wall thickness and bottle height has a significant effect on strength. Then, these analysis results are sorted in Excel to form design rules (such as "when the bottleneck wall thickness increases by 0.1mm, the strength increases by 5%) and empirical knowledge entry sets, which provide a knowledge basis for the subsequent knowledge graph construction. Neo4j (a graph database) is used to semantically annotate and build relationships for the glass container design knowledge entry set. Taking glass beverage bottles as an example, create nodes and relationships in Neo4j, annotate design rules and empirical knowledge items as nodes, and define the relationships between them (such as causal relationships, association relationships). For example, create a node "increase in bottleneck wall thickness" and another node "increase in strength", and define the causal relationship between them. Through the above steps, the knowledge graph construction data is obtained, which provides structured knowledge data for the subsequent knowledge base construction. Based on the knowledge graph construction data, Neo4j and Python (combined with the py2neo library) are used to build a glass container design knowledge base.Taking glass medicine bottles as an example, the knowledge graph data is imported into Neo4j, and Python scripts are used for data verification and preliminary evaluation. By checking the completeness and accuracy of the knowledge graph, a knowledge base quality assessment report is generated. For example, the report points out that the relationship between some nodes needs to be further refined, or some knowledge items need to be supplemented with more details. Through these assessments, the quality and availability of the knowledge base are ensured. According to the knowledge base quality assessment report, Neo4j and Python are used to supplement and optimize the glass container design knowledge base. Taking glass wine bottles as an example, the knowledge base is supplemented and modified according to the problems pointed out in the assessment report. For example, missing knowledge items are added, the relationship between nodes is refined, or incorrect data is corrected. Through these operations, the improved glass container design knowledge base is obtained, which provides high-quality knowledge support for subsequent knowledge-driven evaluation. Through the improved glass container design knowledge base, Python (combined with py2neo library) is used to perform knowledge-driven evaluation on the parametric glass container 3D model. Taking glass beverage bottles as an example, scripts are written in Python to query the relevant knowledge items in the knowledge base and evaluate the 3D model based on this knowledge. For example, according to the rule in the knowledge base that "the bottleneck wall thickness should be greater than 1.3mm to ensure strength", evaluate whether the bottleneck wall thickness of the current model meets the requirements. Through the above steps, the design rationality verification results are obtained, which provides a basis for subsequent model adjustments. According to the design rationality verification results and the comprehensive performance score, Rhinoceros and Grasshopper are used to make final adjustments to the parametric glass container 3D model. Taking the glass medicine bottle as an example, according to the results of the knowledge-driven evaluation (such as the bottleneck wall thickness needs to be further increased) and the comprehensive performance score (such as the strength still needs to be improved), the 3D model is opened in Rhinoceros, and the relevant parameters are adjusted using the Grasshopper plug-in. For example, the bottleneck wall thickness is increased from 1.3mm to 1.4mm, and the model is regenerated. Through these adjustments, the final lightweight glass container structure design scheme is obtained, which achieves the lightweight goal while meeting the performance requirements.

[0042] The present invention ensures the scientificity and feasibility of the optimization process by adjusting the constraints of the optimization model according to the design feedback and performing mathematical feasibility verification. Furthermore, a glass container design knowledge base is constructed by structuring and extracting knowledge from the design case feature data, and semantic annotation and relationship construction of knowledge are performed based on knowledge graph technology, providing rich knowledge support for design optimization. This not only accumulates design experience, but also can verify the rationality of the design scheme through knowledge-driven evaluation and further optimize the design scheme. Finally, the three-dimensional model is adjusted in combination with the evaluation results of the knowledge base and the comprehensive performance score to ensure the scientificity, rationality and efficiency of the design scheme.

[0043] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0044] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A lightweight glass container structure optimization design method based on a three-dimensional simulation model, characterized in that: The following steps are involved: Step S1: constructing a glass container geometric variant library; performing multi-condition simulation on the glass container geometric variant library, and extracting stress distribution features to obtain a container stress distribution feature set; Step S2: performing NURBS control point identification on the glass container to obtain a NURBS control point set of the constrained container; performing gradient conservation deformation control on the NURBS control point set of the constrained container to obtain a container wall thickness gradient constraint model; and generating a container stress sensitive area distribution map based on the container wall thickness gradient constraint model; Step S3: constructing a mass minimization objective function, a strength maximization objective function, and a manufacturing feasibility function for the glass container; constructing a molding multi-objective optimization model based on the mass minimization objective function, the strength maximization objective function, and the manufacturing feasibility function; Solve the molding multi-objective optimization model to obtain the glass container design parameter set; Step S4: constructing a parameterized glass container three-dimensional model according to the glass container design parameter set; performing a molding performance evaluation on the parameterized glass container three-dimensional model to obtain molding simulation performance data; performing a parameter perturbation analysis on the glass container design parameter set to obtain a robustness index evaluation result; Step S5: generating design feedback adjustment suggestions based on the molding simulation performance data and the robustness index evaluation results; Generate a lightweight glass container structure design plan based on the design feedback and adjustment suggestions.

2. The lightweight glass container structure optimization design method based on a three-dimensional simulation model according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: obtaining a list of common glass containers; collecting surface morphology of each common glass container according to the list of common glass containers to obtain a glass container point cloud data set; Step S12: filtering out noise points on the glass container point cloud dataset to obtain a smooth glass container point cloud dataset; Step S13: performing NURBS surface fitting reconstruction based on the smoothed glass container point cloud data set to obtain a continuous surface model of the glass container; Step S14: Structural partitioning of the continuous curved surface model of the glass container is performed to obtain curved surface blocks of the bottleneck area, the bottle shoulder area, the bottle body area and the bottle bottom area; Step S15: extracting characteristic dimensions of the curved surface blocks of the bottleneck area, the bottle shoulder area, the bottle body area and the bottle bottom area to obtain a set of geometric characteristic parameters of the glass container; Step S16: constructing a glass container modeling function set according to the glass container geometric feature parameter set, wherein the glass container modeling function set includes a bottleneck area thermal-mechanical coupling gradient function, a shoulder area rheological stress transfer function, a body area wall thickness gradient constraint function, and a bottom area contact stress attenuation function. The specific construction process of the glass container modeling function set is as follows: Identify the thermal stress concentration area in the bottleneck area of ​​the glass container and obtain the thermal stress distribution data of the bottleneck area; Layout the mold cooling channel according to the thermal stress distribution data of the bottleneck area to obtain the bottleneck-mold cooling channel mapping data; Based on the bottleneck-mold cooling channel mapping data, thermal-mechanical coupling gradient modeling is performed to obtain the thermal-mechanical coupling gradient function of the bottleneck area; The transition curvature of the shoulder area of ​​the glass container is identified, and the pressure field of the blow molding process is inverted to obtain the transition curvature field of the shoulder area and the pressure gradient parameters of the shoulder blow molding; According to the transition curvature field of the bottle shoulder area and the pressure gradient parameters of the bottle shoulder blow molding, the curved surface rheological characteristics of the glass container are constrained, and the rheological stress transfer function of the bottle shoulder area is obtained; Decompose the stacking load direction vector field of the bottle body area of ​​the glass container to obtain the load direction vector diagram of the bottle body area; Locate the mold parting line of the bottle body area of ​​the glass container to obtain the bottle body-mold parting line space topology data; Based on the load direction vector diagram of the bottle body area and the spatial topological data of the bottle body-mold parting line, the wall thickness-parting line dynamic coupling modeling is carried out to obtain the wall thickness gradient constraint function of the bottle body area; Conduct mold ejector pin distribution contact stress identification on the bottom area of ​​the glass container to obtain ejector pin distribution constraint data in the bottom area; Stress attenuation gradient modeling is performed based on the ejector pin distribution constraint data in the bottom area of ​​the bottle to obtain the contact stress attenuation function in the bottom area of ​​the bottle; Step S17: constructing a glass container geometric variant library based on the glass container modeling function set; performing multi-condition simulation on the glass container geometric variant library, and extracting stress distribution features to obtain a container stress distribution feature set.

3. The lightweight glass container structure optimization design method based on a three-dimensional simulation model according to claim 2, characterized in that: Step S17 includes the following steps: Step S171: defining variables for the glass container modeling function set, and constructing dependency constraint relationships between geometric parameters, thereby obtaining a glass container geometric parameter association map; Step S172: sorting the importance of each regional parameter in the glass container geometric characteristic parameter set based on the glass container geometric parameter association map to obtain a glass container key parameter sorting table; Step S173: determining a glass container core parameter control set according to the glass container key parameter ranking table; Step S174: obtaining the value range of each parameter in the glass container core parameter control set, and generating a parameter value range table; Step S175: constructing a design variant for each parameter in the glass container core parameter control set according to the parameter value range table to obtain a glass container parameter variation sequence; Step S176: constructing a glass container geometric variant library based on the glass container parameter variation sequence, wherein the glass container geometric variant library contains no less than 100 design variants; Step S177: Perform multi-condition simulation on the glass container geometry variant library and extract stress distribution features to obtain a container stress distribution feature set.

4. The lightweight glass container structure optimization design method based on a three-dimensional simulation model according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: collecting a glass forming process constraint parameter set; normalizing the glass forming process constraint parameter set to obtain a standard manufacturing process parameter library; Step S22: identifying the glass temperature-viscosity relationship according to the standard manufacturing process parameter library to obtain glass temperature-viscosity relationship data; Step S23: evaluating the material flow characteristics based on the glass temperature-viscosity relationship data to obtain glass molding fluidity constraint conditions; Step S24: Identifying demoulding constraint conditions of a typical glass container production mold to obtain mold demoulding constraint conditions; Step S25: Acquire glass physical property data; identify the thermal stress distribution law based on the glass physical property data to obtain glass thermal gradient constraint conditions; Step S26: Integrate the glass molding fluidity constraint condition, the mold demoulding constraint condition and the glass thermal gradient constraint condition to obtain a container molding process constraint set; Step S27: performing NURBS control point identification on the glass container according to the container molding process constraint set to obtain a NURBS control point set of the constraint container; Step S28: performing gradient conservation deformation control on the NURBS control point set of the constraint container to obtain a container wall thickness gradient constraint model; and generating a container stress sensitive area distribution map based on the container wall thickness gradient constraint model.

5. The method for optimizing the structure of a lightweight glass container based on a three-dimensional simulation model according to claim 4, characterized in that: Step S27 includes the following steps: Step S271: constructing glass forming parameterized process boundary conditions based on the container forming process constraint set; Step S272: construct constraint violation penalty rules for the glass forming parameterized process boundary conditions to obtain process feasibility evaluation rules; Step S273: performing non-uniform rational B-spline control mesh division on the glass container mesh model to obtain a NURBS control mesh of the glass container, wherein the non-uniform rational B-spline control mesh division is specifically as follows: evenly arranging a control point grid on the surface of the glass container, with a control point density of one control point per 10 mm×10 mm area in the bottle body and bottle bottom regions, and a control point density of one control point per 5 mm×5 mm area in the bottleneck and transition regions; Step S274: setting weight coefficients and node vectors for the glass container NURBS control grid to obtain NURBS control points of the complete glass container; Step S275: Constraints are set on the NURBS control points of the complete glass container based on the process feasibility evaluation rules to obtain a constrained container NURBS control point set.

6. The method for optimizing the structure of a lightweight glass container based on a three-dimensional simulation model according to claim 4, characterized in that: Step S28 includes the following steps: Step S281: evaluating the control point parameter sensitivity of the constraint container NURBS control point set to obtain control point parameter sensitivity distribution data; Step S282: constructing a displacement influence factor table based on the control point parameter sensitivity distribution data to generate a control point displacement influence factor table; Step S283: constructing a control point motion chain transfer rule according to glass forming fluidity constraints; Step S284: injecting dynamic damping constraints into the control point motion chain transfer rule according to the stress sensitive area distribution data to obtain a glass container deformation control strategy; Step S285: deforming and adjusting the NURBS control points of the complete glass container according to the glass container deformation control strategy to obtain a deformable mesh control point set; Step S286: constructing a wall thickness gradient restriction model according to the deformable grid control point set; Step S287: discretizing the wall thickness gradient restriction model to obtain a wall thickness gradient restriction discrete model; Step S288: using the wall thickness gradient constraint discrete model to perform gradient conservation deformation control on the deformable grid control point set to obtain a container wall thickness gradient constraint model; Step S289: Calculate the local stress response coefficient of the container wall thickness gradient constraint model according to the container stress distribution feature set to generate stress sensitive area distribution data; and generate a container stress sensitive area distribution map according to the stress sensitive area distribution data.

7. The method for optimizing the structure of a lightweight glass container based on a three-dimensional simulation model according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: performing material density distribution on the glass container mesh model to obtain glass container volume element density data; Step S32: calculating the volume of the grid unit of the glass container grid model based on the glass container voxel density data to obtain a glass container unit mass distribution map; performing global integration on the glass container unit mass distribution map to obtain a total mass value of the glass container; Step S33: Modeling the relationship between mass and design parameters based on the glass container parameterized modeling function and the total mass value of the glass container to obtain a mass minimization objective function; Step S34: extracting stress peak points from the container stress distribution feature set to obtain a key stress evaluation point set for the glass container; performing Weibull distribution fitting on the key stress evaluation point set for the glass container to obtain failure probability distribution data for the glass container; Step S35: calculating the overall structural failure probability based on the failure probability distribution data of the glass container to obtain the overall structural failure probability of the glass container; Step S36: performing an inverse transformation on the failure probability of the entire structure of the glass container to obtain a strength maximization objective function; Step S37: Evaluate the manufacturing process complexity according to the process feasibility evaluation rules to obtain the glass container process difficulty index; Step S38: cost mapping the glass container process difficulty index to obtain glass container manufacturing cost prediction data; constructing a manufacturing feasibility function based on the glass container manufacturing cost prediction data and the glass container process difficulty index; Step S39: constructing a molding multi-objective optimization model based on the mass minimization objective function, the strength maximization objective function and the manufacturing feasibility function; solving the molding multi-objective optimization model to obtain a glass container design parameter set.

8. The method for optimizing the structure of a lightweight glass container based on a three-dimensional simulation model according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: constructing a three-dimensional detailed model of the glass container based on the glass container design parameter set; performing a CAD model integrity check on the three-dimensional detailed model of the glass container to obtain a model quality assessment report; Step S42: performing surface repair on the three-dimensional detailed model of the glass container according to the model quality assessment report to obtain a repaired three-dimensional model of the glass container; Step S43: performing tetrahedral and hexahedral mixed mesh division on the repaired glass container three-dimensional model to obtain a divided glass container three-dimensional model; performing material parameter configuration on the divided glass container three-dimensional model to obtain a parameterized glass container three-dimensional model; Step S44: simulating the filling process of the parameterized glass container three-dimensional model to obtain fluid dynamic load distribution data; Step S45: performing structural force response analysis on the parameterized glass container three-dimensional model according to the fluid dynamic load distribution data to obtain a coupled stress distribution field; performing stacking load simulation on the parameterized glass container three-dimensional model to obtain static bearing capacity data; Step S46: performing temperature gradient thermal stress calculation on the parameterized glass container three-dimensional model to obtain thermal stability evaluation data; performing internal pressure bearing capacity calculation on the parameterized glass container three-dimensional model to obtain internal pressure resistance performance data; Step S47: performing modal decomposition according to the fluid dynamic load distribution data and the coupled stress distribution field to obtain dynamic characteristic data of the glass container; Step S48: performing a drop impact dynamic simulation and evaluation on the parameterized glass container three-dimensional model to obtain impact resistance performance data; recording the static load-bearing capacity data, thermal stability evaluation data, internal pressure resistance performance data and impact resistance performance data as molding simulation performance data; Step S49: performing parameter perturbation analysis on the glass container design parameter set according to the dynamic characteristic data of the glass container to obtain a robustness index evaluation result.

9. The method for optimizing the structure of a lightweight glass container based on a three-dimensional simulation model according to claim 1, characterized in that: Step S5 includes the following steps: Step S51: extracting statistical indicators from the forming simulation performance data to obtain a forming performance indicator set; Step S52: Perform reliability evaluation on the robustness index evaluation result to obtain a design reliability analysis report; Step S53: performing a comprehensive performance score on the glass container according to the molding performance index set and the design reliability analysis report to obtain a design scheme evaluation index; Step S54: obtaining an initial molding design target; quantifying the difference between the design scheme evaluation index and the initial molding design target to obtain glass container performance gap data; Step S55: performing Sobol global sensitivity calculation on the glass container design parameter set based on the molding simulation performance data to obtain a key parameter impact data table; Step S56: sorting the adjustment parameters according to the glass container performance gap data and the key parameter impact data table to obtain a parameter adjustment priority list; and generating design feedback adjustment suggestions based on the parameter adjustment priority list; Step S57: Generate a lightweight glass container structure design solution based on the design feedback adjustment suggestions.

10. The method for optimizing the structure of a lightweight glass container based on a three-dimensional simulation model according to claim 9, characterized in that: Step S57 includes the following steps: Step S571: updating the constraint conditions of the molding multi-objective optimization model according to the design feedback adjustment suggestions to obtain an optimization boundary correction scheme; Step S572: verifying the mathematical feasibility of the optimization boundary correction scheme to obtain an optimization constraint condition update set; constructing a constraint iteration optimization strategy based on the optimization constraint condition update set; Step S573: recording the glass container design parameter set, molding simulation performance data, design feedback adjustment suggestions and constraint iteration optimization strategy as glass container design case feature data; Step S574: extracting design rules and experience knowledge from the glass container design case feature data to obtain a glass container design knowledge entry set; Step S575: semantically annotating and building relationships for the glass container design knowledge item set to obtain knowledge graph construction data; Step S576: constructing a glass container design knowledge base based on the knowledge graph construction data; performing knowledge verification and evaluation on the glass container design knowledge base to obtain a knowledge base quality evaluation report; Step S577: supplementing and optimizing the glass container design knowledge base according to the knowledge base quality evaluation report to obtain a perfect glass container design knowledge base; Step S578: performing knowledge-driven evaluation on the parameterized glass container three-dimensional model by improving the glass container design knowledge base to obtain a design rationality verification result; Step S579: Make final adjustments to the parameterized glass container three-dimensional model based on the design rationality verification results and the comprehensive performance score to obtain a lightweight glass container structure design solution.

Images