Intelligent packaging paper container forming parameter modeling method and system

Abstract

The present application relates to the field of packaging paper container forming, in particular to a smart packaging paper container forming parameter modeling method and system, the method first divides the surface into multiple physical partitions according to the mechanical properties of deep drawing forming of the paper container, and establishes the mapping relationship between the partitions and the 2D image; then the paper container surface image is collected and preprocessed, the wrinkle characteristic parameters of each partition are extracted, and a comprehensive wrinkle quality index is constructed; then based on the material mechanics and plastic forming theory, a physical mechanism model of process parameters and wrinkle characteristics is established; then taking the minimization of the comprehensive wrinkle quality index as the target, a forming parameter optimization model is established and solved to obtain the optimal parameter combination; finally, the model is verified and corrected through experiments. The present application solves the problems that the prior art cannot comprehensively describe the wrinkle spatial distribution, lacks physical interpretability, and cannot realize intelligent optimization, and is suitable for deep drawing forming processes of various packaging paper containers.

Description

Technical Field

[0001] This invention belongs to the field of packaging paper container forming, specifically involving a method and system for modeling intelligent packaging paper container forming parameters. Background Technology

[0002] Paper containers are a widely used packaging form in the food and daily necessities industries, and their forming quality directly affects the product's performance and market competitiveness. Deep drawing is the core process in paper container production, where a flat paper blank is stretched into a three-dimensional container with a certain depth using a mold. During deep drawing, the paper blank undergoes plastic deformation under the combined action of various forces such as edge clamping force, drawing force, and friction, inevitably resulting in wrinkles. The presence of wrinkles reduces the geometric accuracy, compressive strength, and appearance quality of the paper container, and in severe cases, can even lead to product scrap. Therefore, effective control and quantitative evaluation of wrinkles during the paper container forming process are crucial for improving the quality and efficiency of paper container production.

[0003] Currently, the quality assessment of wrinkles in paper containers mainly employs traditional scalar quality indicators, such as dimensional deviation, maximum wrinkle height, and number of wrinkles. These indicators, obtained through manual measurement or simple image analysis methods, can reflect the severity of wrinkles to a certain extent. For example, measuring the diameter deviation at the mouth of the paper container can assess the uniformity of overall deformation during the forming process; measuring the maximum wrinkle height can determine the severity of the wrinkles. These methods are simple to operate and low in cost, and have been widely used in industrial production.

[0004] With the development of computer vision technology, some researchers have begun to explore more advanced image analysis methods to quantify wrinkle features. Among these, the Fast Fourier Transform (FFT) is a commonly used method. This method converts a 2D image of the paper container surface into a frequency domain signal, and quantifies the frequency characteristics of the wrinkles by analyzing the energy distribution in the frequency domain. For example, high-frequency components correspond to fine wrinkles, while low-frequency components correspond to coarse wrinkles. By calculating the energy values ​​in different frequency bands, the overall frequency characteristics of the wrinkles can be obtained, thereby assessing the wrinkle quality.

[0005] Furthermore, researchers have applied machine learning techniques to predict the forming quality of paper containers. They trained machine learning models by collecting a large amount of process parameters and corresponding crease quality data, establishing a mapping relationship between process parameters and crease characteristics. Using the trained model, the crease quality after forming can be predicted based on the input process parameters, thus providing a basis for optimizing process parameters. These methods have improved the efficiency and accuracy of crease quality assessment to some extent, laying the foundation for the intelligent development of paper container forming processes.

[0006] Although existing technologies have made some progress in assessing the wrinkle quality of paper containers and optimizing process parameters, the following shortcomings and deficiencies still exist: First, traditional scalar quality indicators cannot fully describe the spatial distribution characteristics of wrinkles. During the deep drawing process of paper containers, wrinkle formation exhibits significant spatial non-uniformity. Wrinkles at different locations have varying degrees of impact on product quality. For example, wrinkles in the middle of the sidewall of a paper container severely affect the product's compressive strength, while wrinkles at the bottom edge have a relatively smaller impact on the product's performance. Traditional scalar quality indicators, such as dimensional deviations and maximum wrinkle height, can only reflect the overall severity of wrinkles or the most severe local cases, failing to differentiate the impact of wrinkles at different spatial locations on quality. This leads to a significant discrepancy between quality assessment results and the actual performance of the product, failing to provide accurate guidance for optimizing process parameters.

[0007] Second, the wrinkle quantization method based on Fast Fourier Transform (FFT) loses spatial location information. FFT is a global frequency domain analysis method that converts the pixel values ​​of the entire image into a frequency domain signal, failing to preserve the spatial location information of wrinkles. Therefore, this method can only obtain the overall frequency characteristics of wrinkles and cannot distinguish the frequency differences of wrinkles at different spatial locations. For example, when wrinkles of different frequencies appear on the left and right sides of a paper container, FFT can only obtain an average frequency characteristic, unable to describe the characteristics of wrinkles on both sides separately. This makes it impossible for this method to accurately assess the impact of the spatial distribution of wrinkles on product quality, limiting its application in actual production.

[0008] Third, traditional machine learning models can only establish the correlation between process parameters and wrinkle features, but cannot reveal the physical mechanisms. Traditional machine learning models, such as linear regression, support vector machines, and decision trees, are data-driven methods. They establish statistical correlations between process parameters and wrinkle features by fitting large amounts of training data. However, these models are "black box" models and cannot explain the intrinsic physical mechanisms between process parameters and wrinkle features. In other words, they can only tell us "what changes occur in wrinkle features when a certain parameter is adjusted," but cannot explain "why adjusting this parameter leads to such changes." This makes the optimization of process parameters lack theoretical guidance, often requiring extensive trial-and-error experiments to find the optimal parameters, which not only increases production costs but also prolongs product development cycles.

[0009] Fourth, existing methods cannot achieve intelligent optimization of molding parameters. Because they cannot fully describe the spatial distribution characteristics of wrinkles, nor reveal the physical mechanisms between process parameters and wrinkle features, existing methods cannot establish accurate molding parameter models. Therefore, they cannot automatically calculate the optimal combination of molding parameters based on product quality requirements. In actual production, the adjustment of process parameters still mainly relies on the operator's experience, leading to poor product quality stability and low production efficiency.

[0010] In summary, existing technologies have significant shortcomings and deficiencies in assessing the crease quality of paper containers and optimizing process parameters, failing to meet the demands of modern packaging industries for high-quality, high-efficiency, and intelligent production. Therefore, there is an urgent need to develop a method that can comprehensively describe the spatial distribution characteristics of creases, reveal their physical mechanisms, and achieve intelligent optimization of forming parameters. Summary of the Invention

[0011] The purpose of this invention is to provide a method for modeling the forming parameters of intelligent packaging paper containers, and at the same time, to provide a system for modeling the forming parameters of intelligent packaging paper containers, so as to solve the problems mentioned in the background art.

[0012] To solve the above-mentioned technical problems, the present invention provides the following technical solution: A method for modeling the forming parameters of intelligent packaging paper containers, including the following steps: Based on the mechanical properties and quality influence weight of deep drawing of paper containers, the surface of the paper container is divided into multiple physical partitions, and the mapping relationship between each physical partition and the 2D image of the paper container is established. 2D images of the paper container surface are acquired and then processed sequentially using grayscale conversion, median filtering for noise reduction, homomorphic filtering for illumination correction, and Sobel operator edge enhancement. Extract wrinkle feature parameters from the image regions corresponding to each physical partition, and construct wrinkle feature vectors for each partition; The comprehensive fold quality index is calculated by combining the quality impact weight and feature weight of each physical partition; Based on the theories of materials mechanics and plastic forming, a physical mechanism model of the relationship between process parameters and wrinkle characteristics in each physical zone is established. With the goal of minimizing the overall wrinkle quality index, a molding parameter optimization model that includes the range of process parameter values ​​and product performance constraints is established and solved. The accuracy of the model was verified through experiments, and the model parameters were corrected using the least squares method.

[0013] Furthermore, the physical partitions include a flange area, an upper sidewall area, a middle sidewall area, a lower sidewall area, and a bottom area; wherein the flange area is the annular area at the edge of the paper container opening, the upper sidewall area is the area from below the flange area to 1 / 3 of the container height, the middle sidewall area is the area from 1 / 3 to 2 / 3 of the container height, the lower sidewall area is the area from 2 / 3 of the container height to the bottom rounded corner, and the bottom area is the planar area below the bottom rounded corner.

[0014] Furthermore, the wrinkle feature parameters include wrinkle area ratio, average wrinkle height, wrinkle density, wrinkle direction entropy, and wrinkle length distribution characteristics; wherein the wrinkle area ratio is the ratio of the wrinkle area to the total area of ​​the partition, the average wrinkle height is calculated through the calibration relationship between the image grayscale value and the actual height, the wrinkle density is the number of wrinkles per unit area, the wrinkle direction entropy is calculated through the distribution histogram of the main wrinkle directions, and the wrinkle length distribution characteristics include the average wrinkle length, the standard deviation of the length, and the maximum length.

[0015] Furthermore, the calculation process of the comprehensive fold quality index is as follows: the minimum-maximum normalization method is used to convert each fold feature parameter into a dimensionless value in the interval [0,1]; the partition quality index of each physical partition is calculated according to the feature weight; the partition quality index and the corresponding quality influence weight are weighted and summed to obtain the comprehensive fold quality index.

[0016] Furthermore, the process of establishing the physical mechanism model is as follows: analyzing the influence of each key process parameter on the stress and strain state of each physical zone; establishing the mathematical relationship between process parameters and stress and strain of each physical zone; establishing the mathematical relationship between stress and strain and wrinkle characteristics based on the material instability theory; and combining the above two types of mathematical relationships to obtain the physical mechanism model.

[0017] Furthermore, the sequential quadratic programming method is used to solve the optimization model. The specific process is as follows: given the initial point, convergence accuracy, and maximum number of iterations; calculate the objective function value, gradient, constraint function value, and Jacobian matrix at the initial point; construct a quadratic programming subproblem and solve it to obtain the search direction; perform a one-dimensional search to determine the step size; update the iteration point; determine whether the convergence condition is met. If it is met, output the optimal solution; otherwise, continue iterating.

[0018] Furthermore, the model verification and correction process is as follows: design a verification experimental scheme that includes key process parameters at different levels; conduct paper container forming experiments and collect sample images; calculate the comprehensive wrinkle quality index of the experiment; input the process parameters into the model to obtain the prediction results; compare the prediction results with the experimental results, analyze the sources of error and correct the model parameters; repeat the verification and correction process until the model prediction accuracy meets the requirements.

[0019] Intelligent packaging paper container forming parameter modeling system, including The physical partitioning module is used to divide the surface of the paper container into multiple physical partitions based on the mechanical properties and quality influence weight of the deep drawing process, and to establish a mapping relationship between each physical partition and the 2D image of the paper container. The image acquisition and preprocessing module is used to acquire 2D images of the paper container surface and sequentially perform grayscale conversion, median filtering for noise reduction, homomorphic filtering for illumination correction, and Sobel operator edge enhancement processing to obtain an image with enhanced edges. The wrinkle feature extraction module is used to extract wrinkle feature parameters from the image regions corresponding to each physical partition and construct wrinkle feature vectors for each partition. The quality index calculation module is used to calculate the comprehensive wrinkle quality index by combining the quality influence weight and feature weight of each physical partition. The physical mechanism modeling module is used to establish physical mechanism models of process parameters and wrinkle characteristics of each physical zone based on materials mechanics and plastic forming theory. The molding parameter optimization module is used to establish and solve a molding parameter optimization model that includes the range of process parameter values ​​and product performance constraints with the goal of minimizing the comprehensive wrinkle quality index, so as to obtain the optimal combination of molding parameters. The model validation and correction module is used to verify the accuracy of the model through experiments and correct the model parameters using the least squares method.

[0020] This application also discloses an electronic device, including: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the above-described intelligent packaging paper container forming parameter modeling method of the present invention.

[0021] This application also discloses a computer-readable storage medium storing computer-executable instructions for causing a computer to execute the intelligent packaging paper container forming parameter modeling method.

[0022] Beneficial effects: First, this invention establishes a quantitative method for the spatial distribution of wrinkles based on physical partitioning, which can comprehensively describe the spatial distribution characteristics of wrinkles. By dividing the surface of the paper container into regions with different physical meanings, wrinkle characteristic parameters of each region are extracted, and a comprehensive wrinkle quality index is constructed by combining quality influence weights. This solves the problem that traditional scalar quality indices cannot distinguish the quality impact of wrinkles in different spatial locations. This method can accurately reflect the impact of the spatial distribution of wrinkles on the geometric accuracy, compressive strength, and appearance quality of the product, providing a reliable basis for the comprehensive evaluation of the forming quality of paper containers.

[0023] Secondly, this invention constructs a physical mechanism model of process parameters and wrinkle characteristics based on mechanical principles, which can reveal the intrinsic causal relationship between process parameters and wrinkle distribution. Based on materials mechanics and plastic forming theory, this model establishes the mathematical relationship between process parameters and stress-strain in each physical region. Furthermore, it combines materials instability theory to establish the mathematical relationship between stress-strain and wrinkle characteristics, ultimately forming a complete physical mapping model between process parameters and wrinkle characteristics. Compared with traditional "black box" machine learning models, this model has good interpretability, clearly explaining "why adjusting a certain parameter leads to a change in wrinkle distribution," providing theoretical guidance for optimizing process parameters, reducing the number of trial-and-error experiments, lowering production costs, and shortening product development cycles.

[0024] Third, this invention develops a molding parameter optimization algorithm based on a physical model, enabling intelligent optimization of molding parameters. By establishing a molding parameter optimization model with the objective of minimizing the overall wrinkle quality index, using key process parameters as design variables, and incorporating the range of process parameter values ​​and product performance constraints, and solving the model using sequential quadratic programming or a genetic algorithm, the optimal combination of molding parameters that meets quality requirements can be automatically calculated. This algorithm solves the problem that existing methods cannot achieve intelligent optimization of molding parameters, eliminates reliance on operator experience, and improves product quality stability and production efficiency.

[0025] Fourth, this invention establishes a complete intelligent modeling process for paper container forming parameters, providing technical support for the intelligent development of paper container forming technology. This process is highly versatile and operable, applicable not only to the deep-drawing forming process of common paper containers such as round paper cups and bowls, but also, by adjusting the physical partitioning method and feature parameter extraction method, to be extended to the forming processes of special-shaped paper containers such as squares and irregular shapes, as well as the plastic forming processes of other materials such as plastics and metals, demonstrating broad application prospects. Attached Figure Description

[0026] Figure 1 This is a flowchart of the intelligent packaging paper container forming parameter modeling method of the present invention; Figure 2 is a schematic diagram of the wrinkle feature extraction process in an embodiment of the present invention; Figure 3 is a schematic diagram of the process for constructing the comprehensive wrinkle quality index in an embodiment of the present invention. Figure 4 is a schematic diagram of the process parameter-fold feature physical mechanism model establishment process in an embodiment of the present invention; Figure 5 is a schematic diagram of the solution process of the molding parameter optimization model in an embodiment of the present invention. Detailed Implementation

[0027] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0028] This embodiment is applicable to the modeling and optimization of deep-drawing process parameters for various types of packaging paper containers. This method can be executed by an intelligent forming device integrating an image acquisition system, a data processing unit, and a process control module. The "physical partitioning" mentioned in this invention refers to surface regions with different mechanical meanings, divided according to the material flow patterns and stress distribution characteristics during the paper container forming process; the "comprehensive wrinkle quality index" refers to a dimensionless quality evaluation parameter that comprehensively considers the spatial distribution and morphological characteristics of wrinkles.

[0029] The intelligent packaging paper container forming parameter modeling method provided by this invention specifically includes the following steps: S101: Based on the mechanical properties and quality influence weight of deep drawing of paper containers, the surface of the paper container is divided into multiple physical partitions, and the mapping relationship between each physical partition and the 2D image of the paper container is established. S102: Acquire a 2D image of the paper container surface and sequentially perform grayscale conversion, median filtering for noise reduction, homomorphic filtering for illumination correction, and Sobel operator edge enhancement processing to obtain the edge-enhanced image; S103: Extract wrinkle feature parameters from the image regions corresponding to each physical partition, and construct wrinkle feature vectors for each partition; S104: Calculate the comprehensive fold quality index by combining the quality influence weight and feature weight of each physical partition; S105: Based on the theory of materials mechanics and plastic forming, establish a physical mechanism model of process parameters and wrinkle characteristics of each physical zone; S106: With the goal of minimizing the overall wrinkle quality index, establish and solve a molding parameter optimization model that includes the range of process parameter values ​​and product performance constraints to obtain the optimal combination of molding parameters; S107: Verify the accuracy of the model through experiments and correct the model parameters using the least squares method.

[0030] Step S101: Physical partitioning and image mapping; In this embodiment, for a circular paper container, its surface is divided into five physical partitions: flange area, upper sidewall area, middle sidewall area, lower sidewall area, and bottom area. Specifically, the flange area is a 5mm wide annular region at the edge of the paper container opening; the upper sidewall area extends from below the flange area to 1 / 3 of the container height; the middle sidewall area extends from 1 / 3 to 2 / 3 of the container height; the lower sidewall area extends from 2 / 3 of the container height to the bottom rounded corner; and the bottom area is the planar region below the bottom rounded corner.

[0031] Specifically, through 50 sets of comparative experiments and theoretical analysis, the quality influence weights of each physical zone were determined as follows: flange zone w1=0.2, upper sidewall zone w2=0.25, middle sidewall zone w3=0.35, lower sidewall zone w4=0.15, and bottom zone w5=0.05, satisfying w1+w2+w3+w4+w5=1.

[0032] For example, when establishing the mapping relationship between physical partitions and 2D images, the Canny edge detection algorithm is first used to extract the inner and outer contours of the paper cup; then, the transformation ratio between the image coordinate system and the actual coordinate system is calculated based on the actual size of the paper cup and the image pixel size; finally, the five physical partitions are mapped onto the 2D image according to the transformation ratio to obtain the image region coordinates corresponding to each physical partition.

[0033] Optionally, for square paper containers, their surface is divided into nine physical zones: the flange zone, the four corner sidewall zones, the four side wall zones, and the bottom zone. The weight of each zone's quality influence is determined experimentally.

[0034] Step S102 Image Acquisition and Preprocessing; In this embodiment, an image acquisition system is constructed, including an industrial camera, a ring-shaped shadowless light source, a telecentric lens, and an image acquisition card. The formed paper container is placed on a black platform, and the camera position and focal length are adjusted to ensure that the surface of the paper container is completely within the camera's field of view and that the image is clear and distortion-free. A grayscale image of the paper container surface is then acquired.

[0035] Specifically, the acquired images undergo preprocessing, including: For grayscale conversion, if a color image is acquired, a weighted average method is used for grayscale conversion. The formula is: Gray = 0.299 × R + 0.587 × G + 0.114 × B; For filtering and noise reduction, a 3×3 window median filtering algorithm is used to eliminate salt-and-pepper noise and impulse noise in the image; Illumination correction processing employs a homomorphic filtering algorithm to eliminate the effects of uneven illumination. Parameters are set to cutoff frequency D0 = 10 and gain constant γ. H =2.0, γL =0.5; Edge enhancement processing uses the Sobel operator to enhance the edges of the image, highlighting the edge information of wrinkles.

[0036] Step S103: Wrinkle feature extraction; In this embodiment, the image after edge enhancement is binarized and morphologically processed, and then the wrinkle feature parameters of each physical partition are extracted.

[0037] Specifically, such as Figure 2 As shown, an adaptive thresholding method is used for binarization, with a window size of 15×15 and a constant C=2. A 3×3 rectangular structuring element is used, and opening and closing operations are performed first to eliminate small noise points and fill small holes inside the folds.

[0038] For example, the extracted wrinkle feature parameters include: Fold area ratio, the ratio of the number of fold pixels within a partition to the total number of pixels in the partition; The average height of the folds is calculated by calibrating the mapping relationship between grayscale values ​​and actual heights using standard height blocks. Fold density, the ratio of the number of folds in a zone to the area of ​​the zone; The fold orientation entropy is calculated based on the histogram of the distribution of the main fold orientations. The characteristics of fold length distribution include average fold length, standard deviation of length, and maximum length.

[0039] Optionally, the fold width distribution features, curvature distribution features, and spacing distribution features can also be extracted, and a more comprehensive fold feature vector can be constructed after dimensionality reduction through principal component analysis.

[0040] Step S104: Construction of comprehensive wrinkle quality index; in this embodiment, such as... Figure 3 As shown, the feature weights of each feature parameter in each partition are determined by expert scoring. After performing min-max normalization on each feature parameter, the partition quality index of each physical partition is calculated. Finally, the partition quality index and the corresponding quality influence weight are weighted and summed to obtain the comprehensive fold quality index.

[0041] Specifically, the minimum-maximum normalization formula is: x'=(xx) min ) / (x max -x min ), where x is the original eigenvalue, x min and x max These are the minimum and maximum values ​​of the feature parameter, respectively; the formula for the partition quality index is: The formula for the comprehensive wrinkle quality index is: .

[0042] Step S105 Physical mechanism modeling; in this embodiment, such as Figure 4 As shown, the key process parameters affecting the formation of wrinkles in paper containers are determined to be edge clamping force F, drawing speed v, die gap c, paper blank thickness t, and paper blank moisture content m.

[0043] Specifically, the process of establishing the physical mechanism model is as follows: Analyze the impact of each key process parameter on the stress-strain state of each physical zone; Establish the mathematical relationship between process parameters and stress-strain in each physical zone; Based on the theory of material instability, a mathematical relationship between stress and strain and wrinkle characteristics is established; Combining the two types of mathematical relationships mentioned above, we obtain a complete physical mechanism model: F i =f i (F,v,c,t,m), where F i Let be the wrinkle feature vector of the i-th physical partition.

[0044] For example, the compressive stress σ in the flange area f The relationship between σ and the blanking force F, die clearance c, and blank thickness t is: f =0.001×F / (π×(D²-d²) / 4)+0.5×c / t+0.1;Critical instability stress σ in the flange area cr The relationship between σ and the elastic modulus E of the paper blank, the paper blank thickness t, and the average radius R of the flange area is: cr =0.2×E×(t / R)²; Flange area ratio FAR1 and σ f and σ cr The relationship is: when σ f >σ cr At that time, FAR1 = 2 × (σ f -σ cr ) / σ cr When σ f ≤σ cr At that time, FAR1=0.

[0045] Step S106 Molding parameter optimization; such as Figure 5 As shown, in this embodiment, a molding parameter optimization model is established with the goal of minimizing the comprehensive fold quality index, and the sequential quadratic programming method is used to solve it.

[0046] Specifically, the optimization model is as follows: min CQI (F,v,c,t,m); st; 500N≤F≤2000N; 10mm / s≤v≤50mm / s; 0.2mm≤c≤0.5mm; 0.2mm≤t≤0.4mm; 6%≤m≤12%; P≥P min ; Where P is the compressive strength of the paper container, P min This is the minimum compressive strength required by the design.

[0047] For example, the solution process is as follows: given an initial point, convergence precision, and maximum number of iterations; calculate the objective function value, gradient, constraint function value, and Jacobian matrix at the initial point; construct a quadratic programming subproblem and solve it to obtain the search direction; perform a one-dimensional search to determine the step size; update the iteration point; determine whether the convergence condition is met, and if so, output the optimal solution; otherwise, continue iterating.

[0048] Optionally, a genetic algorithm can be used to solve the optimization model, which has the advantage of strong global search capability and can avoid getting trapped in local optima.

[0049] Step S107 Model Verification and Correction: In this embodiment, a verification experimental scheme containing key process parameters at different levels is designed, paper container forming experiments are conducted and sample images are collected, the comprehensive wrinkle quality index of the experiment is calculated, the process parameters are input into the model to obtain the prediction results, the prediction results are compared with the experimental results, the source of error is analyzed and the model parameters are corrected.

[0050] Specifically, the least squares method is used to correct the coefficients in the model, and the verification and correction process is repeated until the model's prediction accuracy meets the requirements of engineering applications.

[0051] This invention also provides an intelligent packaging paper container forming parameter modeling system, comprising: The physical partitioning module is used to divide the surface of the paper container into multiple physical partitions based on the mechanical properties and quality influence weight of the deep drawing process, and to establish a mapping relationship between each physical partition and the 2D image of the paper container. The image acquisition and preprocessing module is used to acquire 2D images of the paper container surface and sequentially perform grayscale conversion, median filtering for noise reduction, homomorphic filtering for illumination correction, and Sobel operator edge enhancement processing to obtain an image with enhanced edges. The wrinkle feature extraction module is used to extract wrinkle feature parameters from the image regions corresponding to each physical partition and construct wrinkle feature vectors for each partition. The quality index calculation module is used to calculate the comprehensive wrinkle quality index by combining the quality influence weight and feature weight of each physical partition. The physical mechanism modeling module is used to establish physical mechanism models of process parameters and wrinkle characteristics of each physical zone based on materials mechanics and plastic forming theory. The molding parameter optimization module is used to establish and solve a molding parameter optimization model that includes the range of process parameter values ​​and product performance constraints with the goal of minimizing the comprehensive wrinkle quality index, so as to obtain the optimal combination of molding parameters. The model validation and correction module is used to verify the accuracy of the model through experiments and correct the model parameters using the least squares method.

[0052] Furthermore, the physical partitioning module is specifically used to: for a circular paper container, divide its surface into five physical partitions: flange area, upper sidewall area, middle sidewall area, lower sidewall area, and bottom area; for a square paper container, divide its surface into nine physical partitions: flange area, four corner sidewall areas, four side sidewall areas, and bottom area.

[0053] Furthermore, the wrinkle feature extraction module is specifically used to: extract wrinkle area ratio, average wrinkle height, wrinkle density, wrinkle direction entropy, and wrinkle length distribution features, and construct wrinkle feature vectors for each partition.

[0054] Furthermore, the physical mechanism modeling module is specifically used to: analyze the influence of key process parameters on the stress and strain state of each physical zone, establish the mathematical relationship between process parameters and stress and strain, as well as the mathematical relationship between stress and strain and wrinkle characteristics, and finally form a complete physical mechanism model.

[0055] Furthermore, the molding parameter optimization module is specifically used to: solve the molding parameter optimization model using a sequential quadratic programming method or a genetic algorithm to obtain the optimal combination of molding parameters.

[0056] This invention also provides an electronic device, including at least one processor and a memory communicatively connected to the at least one processor. The memory stores a computer program executable by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the intelligent packaging paper container forming parameter modeling method described in any of the above method embodiments.

[0057] Specifically, the electronic device may further include input devices and output devices. The processor, memory, input devices, and output devices can be connected via a bus or other means. Memory, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules. The processor executes various functional applications and data processing of the electronic device by running the non-volatile software programs, instructions, and modules stored in the memory.

[0058] This invention also provides a computer-readable storage medium storing computer-executable instructions for causing a computer to execute the intelligent packaging paper container forming parameter modeling method described in any of the above method embodiments.

[0059] The computer-readable storage medium may include, but is not limited to, various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0060] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for modeling the forming parameters of intelligent packaging paper containers, characterized in that, Includes the following steps: Based on the mechanical properties and quality influence weight of deep drawing of paper containers, the surface of the paper container is divided into multiple physical partitions, and the mapping relationship between each physical partition and the 2D image of the paper container is established. 2D images of the paper container surface are acquired and then processed sequentially using grayscale conversion, median filtering for noise reduction, homomorphic filtering for illumination correction, and Sobel operator edge enhancement. Extract wrinkle feature parameters from the image regions corresponding to each physical partition, and construct wrinkle feature vectors for each partition; The comprehensive fold quality index is calculated by combining the quality impact weight and feature weight of each physical partition; Based on the theories of materials mechanics and plastic forming, a physical mechanism model of the relationship between process parameters and wrinkle characteristics in each physical zone is established. With the goal of minimizing the overall wrinkle quality index, a molding parameter optimization model that includes the range of process parameter values ​​and product performance constraints is established and solved. The accuracy of the model was verified through experiments, and the model parameters were corrected using the least squares method.

2. The intelligent packaging paper container forming parameter modeling method according to claim 1, characterized in that, The physical partitions include a flange area, an upper sidewall area, a middle sidewall area, a lower sidewall area, and a bottom area; wherein the flange area is the annular area at the edge of the paper container opening, the upper sidewall area is the area from below the flange area to 1 / 3 of the container height, the middle sidewall area is the area from 1 / 3 to 2 / 3 of the container height, the lower sidewall area is the area from 2 / 3 of the container height to the bottom rounded corner, and the bottom area is the planar area below the bottom rounded corner.

3. The intelligent packaging paper container forming parameter modeling method according to claim 1, characterized in that, The wrinkle feature parameters include wrinkle area ratio, average wrinkle height, wrinkle density, wrinkle direction entropy, and wrinkle length distribution characteristics; wherein the wrinkle area ratio is the ratio of the wrinkle area to the total area of ​​the partition, the average wrinkle height is calculated by the calibration relationship between the image gray value and the actual height, the wrinkle density is the number of wrinkles per unit area, the wrinkle direction entropy is calculated by the distribution histogram of the main wrinkle direction, and the wrinkle length distribution characteristics include the average wrinkle length, the standard deviation of the length, and the maximum length.

4. The intelligent packaging paper container forming parameter modeling method according to claim 1, characterized in that, The calculation process of the comprehensive fold quality index is as follows: the minimum-maximum normalization method is used to convert each fold characteristic parameter into a dimensionless value in the interval [0,1]. The partition quality index for each physical partition is calculated based on the feature weights; the partition quality index is then weighted and summed with the corresponding quality influence weights to obtain the comprehensive fold quality index.

5. The method for modeling the forming parameters of intelligent packaging paper containers according to claim 1, characterized in that, The process of establishing the physical mechanism model is as follows: analyze the influence of each key process parameter on the stress and strain state of each physical zone; establish the mathematical relationship between process parameters and stress and strain of each physical zone; establish the mathematical relationship between stress and strain and wrinkle characteristics based on the material instability theory; and combine the above two types of mathematical relationships to obtain the physical mechanism model.

6. The method for modeling the forming parameters of intelligent packaging paper containers according to claim 1, characterized in that, The sequential quadratic programming method is used to solve the optimization model. The specific process is as follows: given the initial point, convergence accuracy, and maximum number of iterations; calculate the objective function value, gradient, constraint function value, and Jacobian matrix at the initial point; construct the quadratic programming subproblem and solve it to obtain the search direction; perform a one-dimensional search to determine the step size; update the iteration point. Determine if the convergence condition is met. If it is, output the optimal solution; otherwise, continue iterating.

7. The method for modeling the forming parameters of intelligent packaging paper containers according to claim 1, characterized in that, The model verification and correction process is as follows: designing a verification experimental scheme that includes key process parameters at different levels; conducting paper container molding experiments and acquiring sample images; Calculate the overall wrinkle quality index of the experiment; input the process parameters into the model to obtain the prediction results; compare the prediction results with the experimental results, analyze the sources of error and correct the model parameters; repeat the verification and correction process until the model prediction accuracy meets the requirements.

8. A system utilizing the intelligent packaging paper container forming parameter modeling method according to claim 1, characterized in that, include: The physical partitioning module is used to divide the surface of the paper container into multiple physical partitions based on the mechanical properties and quality influence weight of the deep drawing process, and to establish a mapping relationship between each physical partition and the 2D image of the paper container. The image acquisition and preprocessing module is used to acquire 2D images of the paper container surface and sequentially perform grayscale conversion, median filtering for noise reduction, homomorphic filtering for illumination correction, and Sobel operator edge enhancement processing to obtain an image with enhanced edges. The wrinkle feature extraction module is used to extract wrinkle feature parameters from the image regions corresponding to each physical partition and construct wrinkle feature vectors for each partition. The quality index calculation module is used to calculate the comprehensive wrinkle quality index by combining the quality influence weight and feature weight of each physical partition. The physical mechanism modeling module is used to establish physical mechanism models of process parameters and wrinkle characteristics of each physical zone based on materials mechanics and plastic forming theory. The molding parameter optimization module is used to establish and solve a molding parameter optimization model that includes the range of process parameter values ​​and product performance constraints with the goal of minimizing the comprehensive wrinkle quality index, so as to obtain the optimal combination of molding parameters. The model validation and correction module is used to verify the accuracy of the model through experiments and correct the model parameters using the least squares method.

9. An electronic device, characterized in that, The device includes at least one processor and a memory communicatively connected to the at least one processor, the memory storing a computer program executable by the at least one processor, the computer program being executed by the at least one processor to cause the at least one processor to perform the intelligent packaging paper container forming parameter modeling method as described in any one of claims 1-10.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions for causing a computer to perform the intelligent packaging paper container forming parameter modeling method as described in any one of claims 1-10.

Images