Product package online design display method and system

Abstract

The invention relates to the technical field of data processing, in particular to a product package online design display method and system. The method comprises the following steps of obtaining packaging target product data and determining the outer contour of a product, reconstructing a simulation to-be-packaged product frame, obtaining a packaging leakproofness requirement and estimating a packaging use scene, generating packaging process constraint data, analyzing entity packaging constraints, and fusing the packaging constraint data to support packaging process simulation. The method comprises the following steps: generating simulated product packaging data, constructing a packaging demand reverse environment field according to a leakproofness demand, carrying out leakproofness re-measurement on the simulated product packaging data, generating packaging design leakproofness data, carrying out packaging backtracking optimization through the data, finally obtaining qualified product packaging design data, and carrying out online visual display. According to the invention, efficient and intelligent product package online design and display are realized.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method and system for online design and display of product packaging. Background Art

[0002] Traditional packaging design methods often rely on experience and manual operations, resulting in low design efficiency and frequent errors. Existing technologies lack accuracy in obtaining product data and determining outer contours, and often cannot meet the packaging needs of complex products. Especially in a diversified market environment, the flexibility and adaptability of packaging design are insufficient, affecting the product's market competitiveness and consumer experience. In addition, traditional methods for evaluating packaging sealing requirements and estimating usage scenarios often rely on subjective judgment, lack systematic and scientific data support, and the mapping of process constraints is often inaccurate, resulting in the physical limitations in the packaging process not being fully considered. The safety and effectiveness of the packaging design are affected, and real-time monitoring and optimization of the sealing performance cannot be achieved. Summary of the Invention

[0003] Based on this, it is necessary to provide a product packaging online design and display method and system to solve at least one of the above technical problems.

[0004] To achieve the above-mentioned purpose, a method for online design and display of product packaging includes the following steps:

[0005] Step S1: Acquire data of a target product to be packaged; determine the outer contour of the product according to the data of the target product to be packaged; and reconstruct a simulated frame of the product to be packaged based on the outer contour of the product;

[0006] Step S2: Obtaining packaging tightness requirements; estimating packaging usage scenarios based on the packaging tightness requirements; performing process constraint mapping based on the estimated packaging usage scenarios and the packaging tightness requirements to generate packaging process constraint data;

[0007] Step S3: Analyze the physical packaging constraints based on the simulated product framework to be packaged; fuse the product packaging constraints according to the physical packaging constraints and the packaging process constraint data to obtain packaging constraint data;

[0008] Step S4: simulating the packaging process of the simulated product framework to be packaged based on the packaging constraint data to generate simulated product packaging data; and constructing a packaging requirement reverse environment field according to the packaging airtightness requirement;

[0009] Step S5: retest the sealing performance of the simulated product packaging data based on the reverse environmental field of the packaging requirements, thereby generating packaging design sealing performance data; perform packaging retrospective optimization on the simulated product packaging data through the packaging design sealing performance data, thereby obtaining qualified product packaging design data, and visually display the qualified product packaging design data online.

[0010] By acquiring the target product data for packaging and determining the outer contour of the product, the present invention can accurately reconstruct the simulated product framework to be packaged, providing a clear basis for subsequent packaging design. The acquisition of packaging airtightness requirements is combined with the estimated packaging usage scenario to formulate reasonable environmental adaptability for packaging design. The packaging process constraint data generated by process constraint mapping ensures that the packaging design meets the actual operation requirements. The analysis of physical packaging constraints helps to fully understand the physical limitations of the packaging process. The product packaging constraint fusion based on these constraints forms systematic packaging constraint data. The packaging process simulation of the simulated product framework to be packaged based on these data can display the dynamic changes in the packaging process in real time. The generated simulated product packaging data provides a reference for design improvement. The construction of the reverse environmental field of packaging requirements provides a scientific basis for the re-testing of the sealing performance. This environmental field re-tests the sealing performance of simulated product packaging data, which can accurately evaluate the sealing of packaging design. The generated packaging design sealing performance data provides data support for subsequent optimization. By performing packaging retrospective optimization on simulated product packaging data, the qualification of the final product packaging design is ensured. The online visual display of qualified product packaging design data not only improves the transparency of the design, but also facilitates all parties involved to monitor and evaluate the design effect in real time. Through the application of digital technology, the overall system not only improves the efficiency of packaging design, but also enhances the flexibility and adaptability of design, laying the foundation for the intelligent transformation of the modern packaging industry, promoting the digitalization and visualization of packaging design, promoting the rational use of resources and cost control, improving the market competitiveness and customer satisfaction of enterprises, and ensuring the efficient response and rapid iteration of packaging design under diversified needs.

[0011] The present invention also provides a product packaging online design and display system for executing the product packaging online design and display method described above. The product packaging online design and display system includes:

[0012] The product modeling module is used to obtain the packaging target product data; determine the product outline based on the packaging target product data; and reconstruct the simulated product frame to be packaged based on the product outline;

[0013] The requirements mapping module is used to obtain packaging tightness requirements; estimate packaging usage scenarios based on packaging tightness requirements; and perform process constraint mapping based on the estimated packaging usage scenarios and packaging tightness requirements to generate packaging process constraint data;

[0014] The constraint fusion module is used to analyze the physical packaging constraints based on the simulated product framework to be packaged; the product packaging constraints are integrated according to the physical packaging constraints and the packaging process constraint data to obtain the packaging constraint data;

[0015] The packaging simulation module is used to simulate the packaging process of the simulated product framework to be packaged based on the packaging constraint data to generate simulated product packaging data; and to construct a packaging requirement reverse environment field according to the packaging airtightness requirements;

[0016] The sealing optimization module is used to re-test the sealing performance of simulated product packaging data based on the reverse environmental field of packaging requirements, thereby generating packaging design sealing performance data; the simulated product packaging data is optimized through the packaging design sealing performance data to obtain qualified product packaging design data, and the qualified product packaging design data is visualized online.

[0017] Through the application of the product modeling module, the present invention enables the system to accurately obtain the packaging target product data and determine its outer contour. The reconstructed simulated product framework to be packaged provides a clear basis for subsequent design. The packaging airtightness requirements obtained by the demand mapping module are combined with the estimated packaging usage scenarios to formulate reasonable environmental adaptability for the packaging design. The generated packaging process constraint data ensures that the design meets the actual operation requirements. The constraint fusion module can fully understand the physical limitations of the packaging process by analyzing the physical packaging constraints. The product packaging constraint fusion based on these constraints forms systematic packaging constraint data. The packaging simulation module simulates the packaging process of the simulated product framework to be packaged based on the packaging constraint data, and can display the dynamic changes in the packaging process in real time. The generated simulated product packaging data provides an important reference for design improvement. The construction of the reverse environmental field of packaging requirements provides a scientific basis for the re-testing of the sealing performance. The re-testing of the sealing performance of the simulated product packaging data based on the environmental field can accurately The airtightness of the packaging design is evaluated, and the generated packaging design airtightness performance data provides data support for subsequent optimization. By performing packaging retrospective optimization on the simulated product packaging data, the qualification of the final product packaging design is ensured. The online visual display of qualified product packaging design data not only improves the transparency of the design, but also facilitates all parties involved to monitor and evaluate the design effect in real time. The overall system has significantly improved the efficiency of packaging design through the application of digital technology, enhanced the flexibility and adaptability of design, laid the foundation for the intelligent transformation of the modern packaging industry, promoted the digitalization and visualization process of packaging design, promoted the rational use of resources and cost control, improved the company's market competitiveness and customer satisfaction, and ensured the efficient response and rapid iteration of packaging design under diversified needs. The integrated design of the system improves the collaboration efficiency between modules, reduces errors in the design process, enhances the overall quality and reliability of product packaging, and ultimately realizes the intelligence, automation and efficiency of packaging design. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A flowchart of the steps in designing an online display method for a product packaging;

[0019] Figure 2 Detailed implementation flow chart of step S2;

[0020] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0021] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. It is obvious that the embodiments described are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts are within the scope of protection of the present invention.

[0022] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0023] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0024] To achieve this, please refer to Figures 1 to 2 , a product packaging online design and display method, comprising the following steps:

[0025] Step S1: Acquire data of a target product to be packaged; determine the outer contour of the product according to the data of the target product to be packaged; and reconstruct a simulated frame of the product to be packaged based on the outer contour of the product;

[0026] Step S2: Obtaining packaging tightness requirements; estimating packaging usage scenarios based on the packaging tightness requirements; performing process constraint mapping based on the estimated packaging usage scenarios and the packaging tightness requirements to generate packaging process constraint data;

[0027] Step S3: Analyze the physical packaging constraints based on the simulated product framework to be packaged; fuse the product packaging constraints according to the physical packaging constraints and the packaging process constraint data to obtain packaging constraint data;

[0028] Step S4: simulating the packaging process of the simulated product framework to be packaged based on the packaging constraint data to generate simulated product packaging data; and constructing a packaging requirement reverse environment field according to the packaging airtightness requirement;

[0029] Step S5: retest the sealing performance of the simulated product packaging data based on the reverse environmental field of the packaging requirements, thereby generating packaging design sealing performance data; perform packaging retrospective optimization on the simulated product packaging data through the packaging design sealing performance data, thereby obtaining qualified product packaging design data, and visually display the qualified product packaging design data online.

[0030] By acquiring the target product data for packaging and determining the outer contour of the product, the present invention can accurately reconstruct the simulated product framework to be packaged, providing a clear basis for subsequent packaging design. The acquisition of packaging airtightness requirements is combined with the estimated packaging usage scenario to formulate reasonable environmental adaptability for packaging design. The packaging process constraint data generated by process constraint mapping ensures that the packaging design meets the actual operation requirements. The analysis of physical packaging constraints helps to fully understand the physical limitations of the packaging process. The product packaging constraint fusion based on these constraints forms systematic packaging constraint data. The packaging process simulation of the simulated product framework to be packaged based on these data can display the dynamic changes in the packaging process in real time. The generated simulated product packaging data provides a reference for design improvement. The construction of the reverse environmental field of packaging requirements provides a scientific basis for the re-testing of the sealing performance. This environmental field re-tests the sealing performance of simulated product packaging data, which can accurately evaluate the sealing of packaging design. The generated packaging design sealing performance data provides data support for subsequent optimization. By performing packaging retrospective optimization on simulated product packaging data, the qualification of the final product packaging design is ensured. The online visual display of qualified product packaging design data not only improves the transparency of the design, but also facilitates all parties involved to monitor and evaluate the design effect in real time. Through the application of digital technology, the overall system not only improves the efficiency of packaging design, but also enhances the flexibility and adaptability of design, laying the foundation for the intelligent transformation of the modern packaging industry, promoting the digitalization and visualization of packaging design, promoting the rational use of resources and cost control, improving the market competitiveness and customer satisfaction of enterprises, and ensuring the efficient response and rapid iteration of packaging design under diversified needs.

[0031] In an embodiment of the present invention, the method for online design and display of product packaging includes the following steps:

[0032] Step S1: Acquire data of a target product to be packaged; determine the outer contour of the product according to the data of the target product to be packaged; and reconstruct a simulated frame of the product to be packaged based on the outer contour of the product;

[0033] In this embodiment, a 3D point cloud data of the product to be packaged is obtained using a 3D scanner (such as Artec Leo) with a resolution better than 0.2 mm. During scanning, the scanning angle is maintained at not less than 45° and continuously rotated to cover the entire outer surface area. The scanned point cloud data is imported into the modeling system in PLY format, and a closed surface is reconstructed using a Poisson reconstruction algorithm. Point density regularization is performed using a voxel filtering algorithm with a point pitch set to 0.5 mm. Subsequently, an outer contour extraction operation based on the alpha shape method is performed with an alpha value set to 1.8 to retain all external morphological protrusions and ignore concave features. The outer contour data is exported in STL format and entered into a structure reconstruction module. The principal axis vector of the STL model is estimated based on boundary mesh structure analysis. The principal component analysis (PCA) method is used to obtain the length principal axis direction vector and the vertical vector, and a spatial skeleton model is constructed. The number of skeleton nodes is limited to not less than 20, and the connections between the nodes are constructed based on the shortest connected paths in space. Finally, a 3D topological structure skeleton model is generated as the simulated framework of the product to be packaged.

[0034] Step S2: Obtaining packaging tightness requirements; estimating packaging usage scenarios based on the packaging tightness requirements; performing process constraint mapping based on the estimated packaging usage scenarios and the packaging tightness requirements to generate packaging process constraint data;

[0035] In this embodiment, the product type, usage location, climate adaptability level and sealing level parameters are used as input variables and imported into the packaging airtightness level classification database, wherein the climate adaptability level is divided into indicator levels such as high temperature (>35°C), humidity (relative humidity>80%), and salt spray corrosion (NaCl content>5%). Each indicator corresponds to a sealing level mapping interval, and the sealing level is divided into five levels from IP54 to IP68. After coupling the packaging airtightness level with the product material information, structural packaging airtightness requirement data is formed. According to the airtightness requirement data, a table is searched to match the typical usage scenario labels in the packaging environment model library. The labels include transportation scenarios (such as "cold chain transportation", "tropical port transportation", "plateau highway transportation"), storage scenarios (such as "constant Warm warehouse", "temporary open air storage"), each label contains multiple environmental factor boundary data, and the airtightness requirement index vector is matched with the environmental field label feature vector by Euclidean distance measurement. The matching threshold is set to within 0.25, and finally the three packaging usage scenarios with the smallest distance are selected as the estimated usage scenarios. The temperature, humidity and air pressure disturbance models corresponding to the three groups of packaging usage scenarios are cross-mapped with the packaging airtightness index, and the mapping tensor of the usage scenario to the process parameters is established using the graph structure mapping method. The process parameters include packaging pressure, seam method, wrapping tape type and composite sealing material structure configuration. All mapping results are summarized as packaging process constraint data. The constraint data format is a structured JSON structure, and the fields include process step, material number and corresponding closed structure parameter value range.

[0036] Step S3: Analyze the physical packaging constraints based on the simulated product framework to be packaged; fuse the product packaging constraints according to the physical packaging constraints and the packaging process constraint data to obtain packaging constraint data;

[0037] In this embodiment, the generated simulation product framework model is input into the physical packaging configuration analysis module in the B-rep (boundary representation) format. The thickness discontinuity area, rigid protrusion area and connection suture surface are identified by facet segmentation operation based on the structural point cloud model. The structural edge gradient change threshold is set to 2.3. The physical contact prediction model is constructed based on the Euler topology structure to simulate the point-surface contact distribution map under the coverage of the paper packaging box structure. The contact area threshold is set to 10 cm 2 It is used to determine the structural fixation requirements and use the contact prediction results as the entity packaging constraint data. The entity packaging constraint data fields include "node identification", "number of contact points", and "rigid coverage level". The entity packaging constraint data is then fused with the aforementioned generated packaging process constraint data through field standardization and weight matching. A weighted normalization strategy is used to construct a fusion score matrix for each packaging structure unit. The fusion score is calculated by combining the entity contact requirements and the packaging method risk weight. The final output is the packaging constraint data including the packaging structure number, constraint weight, and recommended packaging structure model path.

[0038] Step S4: simulating the packaging process of the simulated product framework to be packaged based on the packaging constraint data to generate simulated product packaging data; and constructing a packaging requirement reverse environment field according to the packaging airtightness requirement;

[0039] In this embodiment, the packaging constraint data is input into the packaging path simulation module. The path simulation module constructs a path generation rule based on spatial curvature priority. The first node of each packaging path is limited to the minimum outer dimension surface of the product, and the tail node is the maximum stress concentration surface. The path generation uses a greedy iterative algorithm to search for the path combination with the minimum curvature change. The total length of the path is limited to within 1.5 times the total length of the structure outer contour line. After all feasible paths are generated, simulations are performed one by one. The packaging physical parameter model is loaded on each path, such as winding angle, packaging stacking order, pressing frequency, etc. The pressing frequency is controlled within the range of 0.30.6Hz to generate each path. The corresponding simulated packaging data under each path is collected and stored as a simulated packaging sample set. At the same time, a reverse environmental field for packaging airtightness and packaging requirements is constructed. First, the limit values ​​of the unsatisfied scenarios in the packaging airtightness level are derived, and the reverse environmental disturbance vector field is constructed. The disturbance vector field includes three directional disturbance models: external air pressure mutation (set to dynamic changes in the range of 15-35kPa), high-frequency humidity shock (humidity oscillation range is 40%-100%, frequency is set to 1Hz), and temperature jump (periodic jump amplitude is 3℃ / min-6℃ / min). The disturbance vector field is input into the packaging simulation structure model through a three-dimensional finite difference simulator (such as COMSOL Multiphysics) for environmental field coupling loading. The simulation step size is set to 0.1 second, and the total time period is not less than 600 seconds. Finally, a complete reverse disturbance environmental data field is generated.

[0040] Step S5: retest the sealing performance of the simulated product packaging data based on the reverse environmental field of the packaging requirements, thereby generating packaging design sealing performance data; perform packaging retrospective optimization on the simulated product packaging data through the packaging design sealing performance data, thereby obtaining qualified product packaging design data, and visually display the qualified product packaging design data online.

[0041] In this embodiment, a reverse perturbation environment data field is co-simulated with packaging simulation data. Physical field response monitoring is used to obtain real-time sealing response indicators. These indicators include the pressure leakage rate at the edge seal (target limit: pressure drop of no more than 1.5 kPa per second), the seam lift angle (maximum no more than 15°), and the edge tear initiation load (no less than 30 N). The sealing test results are recorded as time series data. If the sealing performance of a packaging path does not meet the threshold requirements, the influencing point is retrospectively identified based on structural sensitivity analysis. The path parameters are then locally perturbed and optimized using a simulation optimizer (such as the ANSYS Workbench Optimization Module). The perturbation range is limited to ±10% of the winding tension, ±3° of the angle, and 1 layer of packaging. Ultimately, a qualified packaging structure path that meets all sealing standards is generated. The packaging method under this path is converted into a structural model and a packaging process animation model, which is dynamically displayed on a 3D display platform using WebGL rendering. The model frame rate is controlled at 60 frames per second and supports structural detail magnification, path playback, and material structure layer switching.

[0042] Preferably, step S1 includes the following steps:

[0043] Step S11: Acquire packaging target product data; perform structured analysis on the packaging target product data to obtain product structured data;

[0044] Step S12: extracting multi-dimensional product structure parameters from the product structured data, wherein the size range of the multi-dimensional product structure parameters is limited to a length, width, and height range of 10 mm to 5000 mm;

[0045] Step S13: calibrating the product outer contour according to the multi-dimensional product structural parameters, wherein the outer contour error tolerance is controlled within ±0.5mm;

[0046] Step S14: extracting the principal axis vector data of the product outer contour, wherein the length range of the principal axis vector is limited to 80% of the longest side of the product;

[0047] Step S15: Reconstruct the product structure skeleton according to the principal axis vector data, wherein the skeleton node spacing is limited to 5mm-50mm;

[0048] Step S16: Perform product frame simulation based on the product structure skeleton, thereby obtaining a simulated product frame to be packaged.

[0049] In this embodiment, when obtaining the data of the target product for packaging, a FARO FocusS350 three-dimensional laser scanner is used. The scanner is configured with an optical resolution of 1 mm and a scanning frequency of 976,000 points per second. The scanning area is set to a horizontal range of 300 degrees × a vertical range of 360 degrees. In the experimental environment, the target product is placed in the center of the three-dimensional scanning platform. After starting the scanner, the external point cloud data of the product is collected. The data format is .las. The scanning time is controlled within 15 seconds. The obtained scanned point cloud data is transmitted to the graphics workstation via the USB3.0 interface. Subsequently, the PolyWorks Inspector module is called in the graphics workstation to import the .las format data, and point cloud stitching, background removal and noise filtering operations are performed. The background removal method used is the boundary clustering method. The noise point removal adopts the three-dimensional boundary minimum density removal method, in which the neighborhood radius is set to 10 mm and the removal density threshold is set to 0.0015 points / mm. 3Finally, a non-redundant point cloud data file is obtained. When performing structural analysis on the packaging target product data, the surface fitting is performed based on the B-Spline Surface Reconstruction Algorithm. During the processing, the point cloud data format is converted into the STL mesh format. The preliminary mesh reconstruction is performed through MeshLab, and the reconstruction accuracy is set to 1mm triangle piece side length. After the polygonal mesh model is generated, it is imported into the Open3D framework to perform the structure recognition task. The surface normal vector partitioning (Surface Normal The whole grid data is voxelized using the method of Partitioning, and the voxel side length is set to 2mm. The geometric center and normal direction information of all voxel center points are extracted. The face groups are divided by the K-means clustering algorithm, and the number of clusters is set to 10. The product structured data is generated according to the face normal consistency and boundary connectivity. When extracting the multi-dimensional product structure parameters in the product structured data, the key fields in the structured JSON data set are parsed. According to the standard industrial three-dimensional geometric format parsing method, the product three-dimensional dimension boundary point set is obtained, the maximum span in the X, Y, and Z axes is calculated, and the boundary values ​​of length, width, and height are obtained. For each dimension value, the A correction function removes spikes and outliers. This function uses a bilateral filtering rule based on the median and an outlier threshold of 1.5 times the interquartile range. After removal, the results are recalculated to achieve stable dimensions. During the extraction process, the product structural dimensions are limited to a range of 10 mm to 5000 mm for length, width, and height, respectively. If any dimension exceeds this range, a scanning error message is returned and the process is terminated. The extracted multidimensional product structural parameters are saved in a CSV file, including six dimensions: length, width, height, volume, surface area, and main diagonal length. To calibrate the product contour based on these multidimensional structural parameters, a minimum bounding box (MBB) algorithm is used. The three-dimensional boundary values ​​of the product are read from the CSV file and the cv::minAreaRect3D function in OpenCV is used for spatial contour calibration, with a calibration tolerance of ±0.Within 5mm, during the calculation process, the three-axis coordinate system is used to project the product structure point cloud to the XY, YZ, and XZ planes in sequence, and the minimum circumscribed rectangle is fitted respectively. Then, the three sets of two-dimensional boundaries are combined into a spatial circumscribed rectangular parallelepiped model. Through cross-checking, it is ensured that the three sets of fitting contours are superimposed without error. Finally, the product outer contour model data with spatial position, direction vector, envelope size and error index is generated. When extracting the principal axis vector data of the product outer contour, the principal axis direction is calculated based on the outer contour envelope rectangular parallelepiped model, and PCA (Principal Component Analysis) is called. The main direction vector is extracted using the point cloud coordinate analysis) algorithm. The input is set as a point cloud coordinate set. The covariance matrix is ​​calculated first, and then the eigenvalue decomposition is performed on the covariance matrix. The eigenvector corresponding to the maximum eigenvalue is selected as the main axis direction. Then, the complete main axis vector is constructed through the vector from the origin to the maximum boundary point. The main axis vector length is set to 80% of the longest side value in the three-dimensional size of the product. During the value-taking process, the eigenvector needs to be normalized and rescaled to 80% of the length. The vector drawing function is called in the three-dimensional visualization tool CloudCompare to visually check the extraction results. If the main axis direction deviates from the symmetry direction of the product structure by more than 15 degrees, the coordinate base is readjusted and the main axis vector is extracted again. When reconstructing the product structure skeleton according to the main axis vector data, a skeleton reconstruction method based on linear interpolation is adopted. , the main axis vector is evenly segmented with a unit length of 5mm in the main direction, each segment is used as a skeleton node, and the nodes are connected in pairs in the form of spatial straight line segments to form a preliminary skeleton line. The node spacing is set to be evenly distributed in the range of 5mm to 50mm. The density is adaptively adjusted according to the complexity of the product surface shape. The local normal vector and curvature information of the corresponding envelope surface are extracted at each skeleton node position. The skeleton node position is adjusted through the mapping function to make it fit the outer envelope contour more closely. After completing the skeleton structure fine-tuning, the skeleton data is stored as an .obj format model file with topological relationships, including node coordinates, connection order, skeleton segment length and other information. The 3D view is rendered through the Blender tool to verify the consistency of the skeleton structure. When the product frame simulation is performed based on the product structure skeleton, in Siemens Import the .obj skeleton file into the NX simulation environment, perform solid extrusion operations using the NX modeling module, and construct a rectangular cross-section pipe structure on each skeleton segment. The cross-sectional side length is set to 10mm, and the material is simulated as standard industrial ABS plastic. After assigning physical properties to the model, the finite element analysis module is used to simulate structural integrity. Loading conditions include concentrated loads of 100N, 200N, and 300N in the X, Y, and Z axes. Stress concentration areas and node displacements are examined. A structural stress distribution diagram is generated with a simulation accuracy of 0.1mm. The post-simulation stable frame model is exported in .step format, storing the simulated mechanical property values ​​and stress distribution point information.

[0050] Preferably, step S2 includes the following steps:

[0051] Step S21: Obtain target product usage data; determine its packaging airtightness requirements based on the target product usage data, wherein the target minimum gas leakage rate in the airtightness requirements is limited to no more than 0.005 L / min;

[0052] Step S22: Mapping usage scenarios based on the packaging airtightness requirements and target product usage data to generate estimated packaging usage scenarios;

[0053] Step S23: Determine a packaging structure strength threshold based on the estimated packaging usage scenario and packaging airtightness requirements, wherein the determined static compressive strength threshold is limited to 50-800 kPa;

[0054] Step S24: performing packaging process constraint mapping based on the packaging structure strength threshold, wherein the maximum load deformation limit in the constraint mapping is controlled to be no more than 3% of the short side dimension of the product, so as to generate packaging process constraint data.

[0055] In this embodiment, when obtaining the target product usage data, the industrial product field operation log system interface is called to extract the target product's 120-hour continuous operation data in the actual application environment, including temperature and humidity fluctuation records, operation vibration frequency, external gas pressure difference changes, and collision records during actual transportation. At the same time, the instantaneous air pressure change information recorded by the high-precision air pressure sensor model MS5803 is used in conjunction with the DS18B20 temperature sensor and the SHT35 humidity sensor to obtain the environmental temperature and humidity conditions during the product usage period. All sensor data are collected by the PLC controller and uploaded to the data processing server through the Modbus protocol. The server's built-in structured data cleaning program performs data segmentation, error value elimination, and unified unit conversion on the original log data. The gas pressure difference unit is uniformly converted to kilopascals (kPa), and the vibration record is recorded in acceleration (m / s 2) measurement, and then the threshold judgment function is used to hierarchically encode the usage conditions in different scenarios into three levels: strong, medium and weak, and bind them to the product structured data at the field level. Based on the processed usage data, the environmental response assessment model is used to judge the packaging sealing requirements. The model is implemented in Python and embedded in the data analysis module. The sealing requirement takes the target minimum gas leakage rate as the core indicator, and the obtained air pressure change amplitude is dynamically coupled with the product cavity volume data. Assuming that the initial volume of the cavity is 3.2L, the pressure difference change value in 30 minutes shall not exceed 0.075kPa. According to the conversion formula, the upper limit of the target leakage rate is calculated to be 0.0049L / min. It is concluded that the current product sealing requirement level is high. At the same time, the corresponding sealing level standard template is called in the database and matched to the product usage scenario. Finally, the sealing requirement generation and output corresponding to the target product are completed. According to the generated sealing requirement data and product usage data, the graph neural network based on usage scenario label matching is used. The model uses a scene feature mapping network model. The input of the model receives the standardized usage environment parameter matrix and the airtightness requirement intensity scalar respectively. The environmental parameter matrix is ​​a 6-dimensional input vector, including temperature change rate, humidity fluctuation amplitude, impact frequency, air pressure fluctuation amplitude, vibration intensity and site movement frequency. The output is one of 12 typical packaging usage scenario categories. The model has been trained with 12,000 sets of product application data. The current output category is "remote vibration-low pressure sealing scenario". At the same time, a weight value vector is returned to characterize the criticality of parameters in this scenario. The joint weight of vibration intensity and air pressure fluctuation exceeds 0.72. The system automatically loads the scenario parameter template and anti-interference standard according to the output category to form an estimated packaging usage scenario description file for the current product. According to the estimated packaging usage scenario and airtightness requirement, the finite element analysis module is called to perform static loading simulation on the packaging carrier material properties and product structure parameters. The material used in the simulation model has a density of 980kg / m 3, high-strength polypropylene material with an elastic modulus of 2100MPa and a Poisson's ratio of 0.3 is used as the reference packaging material. The external pressure is set to gradually increase from 0kPa to 1000kPa, and the total deformation and stress distribution of the product structure skeleton under different pressure levels are recorded. The deformation at maximum static compressive strength is controlled at 1% of the minimum side length of the product. The external pressure corresponding to the maximum stress is taken as the compressive strength threshold. Through regression analysis of 10 groups of simulation results, it is concluded that the static compressive strength threshold required for the current packaging structure is 620kPa. After storage, this threshold is called in the subsequent constraint mapping module, and the relevant area is marked in red in the three-dimensional packaging structure preview module to indicate the high strength demand segment. With 620kPa as the static compressive strength threshold of the packaging structure, the packaging deformation limit model in the structural constraint mapping module is called, and the input parameter is the product shortest side data 230mm. The maximum packaging load is set to 750kPa. The maximum surface deformation of the packaging structure under different loads is recorded through step-by-step stress loading analysis. The surface deformation contour is fitted using point cloud data and differentially compared with the original unloaded model. The maximum deformation boundary value is extracted to be 6.4mm, which accounts for 2.78% of the shortest side of the product, meeting the requirement that the maximum load deformation limit shall not exceed 3% of the short side size of the product. Therefore, it is determined that the current packaging structure constraint is qualified. By superimposing the deformation limit threshold with the skeleton node vector data, a new packaging process constraint matrix is ​​formed, in which each skeleton node additionally stores a compression tolerance value field. This matrix is ​​used as the structural strength boundary input of the subsequent online packaging design module and is used in the real-time modeling module to limit the boundary parameter constraints when the packaging scheme is automatically generated, and finally the packaging process constraint data is generated.

[0056] Preferably, analyzing the physical packaging constraints based on the simulated product framework in step S3 includes:

[0057] Perform spatial topological projection on the simulated product frame to be packaged, thereby obtaining spatial topological projection data;

[0058] Identify structural node feature data in spatial topological projection data;

[0059] Identify and simulate the protruding node areas in the frame of the product to be packaged based on the structural node feature data;

[0060] The irregular area frame and the regular area frame in the frame of the product to be packaged are simulated based on the prominent node area segmentation;

[0061] Performing entity packaging simulation on the irregular area framework to obtain irregular area packaging simulation data;

[0062] Performing entity packaging simulation on the regular area framework to generate regular area packaging simulation data;

[0063] Inferring the impact data of packaging area morphology based on irregular area packaging simulation data and regular area packaging simulation data;

[0064] Map the packaging area morphology impact data to the packaging area morphology impact data.

[0065] In this embodiment, in the process of performing spatial topological projection on the simulated product frame to be packaged, the simulated product frame data input in the three-dimensional model format is imported into the geometric modeling engine built based on the OpenCASCADE geometric kernel, and the TopExp_Explorer class is used to iteratively access all the face elements in the model, and perform face element projection in the Z-axis direction. By reading the three-dimensional coordinate data of the point set in each TopoDS_Face structure and converting it into a two-dimensional XY plane coordinate through the matrix projection formula, a two-dimensional plane closed figure set composed of the face element contour boundary is formed, and then the BRepTools class provided by OCC is used. All closed contours are exported as a two-dimensional topological set consisting of TopoDS_Wire objects as spatial topological projection data. The above two-dimensional topological data in SVG format is imported into the image analysis engine based on OpenCV, and the edge extraction operation is performed after the image is grayed out. By calling the Canny operator in OpenCV and setting the upper and lower thresholds to 80 and 150, all contour lines are obtained in the edge image. Then, the corner detection function cv::goodFeaturesToTrack is used, and the maximum number of corner points is set to 500, the minimum quality level is set to 0.01, and the minimum corner point spacing is set to 10 pixels. The geometric inflection points in the two-dimensional plane are extracted as the initial The initial node feature point set is obtained, and each feature point is mapped to its corresponding SVG topological boundary through Hash mapping to generate structural node feature data, and the two-dimensional coordinates, number of connected edges and angle information between adjacent nodes of each node are recorded in JSON format. Based on the number of edge connections of all nodes extracted from JSON, high-connectivity nodes with the number of edge connections greater than or equal to 5 are screened out, and the angle difference between adjacent edges of these high-connectivity nodes is further calculated. By setting the angle difference threshold to 45 degrees, nodes with high edge connectivity and uneven angle distribution are identified as morphological mutation nodes, and all nodes that meet this condition are used as the center identification points of the prominent node area, and then through KMe The ans clustering algorithm clusters all nodes in XY space, setting the number of clusters to 10% of the total number of nodes in the model. A geometric envelope operation is then performed on the contours of the surrounding adjacent edges, with each prominent node as the cluster centroid. The enclosed region around each prominent node is output as the basis for subsequent segmentation of the prominent node region. A polygonal set of contours for all prominent node regions is constructed in a two-dimensional topological plane. The topology of the overall simulation framework is then segmented into multiple regions based on these contour polygons. Within each region, OpenCV's contour fitting function, cv::fitEllipse, is called to fit the shape of each closed region and calculate the ratio of its major axis to its minor axis. The aspect ratio threshold is set to 1.2. If the aspect ratio of any fitted figure in the region is greater than the threshold and contains more than one boundary concave point, it will be marked as an irregular region frame, and the rest will be classified as regular region frames. Finally, two sets are generated to record the region number, boundary node coordinates and structure tags respectively and stored as binary structured files for subsequent use. In the process of physical packaging simulation of the irregular region frame, the finite element analysis module embedded in Ansys Mechanical is used to select the CAD model corresponding to each contour in the irregular region frame as the input object, and the quadrilateral structure is divided with a unit side length of 10 mm. The boundary nodes are set as fixed support conditions. A uniformly distributed pressure was applied to the central region, with a step-by-step loading range of 100 kPa to 600 kPa, with each step being 50 kPa. After the simulation, the deformation and principal stress values ​​of each node were extracted. By comparing the deformation distribution with the area of ​​maximum stress concentration, the stress tolerance range and recommended structural thickness of the packaging material in the irregular region were determined. Each simulation result was recorded as a separate entry and exported as a packaging simulation data XML file. This file records the deformation data, contact area, node offset, and coordinates of the maximum stress point for each element at each pressure. A solid packaging simulation of the regular region framework was performed to generate the regular region packaging simulation data. The SolidWorks Simulation plug-in was used to construct the 3D structural boundary. The external boundaries of each regular region in the regular region framework were extracted and a closed volume structure was constructed. The material parameters were set to PET plastic with a density of 1.38 g / cm. 3, Young's modulus is 3.1GPa, Poisson's ratio is 0.38, the upper surface is set as a uniform force surface, the loading direction is vertical to the Z axis, the loading value is set to a gradually increasing constant surface pressure, from 0.1MPa to 0.8MPa, with each step of 0.1MPa, the meshing parameters are set to a maximum unit side length of 8mm, and the contact surface adaptive adjustment function is turned on. After each loading, the total deformation of the overall model, maximum contact force, critical displacement and Von Mises stress distribution data of each node are collected. All simulation data are exported as structured table data containing area ID, loading steps, maximum stress position, and model deformation path. In the process of inferring the packaging area morphology influence data based on irregular area packaging simulation data and regular area packaging simulation data, a customized multi-source data fusion analysis module is used. First, the packaging simulation data of two different source formats are converted into a unified Pandas. DataFrame structure, the fields include area ID, loading value, maximum node offset, principal stress range, model boundary change rate, material compression ratio, and then the RandomForestRegressor model in Scikit-learn is called for training. The training goal is set to predict the deformation trend value and stress overload risk level of the packaging structure. The input features include the difference in deformation amplitude and stress peak value between irregular areas and regular areas under the same loading conditions. The model is cross-validated for 100 rounds to obtain feature weights. Finally, the predicted morphological impact level is mapped to each area to generate packaging area morphological impact data, and the packaging area morphological impact data is mapped to the process of packaging area morphological impact data. In the process, based on the established simulation product packaging framework model, the morphological impact data obtained by the above inference are first embedded region by region in the three-dimensional model. The vertex attribute injection method in the OpenGL rendering engine is used to map the morphological impact level corresponding to each region to the RGB attribute value of the vertices on the three-dimensional structure surface through color coding. The higher the level, the more red the color, and the lower the level, the more blue the color. The mapping structure is displayed in an interactive and draggable manner in a three-dimensional visualization environment. In addition, the region number and the corresponding morphological impact data are exported to JSON format for subsequent online display system calls, thereby realizing the spatial positioning of the packaging region morphological impact data on the three-dimensional model, visualization of the intensity distribution, and interactive reading of the structural analysis data.

[0066] Preferably, in step S3, fusing product packaging constraints according to the physical packaging constraints and the packaging process constraint data includes:

[0067] Labeling entity packaging constraint data with constraint attributes to obtain packaging constraint label data;

[0068] Decompose the packaging process constraint data into process nodes to generate process node constraint data;

[0069] Perform interactive correlation mapping on the packaging constraint label data and the process node constraint data to obtain label node mapping data;

[0070] The packaging constraint label data and the process node constraint data are constrained and integrated based on the label node mapping data, thereby obtaining the packaging constraint data.

[0071] In this embodiment, a structural constraint attribute matrix is ​​constructed based on the irregular area packaging simulation data and the regular area packaging simulation data generated in the step. The matrix takes the geometric shape data of each area as the main dimension, and extracts the length, width, height, protrusion point coordinate set, embedding point coordinate set, number of rigid connection segments, detachable connection segment identifier and docking surface coordinate area set respectively. Each set of data is defined as a constraint attribute field, and a field dictionary set is constructed. Subsequently, the Pandas tool in the Python language is used to standardize the structural matrix in a DataFrame manner, and the numerical or Boolean content corresponding to each field is mapped to an attribute label. For example, a length field value less than 100 mm is mapped to a "short side structure", and a number of embedding point coordinate sets greater than 3 is mapped to a "multi-embedded surface structure". Finally, no less than 8 attribute labels are generated for each packaging area and stored in a JSON structure. The key field in the structure is the area identifier, and the value field is the corresponding attribute label set. A standard packaging process diagram is constructed. The diagram uses the process modeling language BPMN (Business Process Model and The modeling process is modeled using the Business Process Modeling and Notation (BPM) format. The node types set include packaging material supply nodes, positioning and calibration nodes, alignment and forming nodes, pressing and bonding nodes, nesting and fixing nodes, corner trimming nodes, and barcode pasting nodes. Each node must be accompanied by operation time, packaging action dimensions, fixture type requirements, and node activation restrictions. The process modeling is performed using the Camunda modeling tool and exported as a .bpmn file format. The Camunda parsing engine is then used to analyze the process.The bpmn file is structurally parsed node by node, each node is disassembled into a key-value pair structure with constraint fields, and a node dependency mapping table based on sequential flow control is formed, and finally the process node constraint data is generated. The data adopts a nested Map structure, in which the first-level key is the node number, the second-level key is the constraint field name, and the value is the field value. The query mapping relationship between the label and the node is defined in GraphQL. First, the package constraint label data is loaded into the Neo4j graph database, each package attribute label is mapped to an attribute node, and the label type is set to "AttrLabel". Each node in the process node constraint data is mapped to a node entity with the type "ProcNode". At the same time, bidirectional edges are set to connect the label and the node based on the relevance of the package structure field. The edge type is "ConstraintLink". The edge weight is set according to the matching degree of the associated field. The matching degree calculation formula is the frequency ratio of the same field value in the corresponding field set. In this way, the graph structure generation is completed, and then the Cypher query is written The language script traverses the edges connecting nodes and labels node by node and generates a label-node mapping table. This mapping table is formatted as a CSV file, with fields including label name, node number, weight value, and field source. This CSV data ultimately serves as the label-node mapping data. A threshold of 0.35 is set based on the weight value in the mapping table, retaining only links between labels and process nodes with weights greater than or equal to this value. A fusion constraint object structure is then constructed, with the process node as the primary key. Constraints for the corresponding label are fused under each node. If the label is "Multi-embedded Surface Structure" and the process node type is "Nested Fixed Node," a "Nested Insert Calibration" action is added to the node's packaging action field. If the label is "Short Edge Structure" and the process node is "Aligned Forming Node," the "Clamping Mechanism Short Edge Guide Groove" restriction is added to the fixture type requirement field. A fusion function is constructed using Python to merge and overwrite fields. Finally, all fused node structures are saved in YAML format. This structure serves as the packaging constraint data, comprising a complete constraint configuration set encompassing both structural attributes and process behavior constraints.

[0072] Preferably, in step S4, simulating the packaging process of the simulated product framework to be packaged based on the packaging constraint data includes:

[0073] Enumerate the packaging paths of the simulated product framework to be packaged, thereby obtaining a collection of candidate product packaging processes;

[0074] Based on the packaging constraint data, the path stability of the candidate product packaging process set is evaluated to generate path characteristic parameters;

[0075] Performing adaptive path screening on the candidate product packaging process collection according to the path characteristic parameters to generate an adaptive packaging process;

[0076] By adapting the packaging process, the packaging process of the simulated product framework to be packaged is simulated to generate simulated product packaging data.

[0077] In this embodiment, the structure level flattening process is performed based on the assembly structure diagram of the 3D model. First, the OpenCascade 3D geometry kernel is used to traverse the topological structure tree of the 3D model of the packaging product to be packaged, extract all the connected entity units and connection relationship nodes, record the parent-child connection relationship, connection type (such as plug-in, screw connection, slide rail docking), facing direction and assembly direction vector of each subassembly or part, generate the structure level table and direction table, and then use the depth-first traversal method (Depth First Traversal) to traverse the 3D model of the packaging product to extract all the connected entity units and connection relationship nodes. Search) generates a set of all feasible disassembly and assembly path sequences, records them in reverse as a set of packaged paths, and calculates the path turning points and rotation points based on the packaging direction and center of gravity offset. Each path contains the step number, the corresponding node number, the relative displacement vector, the clamping surface identification mark and the operation direction mark, and finally obtains a collection of candidate product packaging processes. Each path is stored in JSON format, and the fields include the path sequence number, the operation step sequence list, the path length, the number of actions, the node operation vector sequence, and the visual angle change trajectory value set. The action steps of each path are mapped to the node operation restriction set in the constraint data, and the clamping device restriction conditions, minimum operation space threshold, dynamic torque limit threshold and path allowable redundancy range corresponding to each process action in the packaging constraint data are read. For the action vector in each path, the center of gravity position change value of the corresponding assembly surface, the path rotation matrix change amplitude, and the local space envelope volume change rate are calculated, and three characteristic parameter indicators are set respectively: the center of gravity drift rate CgRate, the operation Redundancy (OpRedundancy) and path gripping stability (GripStability) are measured. CgRate is defined as the rate of change of the three-dimensional coordinates of the center of gravity during the continuous motions before and after the path. GripStability is calculated as the cosine of the angle between the gripping surface normal and the path displacement vector. OpRedundancy is calculated as the ratio of the difference between the continuous motion direction vectors and the intersection volume of the operating space. All indicators are numerically represented as a path characteristic parameter vector. Each path ultimately generates a record row containing the path ID, CgRate, OpRedundancy, and GripStability. All path records are aggregated in a CSV file format with the fields PathID, CgRate, OpRedundancy, and GripStability. A set of filtering thresholds is set: CgRate must not exceed 0.12, OpRedundancy must be greater than 0.45, and GripStability must be greater than 0.85, traverse each path record in the CSV record file, compare each field to see if it meets the set threshold conditions, if any field does not meet the conditions, then remove the path, and include the paths that meet all the conditions as the adaptation path set. At the same time, construct a comprehensive adaptation score value based on the weighted average of the two indicators OpRedundancy and GripStability, with a weight ratio of 0.4 and 0.6. After sorting the adaptation path set from high to low according to the comprehensive score, take the top five paths as the final adaptation packaging process, renumber each process path, and add the process ID and sorting weight label. Finally, reorganize the adaptation path content in JSON structure, including path dynamics. The operation sequence, clamping surface, action direction, posture change value and score value are stored in independent fields in the structure. The Rigidbody physical simulation module in the Unity3D engine is used to simulate the motion of the three-dimensional packaging model. The operation steps in each path are converted into a control instruction sequence, including displacement control, rotation control, clamping control, and collision detection control. In each step of control, the operation vector in the path is used to control the movement direction and distance of the three-dimensional model rigid body. The clamping control simulates the clamping contact point by hanging a virtual fixture model with a Trigger attribute. The collision detection module is set to a transparent shell of the packaging structure and the physical material is turned on as "Box Collider", set the collision feedback threshold to no penetration within 0.002 units, and the motion control step length to 20ms per frame. Each path execution sequence outputs the action state, model posture quaternion, operation feedback flag, and collision flag at each moment after the physical frame is executed. After each path simulation is executed, a complete packaging simulation record is generated. The record format is time-serialized JSON, including the timestamp, operation ID, 3D coordinates, fixture contact surface name, operation result, and success / failure flag. The simulated product packaging data generated by each path is archived and saved in a result directory named by the path ID.

[0078] Preferably, in step S4, constructing a packaging requirement reverse environment field according to the packaging airtightness requirement includes:

[0079] Screen and estimate suitable packaging usage scenario data in the estimated packaging usage scenarios based on packaging airtightness requirements;

[0080] Identify the adaptive scenario features in the packaging usage scenario data that meet the packaging airtightness requirements;

[0081] Direction adjustment simulation is performed based on the adapted scene features, thereby generating unadapted scene features;

[0082] Construct a reverse environment field for packaging requirements based on the characteristics of unsuitable scenarios.

[0083] In this embodiment, a set of package sealing requirement parameters is defined. This set is determined by the product type and includes five items: sealing holding time, maximum air pressure difference, humidity tolerance threshold, particle intrusion tolerance level, and corrosion factor exposure limit concentration value. Each item is expressed in standard dimensional units, such as sealing holding time in hours, maximum air pressure difference in kilopascals, humidity tolerance in relative humidity percentage, and particle intrusion level corresponding to ISO 9001. 14644 standard cleanliness level, the exposure concentration of the corrosion factor is measured in ppm. Then, all scenario data containing complete records of the above five environmental parameters are screened from the constructed packaging usage scenario database. The database records are represented by a two-dimensional structure, and the fields include scenario number, usage area code, transportation mode, storage time period, ambient temperature, humidity and pressure record sequence, particle counter output value sequence and corrosion concentration sensor sampling sequence. A Python script is used for traversal and screening. The script uses Pandas as the basis to build a Boolean index mask, and extracts the record rows that meet the lower limit conditions of all packaging airtightness requirement parameters as the adapted packaging usage scenario data. Each scenario data in the above screening results is subjected to feature extraction processing, and PCA (Principal Component Analysis) in Scikit-learn is used. The covariance matrix and eigenvector extraction of five environmental parameters are constructed using the PCA method. During the extraction process, all parameters are first normalized to zero mean. The mean of each column is subtracted and divided by the standard deviation, so that each parameter is transformed into a principal component under unit variance. Then, the threshold of the principal component variance contribution rate is set to 95%, and the first n principal components are automatically selected, generally the first 2 to 3. The generated principal components are the linear combination vectors of each environmental parameter. The PCA processing results include the principal component score matrix and the projection coordinates of each scene data in the principal component space. The projection coordinates are the adaptation The scenario eigenvalue is used to characterize the distribution of comprehensive factors affecting the package tightness of the scenario and serves as the initial parameter input for subsequent directional adjustment simulations. Ultimately, the eigenvalues ​​of all compatible scenarios are recorded as three-dimensional vectors, with the dimension being the number of scenarios multiplied by the number of principal components. A characteristic perturbation model is constructed to generate the characteristics of unsuitable scenarios. This model is based on the directional rotation matrix construction method. That is, in the principal component space, with each scenario eigenvector as the reference, a perturbation vector in a specific direction is applied and the resulting offset is observed. The perturbation vector is generated by uniform sampling of the three-dimensional sphere, and the modulus of each perturbation vector is set to 0.1 to 0.5, each disturbance is equivalent to rotating around the unit vector axis in the principal component space of the current scene environment feature. The rotation matrix is ​​generated using the Rodrigues formula (vector rotation formula). A set of disturbances is performed on the original scene features through the perturbation rotation operation to generate new features. Five sets of disturbance samples are generated for each original scene, and the number of scenes is multiplied by five unsuitable scene feature vectors. These unsuitable features are then remapped back to the original parameter space. The five packaging environment parameter values ​​are restored through the inverse transformation matrix, and then compared with the lower limit of the packaging airtightness requirement parameter item by item. If any item does not meet the requirements, it is marked as a valid unsuitable sample, and the disturbance vector parameters, the principal component values ​​after disturbance, and the original restored parameter values ​​are output. All unsuitable features finally form a three-dimensional array with the dimension of the original number of scenes multiplied by five times the number of principal components. A distribution fitting model is constructed for the unsuitable scene feature values. First, KDE (Kernel A kernel density estimation (KD) algorithm performs joint probability density modeling on all mismatched principal component eigenvalues. Using a Gaussian kernel function with a bandwidth coefficient of 0.15, each dimension in the principal component space is jointly modeled to capture high-probability regions. The resulting density model represents characteristic regions with a high risk of leak-tightness failure. Based on this model, a point set representing typical reverse environmental characteristics is sampled, with 300 samples taken at a time. The sampled points are then projected back into the original environmental parameter space, and each parameter value is recovered using an inverse PCA transform. This value constitutes a reverse environmental field parameter sample set, where each sample contains five parameters: humidity, air pressure, corrosion concentration, particle count, and seal retention perturbation factor. This parameter sample set is ultimately used to drive the construction of a virtual simulation environment. The parameters are loaded through the environment control module in the Unity3D environment rendering engine. Each set of parameters is used to generate an independent packaging simulation environment, including adjustable temperature and humidity dynamic curves, particle injection simulation, air pressure perturbation simulation, and corrosive gas generation. All reverse environmental field samples are recorded in a database with fields including the environment number, perturbation parameter, environment configuration file path, and mapping feature index number.

[0084] It is particularly important to construct a reverse environment field for packaging requirements based on the characteristics of unsuitable scenarios, including:

[0085] Perform field inverse tensor mapping on the unsuitable scene feature data to obtain inverse tensor field data;

[0086] Introducing multi-dimensional field intensity perturbation into the inverse tensor field data to generate perturbation field intensity data;

[0087] Performing adaptive threshold replacement on the disturbance field strength data to obtain adaptive threshold offset data;

[0088] Performing reverse environmental load superposition on the adaptation threshold offset data to generate load superposition field data;

[0089] Construct the reverse environmental field of packaging demand based on load superposition field data.

[0090] In this embodiment, the generated unsuitable scene feature data is converted into a fixed-length vector group through a feature vectorization module. Each vector group contains 9-dimensional data, which respectively represent the upper limit of packaging pressure, humidity critical value, temperature change rate, light intensity, particulate matter content, air flow rate, gas erosion coefficient, time period factor and boundary deformation amplitude. The conversion tool used is a feature parsing network trained based on the TensorFlow framework. The network input accepts the original feature map data with a floating-point matrix size of 512×512, and performs feature extraction through a three-layer convolutional neural network (the convolution kernel size of each layer is 3×3, the step size is 1, and the output channels are 32, 64, and 128 respectively), and finally generates a feature map at the output. The vector data after tensor encoding is converted into tensor-encoded data, and then the forward tensor vector group is mapped to the tensor inverse domain in the packaging parameter space through the reverse tensor mapping function R(T) through the function inversion operation. In this operation, the tensor inverse domain uses the mapping constraint function G(x)=-αx+b, where α takes the value of 1.5 and b takes the value of 0. The reverse recoding of all vector groups is completed and the reverse tensor field data is regenerated. The output dimension remains consistent with the 9-dimensional structure of the original vector group. The data excitation process is completed by introducing an external perturbation tensor perturbation function set. The perturbation function set is constructed from the superposition of three groups of Gaussian distribution functions and two groups of Poisson distribution functions. Its parameters are set to the Gaussian function mean μ is 0.3, 0.5, and 0.7, and the standard deviation σ is 0.1, 0.2, 0.15, the Poisson distribution λ is 3 and 5, the perturbation function corresponds to the position perturbation intensity and amplitude variation coefficient of each dimension of the tensor in the tensor space respectively. During the processing, the perturbation function parameters are used as the input of the perturbation layer and applied to each dimension of the inverse tensor field data through the tensor bias generation module. The maximum perturbation amplitude is controlled within the range of ±25% of the original value. When all perturbation tensors are applied, a synchronous perturbation superposition strategy is adopted, that is, the perturbation amplitude of all dimensions is applied simultaneously in a single processing cycle. The generated perturbation field strength data still maintains a 9-dimensional structure, but the numerical distribution shows nonlinear random fluctuation characteristics. The critical trigger point in the perturbation tensor data is extracted through the threshold mapping module, and the threshold setting is based on the historical package The 95th percentile damage initiation parameter values ​​from the damage database are collected. This database contains 4,000 sets of damage threshold values ​​for different material packages in various external environments. Values ​​above the 95th percentile in each dimension of the tensor are extracted as key feature points exceeding the adaptation threshold. A permutation operation is then performed using the threshold permutation function F(x) = x - Δv, where Δv is the difference between the current perturbation value and the target adaptation value. All identified critical points are permuted and adjusted based on the target adaptation value of the corresponding dimension to relocate them to within the maximum adaptation range. A dynamic step-size adjustment mechanism is used during the permutation process, with an initial step size of 0.01. If the adjusted data still exceeds the target adaptation range, the step size is increased to 1 of the current step size.5 times until the adjustment is successful. The replaced data constitutes the adaptation threshold offset data. The 9-dimensional structure of the data remains unchanged, but the value of each dimension is within the specified adaptation threshold range. A load field function set is constructed to simulate the composite superposition effect of the external field pressure, wind pressure, chemical etching and structural compression of the packaging. The function set includes 4 types of load types. Each type of load is mapped to the corresponding dimension of the tensor structure. The superposition method is multiplicative composite superposition, that is, Vn=Vn×(1+Ln), where Vn is the offset data dimension value, Ln is the current load superposition coefficient, and the four types of load coefficients are set to external field pressure 0.12, wind pressure 0.18, etching 0.09, and compression 0.16, respectively. Each set of coefficients is fixed based on the physical simulation model. The superposition operation is batch calculated through the reverse environment simulation module. After all loads are superimposed, each set of tensor data obtains a new set of tensor data. The data maintains the original structure, but all dimensions undergo dynamic expansion or compression changes based on the load. The processed tensor is Load superposition field data is used as input for packaging structure scene rendering through a multi-channel tensor mapping network. This network consists of three layers, each of which includes a set of mapping matrices. The first channel maps the 9-dimensional tensor into 3D environmental field variables, namely, a closed pressure point distribution map, a boundary stress map, and a local tension distribution map. The second channel constructs these variables into a distributed point cloud model. The third channel uses a lighting rendering engine to perform 3D dynamic rendering on a GPU to generate images. This image data is ultimately output as a dynamic image sequence with a 0.5-second refresh rate. Each frame in the image sequence provides real-time visualization of packaging sealing mismatch locations. All output images are integrated into a reverse environmental field scene data file through an encapsulation module that can be directly called by the online design engine. The format is a three-dimensional data package consisting of OBJ, MTL, and PNG, with a standard binary attribute table recording all the original tensor inputs and transformation parameters, completing the construction and output of the reverse environmental field for the final packaging requirements.

[0091] Preferably, in step S5, retesting the sealing performance of the simulated product packaging data based on the reverse environmental field of the packaging requirements includes:

[0092] Based on the reverse environment field of packaging requirements, the simulated product packaging data is scene-coupled and projected to generate reverse environment loading data;

[0093] Conduct airtightness test on simulated product packaging data according to reverse environment loading data to obtain reverse environment airtightness performance;

[0094] Re-determine the packaging design sealing performance data of the simulated product packaging data based on the reverse environment sealing performance.

[0095] In this embodiment, in the process of scene coupling projection of simulated product packaging data based on the reverse environment field of packaging requirements, the three-dimensional disturbance axis component, directional flow density component and critical load threshold component are extracted from the constructed load superposition field data, which respectively represent the spatial dynamic disturbance, pressure projection direction and stress boundary limit of the reverse environment field at the airtightness impact level. The scene coupler module in the three-dimensional environment loading engine is used to set the spatial mapping function G(u,v,w)=F(E×T), where u, v, and w are the spatial voxel coordinates in the simulated product packaging model, E is the reverse environment loading vector matrix, and T is the packaging material structure stress response parameter matrix. The structure unit partition is generated according to the structure unit partition. The mapping function F loads the superposition field energy vector frame by frame according to the local field strength parameters of each voxel unit, thereby completing the spatial superposition mapping between the simulated packaging structure and the reverse environmental field of the packaging demand in the virtual space, and then generating the reverse environmental loading data. The output data structure is a three-dimensional closed influence field block matrix, each element of which contains the closed disturbance component value, the structural response coefficient and the superposition energy boundary value. The operation process of performing a tightness test on the simulated product packaging data according to the reverse environmental loading data includes: first, according to the disturbance value of each three-dimensional point and the corresponding structural unit number provided in the loading data, starting the packaging pressurization simulation module in the sealing performance test system, using the external intelligent loading device (Environmental Loading Actuator) to load the simulated disturbance signal region by region according to the voxel unit, and the disturbance parameter control range is set to 0.4 to 2.5N / cm 2The pressure pulse frequency band is between 25Hz and 200Hz. When loading, the local deformation, crack response and joint stress diffusion value of the structure are collected frame by frame through the micro-structure response sensing unit (Micromechanical Response Probe). The collection cycle is 10ms. A multi-modal seal status recognition network (Multi-modal SealStatus Network) is embedded in the test system. The collected acoustic attenuation signal, pressure leakage threshold response and internal humidity change value are heterogeneously fused and compared to generate a reverse environment sealing performance score corresponding to each voxel structure area. The scoring dimension is 0 to 1, and the area with a threshold value less than 0.6 is defined as a low sealing area. In the process of re-determining the packaging design sealing performance data of the simulated product packaging data based on the reverse environment sealing performance, the three-dimensional scoring matrix obtained above is subjected to sealing reconstruction analysis, and the built-in sealing adjustment module (Seal Adjustment Synthesizer) takes each area with a sealing score less than 0.6 as the target area, extracts its local material structure parameters, disturbance response records, and edge connection properties, and recalculates the required sealing enhancement factor K for this area. The K value is calculated using the material interface reconstruction function H = α × L + β × θ, where L is the connection edge length parameter, θ is the local stress angle response, and α and β are preset material response constants with set values ​​of 1.25 and 0.88, respectively. Based on the K value, the sealing structure parameters of this area in the virtual packaging model are updated, including increasing the edge sealing thickness to 1.2 times the original value, increasing the adhesive density by 30%, and increasing the distribution frequency of the connection angle reinforcement ribs by 1.5 times. Finally, the packaging design sealing performance data containing all the updated sealing structure information is generated. The output format is the enhanced sealing voxel table data structure, and the 3D packaging state model is rendered in the visualization engine at the same time for the next step of displaying the simulation analysis process call preparation.

[0096] Preferably, in step S5, performing packaging retrospective optimization on the simulated product packaging data using the packaging design sealing performance data to obtain qualified product packaging design data, and visually displaying the qualified product packaging design data online includes:

[0097] Based on the packaging design sealing performance data, the exposure path of the simulated product packaging data is tracked to obtain the exposure path distribution data;

[0098] Performing local sealing enhancement on the exposure path data to generate local sealing enhancement data;

[0099] Perform design adjustments on the simulated product packaging data based on the local sealing reinforcement data, thereby generating reinforced sealing packaging data;

[0100] Perform packaging retrospective optimization based on enhanced sealing packaging data to obtain qualified product packaging design data;

[0101] Visually display qualified product packaging design data online.

[0102] In this embodiment, in the implementation process of exposure path tracing of simulated product packaging data based on the sealing performance data of the packaging design, the three-dimensional structural model of the simulated product packaging and the corresponding sealing performance data matrix are first imported. The path tracing algorithm is used to simulate the potential leakage paths of gas molecules or liquid molecules from the inside of the package to the outside in three-dimensional space. The initial molecular leakage point is set as the voxel unit with a sealing performance score below the threshold of 0.6. The Monte Carlo random sampling technique is used to iteratively track the leakage path. The number of iterations is set to 100,000. During the path tracing process, the energy loss and diffusion probability of each path are calculated to evaluate the leakage risk intensity. The tracing results generate exposure path distribution data, which includes the starting and ending coordinates of the path, the path length, the leakage probability, and the voxel area covered by the path. The exposure path distribution data is stored in the form of a sparse matrix to facilitate the subsequent local sealing reinforcement step. The specific operation of local sealing reinforcement of the exposure path data is to first extract the area where the 10% paths with the highest leakage probability are located as the target reinforcement area based on the exposure path distribution data. Based on the three-dimensional coordinate information of the target area, the predefined sealing reinforcement unit in the material reinforcement library is called, including high-density polyethylene Sealing strips, two-component epoxy resin adhesive, and high-elastic sealing foam were deformed and adjusted to the local geometric shape of the packaging structure corresponding to the target area using a computer-aided design (CAD) system. The local sealing thickness was increased by no less than 0.8 mm. The bonding strength of the sealing material was set to 12 MPa, and the elastic modulus was set to 0.25 GPa. Finite element analysis (FEA) software was then used to simulate the stress distribution and airtightness improvement of the local sealing structure. By comparing the structural deformation and stress-strain curves before and after reinforcement, the effective load-bearing range of the local reinforcement measures was confirmed. Finally, local sealing reinforcement data was generated, including updated 3D model information of the reinforced area, material property parameters, and simulation verification results. The operational process for designing and adjusting the simulated product packaging data based on the local sealing reinforcement data includes first importing the local sealing reinforcement data into the packaging design 3D model, calling the parametric design module to perform automated geometric modifications to the local structure, including increasing the edge sealing profile thickness to 1.The thickness of the sealing reinforcement material is evenly laid, and the length and spacing of the reinforcement ribs are adjusted to 15 mm and 20 mm respectively. During the design adjustment process, the rule-based geometric transformation algorithm is applied. Its core is to adjust the model grid structure by controlling the point displacement matrix. After the adjustment, the inverse mapping function is used to synchronize the changes to the overall packaging model data to generate a reinforced sealing packaging data file containing all design adjustment information. This file records the history of design parameter changes and the correspondence between the three-dimensional models. The data format is compatible with STEP and IGES standards to ensure that subsequent data transmission is accurate. The implementation details of packaging backtracking optimization based on the reinforced sealing packaging data are reflected in the application of multiple rounds of iterative optimization algorithms. First, the reinforced sealing packaging data is imported into the multi-objective genetic algorithm (Multi-objective Genetic The optimization objectives include maximizing the sealing performance score and minimizing the material usage. The initial population size is 200 design individuals. The genetic operators include a single-point crossover probability of 0.7 and a mutation probability of 0.03. The fitness function is calculated by weighted comprehensive calculation of the sealing performance data generated by simulation tests and material costs. After 50 generations of evolutionary calculations, the top 10% optimal design solutions are screened out. The structural parameters of the optimal design solution are then fine-tuned using the local gradient descent algorithm to complete the final packaging design data correction. The qualified product packaging design data is output in the form of a standard 3D CAD model with a complete structural parameter description file to ensure that the design data meets all process and material usage restrictions. The qualified product packaging design data is then processed online. For visualization, the design data is first imported into the cloud-based display platform via a 3D rendering engine. This rendering engine utilizes a real-time rendering module based on ray tracing technology, supporting high-precision texture mapping and global illumination calculations. The rendering resolution is set to 1920×1080, and the frame rate is maintained at 60 frames per second. The rendering results are presented in the browser using a WebGL interface, allowing users to rotate, zoom, and perform partial sectioning operations using a mouse or touch device. The display module supports transparency adjustment for multi-layer structures and highlights sealed and reinforced areas. It also incorporates a physical property shading algorithm to realistically represent the gloss and roughness of different materials. The system backend responds to user interaction requests in real time, dynamically loading detailed textures, ultimately achieving a complete online interactive visualization of qualified product packaging design data.

[0103] Of particular importance is the exposure path tracing of simulated product packaging data based on packaging design sealing performance data, including:

[0104] Perform sealing performance threshold mapping on the packaging design sealing performance data to obtain sealing performance threshold field data;

[0105] Extract the interface penetration nodes from the sealing performance threshold field data and the simulated product packaging data to generate penetration node distribution data;

[0106] Deconstruct the path connectivity of the penetration node distribution data to obtain the connectivity path framework;

[0107] The exposure path is tracked based on the connected path framework, and the path distribution of the exposure path is analyzed based on the connected path framework to generate exposure path distribution data.

[0108] In this embodiment, the implementation details of the sealing performance threshold mapping based on the sealing performance data of the packaging design are as follows: first, the sealing performance data matrix obtained in the packaging design is imported into the computing environment. The data matrix is ​​in voxels, and each voxel corresponds to a sealing performance score. The score range is set to 0 to 1, where the lower the value, the worse the sealing performance. The sealing performance data is binarized using the threshold mapping algorithm, and the threshold is set to 0.7. Voxels with scores lower than this value are marked as weak sealing performance areas. The generated sealing performance threshold field data is a three-dimensional Boolean matrix. The Boolean value "true" indicates an area with insufficient sealing performance, and "false" indicates an area with good sealing performance. The data structure storage adopts a sparse matrix format to save memory. The voxel coordinate index is used to establish a rapid retrieval system to extract the specific operations of the interface penetration nodes in the sealing performance threshold field data and the simulated product packaging data, including first mapping the sealing performance threshold field data to the three-dimensional geometric grid of the packaging data. The packaging data grid is composed of nodes (Node) and elements. The grid boundary detection algorithm is used to identify the interface between all the weak sealing performance voxels and the packaging structure surface grid. The penetration node is determined for the node set of the interface. The penetration node is defined as the grid node that is connected to the weak sealing performance area with the external environment. The node determination algorithm detects the node connection status through the adjacency matrix, screens out all the node sets that are directly or indirectly connected to the external space, and generates the penetration node distribution data. The data contains node coordinates, node connection relationships, and mapping relationships with sealing performance scores. The data format is a node list plus an adjacency list to ensure the efficient operation of subsequent connectivity analysis. The specific implementation process of path connectivity deconstruction of penetration node distribution data adopts the connected subgraph segmentation method in graph theory. First, the penetration node distribution data is converted into an undirected graph structure. The nodes correspond to the vertices of the graph, and the connection relationships between nodes correspond to the edges of the graph. The edge weight is calculated based on the spatial distance between the nodes and the sealing performance score. The edge weight calculation adopts the weighted Euclidean distance. The weight is defined as the distance between the nodes multiplied by the average of the sealing performance scores of the two nodes. The undirected graph is traversed using the depth-first search (DFS) or breadth-first search (BFS) algorithm. Identify all connected subgraphs and divide them into several connected path frames. Each connected path frame describes the reachable path structure between penetrating nodes. The connected path frame data includes the path starting point, end point, path length and path coverage node list. This data is stored in a joint structure of graph adjacency matrix and path linked list to ensure the structural integrity and fast query capability of path analysis. The operation steps of tracking exposure paths and analyzing the path distribution of exposure paths based on the connected path framework are as follows: first, apply the path search algorithm to each connected path frame, use the A* heuristic algorithm combined with the sealing performance score as the heuristic function, and perform the shortest path tracing calculation on the path. The heuristic function is defined as the straight-line distance from the current node to the target node divided by the weighted average of the sealing performance scores of the path nodes.During the search process, the path node sequence and the cumulative risk value of the path are recorded. The path risk value is the inverse of the sum of the sealing performance scores of the nodes in the path. The lower the cumulative score, the higher the risk. After all paths are searched, the path dataset is input into the path statistical analysis module to calculate the path frequency, path length distribution, and path risk distribution. The exposure path distribution data is generated. The data contains the spatial coordinate sequence of multiple paths, risk scores, and path weight information. The output format is a multidimensional array and path attribute table. This ensures that the spatial and attribute information of the exposure path accurately reflects the risk exposure distribution of weak areas of the packaging design sealing performance.

[0109] The present invention also provides a product packaging online design and display system for executing the product packaging online design and display method described above. The product packaging online design and display system includes:

[0110] The product modeling module is used to obtain the packaging target product data; determine the product outline based on the packaging target product data; and reconstruct the simulated product frame to be packaged based on the product outline;

[0111] The requirements mapping module is used to obtain packaging tightness requirements; estimate packaging usage scenarios based on packaging tightness requirements; and perform process constraint mapping based on the estimated packaging usage scenarios and packaging tightness requirements to generate packaging process constraint data;

[0112] The constraint fusion module is used to analyze the physical packaging constraints based on the simulated product framework to be packaged; the product packaging constraints are integrated according to the physical packaging constraints and the packaging process constraint data to obtain the packaging constraint data;

[0113] The packaging simulation module is used to simulate the packaging process of the simulated product framework to be packaged based on the packaging constraint data to generate simulated product packaging data; and to construct a packaging requirement reverse environment field according to the packaging airtightness requirements;

[0114] The sealing optimization module is used to re-test the sealing performance of simulated product packaging data based on the reverse environmental field of packaging requirements, thereby generating packaging design sealing performance data; the simulated product packaging data is optimized through the packaging design sealing performance data to obtain qualified product packaging design data, and the qualified product packaging design data is visualized online.

[0115] Through the application of the product modeling module, the present invention enables the system to accurately obtain the packaging target product data and determine its outer contour. The reconstructed simulated product framework to be packaged provides a clear basis for subsequent design. The packaging airtightness requirements obtained by the demand mapping module are combined with the estimated packaging usage scenarios to formulate reasonable environmental adaptability for the packaging design. The generated packaging process constraint data ensures that the design meets the actual operation requirements. The constraint fusion module can fully understand the physical limitations of the packaging process by analyzing the physical packaging constraints. The product packaging constraint fusion based on these constraints forms systematic packaging constraint data. The packaging simulation module simulates the packaging process of the simulated product framework to be packaged based on the packaging constraint data, and can display the dynamic changes in the packaging process in real time. The generated simulated product packaging data provides an important reference for design improvement. The construction of the reverse environmental field of packaging requirements provides a scientific basis for the re-testing of the sealing performance. The re-testing of the sealing performance of the simulated product packaging data based on the environmental field can accurately The airtightness of the packaging design is evaluated, and the generated packaging design airtightness performance data provides data support for subsequent optimization. By performing packaging retrospective optimization on the simulated product packaging data, the qualification of the final product packaging design is ensured. The online visual display of qualified product packaging design data not only improves the transparency of the design, but also facilitates all parties involved to monitor and evaluate the design effect in real time. The overall system has significantly improved the efficiency of packaging design through the application of digital technology, enhanced the flexibility and adaptability of design, laid the foundation for the intelligent transformation of the modern packaging industry, promoted the digitalization and visualization process of packaging design, promoted the rational use of resources and cost control, improved the company's market competitiveness and customer satisfaction, and ensured the efficient response and rapid iteration of packaging design under diversified needs. The integrated design of the system improves the collaboration efficiency between modules, reduces errors in the design process, enhances the overall quality and reliability of product packaging, and ultimately realizes the intelligence, automation and efficiency of packaging design.

[0116] The present invention is therefore intended to be illustrative and non-restrictive in all respects, with the scope of the invention being defined by the appended claims rather than the foregoing description, and all changes that come within the meaning and range of equivalents of the application documents are intended to be embraced therein.

[0117] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily 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 is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A product packaging online design and display method, characterized in that: The following steps are involved: Step S1: Obtaining packaging target product data; determining the product outer contour according to the packaging target product data; Reconstruct the simulated frame of the product to be packaged based on the outer contour of the product; Step S2: Obtaining packaging airtightness requirements; Estimate packaging usage scenarios based on packaging airtightness requirements; Perform process constraint mapping based on the estimated packaging usage scenarios and packaging tightness requirements to generate packaging process constraint data; Step S3: Analyze the physical packaging constraints based on the simulated product framework to be packaged; fuse the product packaging constraints according to the physical packaging constraints and the packaging process constraint data to obtain packaging constraint data; Step S4: simulating the packaging process of the simulated product framework to be packaged based on the packaging constraint data to generate simulated product packaging data; and constructing a packaging requirement reverse environment field according to the packaging airtightness requirement; Step S5: retest the sealing performance of the simulated product packaging data based on the reverse environmental field of the packaging requirements, thereby generating packaging design sealing performance data; perform packaging retrospective optimization on the simulated product packaging data through the packaging design sealing performance data, thereby obtaining qualified product packaging design data, and visually display the qualified product packaging design data online.

2. The product packaging online design and display method according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: Acquire packaging target product data; perform structured analysis on the packaging target product data to obtain product structured data; Step S12: extracting multi-dimensional product structure parameters from the product structured data, wherein the size range of the multi-dimensional product structure parameters is limited to a length, width, and height range of 10 mm to 5000 mm; Step S13: calibrating the product outer contour according to the multi-dimensional product structural parameters, wherein the outer contour error tolerance is controlled within ±0.5mm; Step S14: extracting the principal axis vector data of the product outer contour, wherein the length range of the principal axis vector is limited to 80% of the longest side of the product; Step S15: Reconstruct the product structure skeleton according to the principal axis vector data, wherein the skeleton node spacing is limited to 5mm-50mm; Step S16: Perform product frame simulation based on the product structure skeleton, thereby obtaining a simulated product frame to be packaged.

3. The product packaging online design and display method according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: Obtain target product usage data; determine its packaging airtightness requirements based on the target product usage data, wherein the target minimum gas leakage rate in the airtightness requirements is limited to no more than 0.005 L / min; Step S22: Mapping usage scenarios based on the packaging airtightness requirements and target product usage data to generate estimated packaging usage scenarios; Step S23: Determine a packaging structure strength threshold based on the estimated packaging usage scenario and packaging airtightness requirements, wherein the determined static compressive strength threshold is limited to 50-800 kPa; Step S24: performing packaging process constraint mapping based on the packaging structure strength threshold, wherein the maximum load deformation limit in the constraint mapping is controlled to be no more than 3% of the short side dimension of the product, so as to generate packaging process constraint data.

4. The product packaging online design and display method according to claim 1, characterized in that: In step S3, analyzing the physical packaging constraints based on the simulated product framework to be packaged includes: Perform spatial topological projection on the simulated product frame to be packaged, thereby obtaining spatial topological projection data; Identify structural node feature data in spatial topological projection data; Identify and simulate the protruding node areas in the frame of the product to be packaged based on the structural node feature data; The irregular area frame and the regular area frame in the frame of the product to be packaged are simulated based on the prominent node area segmentation; Performing entity packaging simulation on the irregular area framework to obtain irregular area packaging simulation data; Performing entity packaging simulation on the regular area framework to generate regular area packaging simulation data; Inferring the impact data of packaging area morphology based on irregular area packaging simulation data and regular area packaging simulation data; Map the packaging area morphology impact data to the packaging area morphology impact data.

5. The product packaging online design and display method according to claim 1, characterized in that: In step S3, product packaging constraint fusion is performed based on the physical packaging constraint and packaging process constraint data, including: Labeling entity packaging constraint data with constraint attributes to obtain packaging constraint label data; Decompose the packaging process constraint data into process nodes to generate process node constraint data; Perform interactive correlation mapping on the packaging constraint label data and the process node constraint data to obtain label node mapping data; The packaging constraint label data and the process node constraint data are constrained and integrated based on the label node mapping data, thereby obtaining the packaging constraint data.

6. The product packaging online design and display method according to claim 1, characterized in that: The step S4 of simulating the packaging process of the simulated product frame to be packaged based on the packaging constraint data includes: Enumerate the packaging paths of the simulated product framework to be packaged, thereby obtaining a collection of candidate product packaging processes; Based on the packaging constraint data, the path stability of the candidate product packaging process set is evaluated to generate path characteristic parameters; Performing adaptive path screening on the candidate product packaging process collection according to the path characteristic parameters to generate an adaptive packaging process; By adapting the packaging process, the packaging process of the simulated product framework to be packaged is simulated to generate simulated product packaging data.

7. The product packaging online design and display method according to claim 1, characterized in that: In step S4, the reverse environment field of packaging requirements is constructed according to the packaging airtightness requirements, including: Screen and estimate suitable packaging usage scenario data in the estimated packaging usage scenarios based on packaging airtightness requirements; Identify the adaptive scenario features in the packaging usage scenario data that meet the packaging airtightness requirements; Direction adjustment simulation is performed based on the adapted scene features, thereby generating unadapted scene features; Construct a reverse environment field for packaging requirements based on the characteristics of unsuitable scenarios.

8. The product packaging online design and display method according to claim 1, characterized in that: In step S5, retesting the sealing performance of the simulated product packaging data based on the reverse environmental field of the packaging requirements includes: Based on the reverse environment field of packaging requirements, the simulated product packaging data is scene-coupled and projected to generate reverse environment loading data; Conduct airtightness test on simulated product packaging data according to reverse environment loading data to obtain reverse environment airtightness performance; Re-determine the packaging design sealing performance data of the simulated product packaging data based on the reverse environment sealing performance.

9. The product packaging online design and display method according to claim 1, characterized in that: In step S5, the simulated product packaging data is optimized retrospectively by using the packaging design sealing performance data to obtain qualified product packaging design data, and the qualified product packaging design data is displayed online visually, including: Based on the packaging design sealing performance data, the exposure path of the simulated product packaging data is tracked to obtain the exposure path distribution data; Performing local sealing enhancement on the exposure path data to generate local sealing enhancement data; Perform design adjustments on the simulated product packaging data based on the local sealing reinforcement data, thereby generating reinforced sealing packaging data; Perform packaging retrospective optimization based on enhanced sealing packaging data to obtain qualified product packaging design data; Visually display qualified product packaging design data online.

10. A product packaging online design and display system, characterized in that: For executing the product packaging online design and display method according to claim 1, the product packaging online design and display system comprises: The product modeling module is used to obtain the packaging target product data; determine the product outline based on the packaging target product data; and reconstruct the simulated product frame to be packaged based on the product outline; The requirements mapping module is used to obtain packaging tightness requirements; estimate packaging usage scenarios based on packaging tightness requirements; and perform process constraint mapping based on the estimated packaging usage scenarios and packaging tightness requirements to generate packaging process constraint data; The constraint fusion module is used to analyze the physical packaging constraints based on the simulated product framework to be packaged; the product packaging constraints are integrated according to the physical packaging constraints and the packaging process constraint data to obtain the packaging constraint data; The packaging simulation module is used to simulate the packaging process of the simulated product framework to be packaged based on the packaging constraint data to generate simulated product packaging data; and to construct a packaging requirement reverse environment field according to the packaging airtightness requirements; The sealing optimization module is used to re-test the sealing performance of simulated product packaging data based on the reverse environmental field of packaging requirements, thereby generating packaging design sealing performance data; the simulated product packaging data is optimized through the packaging design sealing performance data to obtain qualified product packaging design data, and the qualified product packaging design data is visualized online.

Images