Design apparatus and design method for multilayer films

Abstract

A design apparatus and method for multilayer films are disclosed. The design apparatus for multilayer films models the laminated structure of the multilayer film to be designed, collects cone strength data for single-layer films forming the laminated structure as stress-strain data, reads values ​​pre-stored in a memory space accessible to the design apparatus for multilayer films, obtains a feature setting mode to specify a different feature setting method according to the read value, sets multiple physical indicators selected from the cone strength data and the thickness of the single-layer film as features according to the feature setting mode, predicts the cone strength of the multilayer film using a machine learning model selected from multiple supervised learning models, with features as independent variables and the cone strength of the multilayer film as the target variable, and generates design data for the multilayer film by combining predicted values ​​for other properties with the predicted value of cone strength.

Description

Technical Field

[0001] [Cross - Reference to Related Applications] This application claims the benefit of priority based on Korean Patent Application No. 10 - 2023 - 0116487 filed on September 1, 2023 and Korean Patent Application No. 10 - 2024 - 0117595 filed on August 30, 2024, and all the contents disclosed in the documents of these Korean patent applications are included as part of this specification.

[0002] The disclosure relates to a device and method for designing a multilayer film.

Background Art

[0003] A polymer film refers to a thin and flat structure made of a polymer material. Here, a polymer can have a long - chain molecular structure in which monomers, which are repeating units, are linked. The thickness of the polymer film can be variously set up to the millimeter (mm) unit. Polymer films can have various advantages such as flexibility, light weight, cost - efficiency, and processability. Specifically, since polymer films are excellent in flexibility and can be manufactured in various forms and sizes, they can be applied to various product designs. Also, due to the light weight of polymer films, they are suitable for products that require portability and can reduce transportation and handling costs. In addition, they can be made from relatively inexpensive raw materials and are cost - efficient because mass production is possible. Moreover, polymer films can be applied to various manufacturing and processing techniques such as thermal processing, extrusion, and adhesion, so the form, size, thickness, etc. of the product can be easily adjusted. Additionally, polymer films have high durability and strength, some have excellent optical properties, are excellent in water - resistant functions, and may have air permeability as required.

[0004] A multilayer film may be a film composed of two or more polymer or other material layers. Such a multilayer structure can improve the performance of the overall film by combining the unique properties of each layer. For example, one layer may be used to increase durability and strength, another layer may provide waterproofing or moisture resistance, and yet another layer may provide resistance to specific chemicals. In this way, by combining layers with diverse properties, multilayer films can be adapted to a wider range of application fields and specific requirements than single-layer films. [Overview of the project] [Problems that the invention aims to solve]

[0005] One problem we aim to solve is to provide a multilayer film design apparatus and method that can predict the physical properties of a multilayer film using only information about the single-layer films that make up the multilayer film. [Means for solving the problem]

[0006] A multilayer film design apparatus according to one embodiment is a multilayer film design apparatus that executes program code loaded into one or more memory devices by one or more processors and designs the multilayer film based on the physical properties of the multilayer film predicted using only information on single-layer films, wherein the program code is executed, the laminated structure of the multilayer film to be designed is modeled, the fall strength data for the single-layer films forming the laminated structure is collected as stress-strain data, values ​​are read from a memory space accessible to the multilayer film design apparatus, and different feature setting methods are specified according to the read values. A feature setting mode is obtained for this purpose, and according to the feature setting mode, a plurality of physical indicators selected from the cone strength data and the thickness of the single-layer film are set as features. At least one of a plurality of supervised learning models capable of regression analysis is selected as a machine learning model, and the cone strength of the multilayer film is used to predict the cone strength of the multilayer film using the machine learning model which has been trained with the features as independent variables and the cone strength of the multilayer film as the target variable. Design data for the multilayer film can be generated by combining predicted values ​​for other properties with the predicted value of the cone strength to satisfy the design requirements of the multilayer film.

[0007] In one embodiment, when the feature setting mode includes a first feature setting mode, setting as a feature may include setting at least some of the following as features: yield point, yield stress, resilience, strain hardening, failure, strain hardening modulus, flow energy, strain hardening energy, toughness, 1% modulus, and 2% modulus, and the thickness of the single-layer film.

[0008] In one embodiment, selecting the machine learning model includes selecting one of the PLS (partial least squares) model, the Lasso regression model, and the random forest model, wherein the selected machine learning model is trained to predict the deconing strength of the multilayer film by receiving input from a vector generated from at least some of the yield point, yield stress, elasticity, strain hardening, fracture, strain hardening coefficient, flow energy, strain hardening energy, toughness, the 1% coefficient, and the 2% coefficient, as well as the thickness of the single-layer film.

[0009] In one embodiment, when the feature setting mode includes a second feature setting mode, setting as a feature may include setting at least some of the yield stress, yield strain, necking stress, necking strain, breaking stress, and breaking strain, and the thickness of the single-layer film as features.

[0010] In one embodiment, selecting the machine learning model includes selecting one of the PLS (partial least squares) model, the Lasso regression model, and the random forest model selected from the PLS model, the Lasso regression model, and the random forest model may be trained to predict the deconing strength of the multilayer film by receiving input from at least a portion of the yield stress, the yield strain rate, the necking point stress, the necking point strain rate, the fracture stress, and the fracture strain rate, as well as a vector generated from the thickness of the single-layer film.

[0011] In one embodiment, when the feature setting mode includes a third feature setting mode, setting as a feature may include setting the yield stress, yield strain, and the yield stress / yield strain index, and the thickness of the single-layer film as features.

[0012] In one embodiment, selecting the machine learning model includes selecting one of the PLS (partial least squares) model, the Lasso regression model, and the random forest model selected from the PLS model, the Lasso regression model, and the random forest model may be trained to predict the delamination strength of the multilayer film by receiving input from the yield stress, the yield strain ratio, an index obtained by dividing the yield stress by the yield strain ratio, and a vector generated from the thickness of the single-layer film.

[0013] In one embodiment, generating design data for the multilayer film may include combining predicted values ​​for at least one of the following properties of the multilayer film: stiffness, strength, strain rate, tear strength, high-speed detent strength, sealing initiation temperature (SIT), haze, moisture permeability, and air permeability, with a predicted value for the detent strength; generating design data based on the combination of predicted values ​​if the multilayer film designed based on the combination satisfies the design requirements; and redesigning the multilayer film by changing the combination to another combination if the multilayer film designed based on the combination of predicted values ​​does not satisfy the design requirements.

[0014] In one embodiment, the multilayer film is modeled as a multilayer film model, the multilayer film includes a first single-layer film and a second single-layer film laminated together, the multilayer film model includes n (where n is a multiple of 6) model layers laminated together, the first single-layer film can correspond to two or more of the model layers, and the second single-layer film can correspond to two or more other model layers.

[0015] In one embodiment, the program code is executed, the model layer is displayed on a display device, and a user interface is provided that allows the user to set the model layer corresponding to the first single-layer film and the model layer corresponding to the second single-layer film.

[0016] A multilayer film design method according to one embodiment is a method for designing a multilayer film based on the physical properties of a multilayer film predicted solely from information on a single-layer film, performed by a computing device including one or more processors and one or more memory devices, comprising the steps of: modeling the laminated structure of the multilayer film to be designed; collecting fall strength data for the single-layer film forming the laminated structure as stress-strain data; reading values ​​pre-stored in a memory space accessible to the computing device, and obtaining a feature setting mode for specifying different feature setting methods according to the read values. The process may include the steps of: setting a plurality of physical indicators selected from the cone strength data and the thickness of the single-layer film as features, depending on the feature setting mode; selecting at least one of a plurality of supervised learning models capable of regression analysis as a machine learning model; predicting the cone strength of the multilayer film using the machine learning model, which has been trained with the features as independent variables and the cone strength of the multilayer film as the target variable; and generating design data for the multilayer film by combining predicted values ​​for other properties with the predicted value of the cone strength to satisfy the design requirements of the multilayer film.

[0017] In one embodiment, when the feature setting mode includes a first feature setting mode, the step of setting as a feature may include setting at least some of the following as features: yield point, yield stress, resilience, strain hardening, failure, strain hardening modulus, flow energy, strain hardening energy, toughness, 1% modulus, and 2% modulus, and the thickness of the single-layer film.

[0018] In one embodiment, the step of selecting the machine learning model includes selecting one of the PLS (partial least squares) model, the Lasso regression model, and the random forest model as the machine learning model, wherein the selected machine learning model from the PLS model, the Lasso regression model, and the random forest model may be trained to predict the deconing strength of the multilayer film by receiving input from a vector generated from at least some of the yield point, the yield stress, the elasticity, the strain hardening, the fracture, the strain hardening coefficient, the flow energy, the strain hardening energy, the toughness, the 1% coefficient, and the 2% coefficient, as well as the thickness of the single-layer film.

[0019] In one embodiment, when the feature setting mode includes a second feature setting mode, the step of setting as a feature may include setting at least some of the yield stress, yield strain, necking stress, necking strain, breaking stress, and breaking strain, and the thickness of the single-layer film as features.

[0020] In one embodiment, the step of selecting the machine learning model includes selecting one of the PLS (partial least squares) model, the Lasso regression model, and the random forest model selected from the PLS model, the Lasso regression model, and the random forest model may be trained to predict the deconing strength of the multilayer film by receiving input from at least a portion of the yield stress, the yield strain rate, the necking point stress, the necking point strain rate, the fracture stress, and the fracture strain rate, as well as a vector generated from the thickness of the single-layer film.

[0021] In one embodiment, when the feature setting mode includes a third feature setting mode, the step of setting as a feature may include setting the yield stress, yield strain, and the yield stress / yield strain index, and the thickness of the single-layer film as features.

[0022] In one embodiment, the step of selecting the machine learning model includes selecting one of the PLS (partial least squares) model, the Lasso regression model, and the random forest model, wherein the selected machine learning model from the PLS model, the Lasso regression model, and the random forest model may be trained to predict the delamination strength of the multilayer film by receiving input from the yield stress, the yield strain ratio, an index obtained by dividing the yield stress by the yield strain ratio, and a vector generated from the thickness of the single-layer film.

[0023] In one embodiment, the step of generating design data for the multilayer film may include: combining predicted values ​​for at least one of the following properties of the multilayer film: stiffness, strength, strain rate, tear strength, high-speed tapering strength, sealing initiation temperature (SIT), haze, moisture permeability, and air permeability, with a predicted value for the tapering strength; generating design data based on the combination of predicted values ​​if the multilayer film designed based on the combination of predicted values ​​satisfies the design requirements; and redesigning the multilayer film by changing the combination to another combination if the multilayer film designed based on the combination of predicted values ​​does not satisfy the design requirements.

[0024] In one embodiment, the multilayer film is modeled as a multilayer film model, the multilayer film includes a first single-layer film and a second single-layer film laminated together, the multilayer film model includes n (where n is a multiple of 6) model layers laminated together, the first single-layer film can correspond to two or more of the model layers, and the second single-layer film can correspond to two or more other model layers.

[0025] In one embodiment, the method may further include providing a user interface that displays the model layer on a display device and allows a user to set the model layer corresponding to the first single-layer film and the model layer corresponding to the second single-layer film.

Advantages of the Invention

[0026] According to an embodiment, at the time of designing a multilayer film, it is possible to predict the physical properties of the multilayer film only from information on the single-layer films constituting the multilayer film, and to design a multilayer film that can meet given design requirements in consideration of the prediction results regarding other characteristics.

Brief Description of the Drawings

[0027] [Figure 1] FIG. 1 is a block diagram for explaining a multilayer film design apparatus according to an embodiment. [Figure 2] FIG. 2 is a schematic diagram for explaining stress-strain data for a single-layer film according to an embodiment. [Figure 3] FIG. 3 is a flowchart for explaining a multilayer film design method according to an embodiment. [Figure 4] FIG. 4 is a block diagram for explaining a multilayer film design apparatus according to an embodiment. [Figure 5] FIG. 5 is a block diagram for explaining a multilayer film design apparatus according to an embodiment. [Figure 6] FIG. 6 is a schematic diagram for explaining a user interface according to an embodiment. [Figure 7] FIG. 7 is a schematic diagram for explaining a user interface according to an embodiment. [Figure 8] FIG. 8 is a schematic diagram for explaining a computing device according to an embodiment.

Modes for Carrying Out the Invention

[0028] Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings so that they can be easily implemented by a person ordinary skill in the art to which the present invention pertains. However, the present invention can be embodied in a variety of different forms and is not limited to the embodiments described herein. Furthermore, unnecessary parts have been omitted in order to clearly illustrate the present invention in the drawings, and similar parts are denoted by similar reference numerals throughout the specification.

[0029] Throughout this specification and the claims described later, when a part of a provision "includes" a component, this means, unless otherwise stated, that it does not exclude other components and may include other components. Ordinal terms such as First, Second, etc., can be used to describe a variety of components, but such components are not limited by such terms. Such terms are used solely for the purpose of distinguishing one component from another.

[0030] The terms "~part," "~device," and "module" as used in this specification may mean a unit capable of performing at least one function or operation as described herein, which can be embodied in hardware or circuitry, software, or a combination of hardware or circuitry and software. Furthermore, at least some of the configurations or functions of the multilayer film design apparatus and method according to the embodiments described later can be embodied in a program or software, which can be stored on a computer-readable medium.

[0031] Figure 1 is a block diagram illustrating a design apparatus for a multilayer film according to one embodiment, and Figure 2 is a schematic diagram illustrating stress-strain data for a single-layer film according to one embodiment.

[0032] Referring to Figure 1, a multilayer film design apparatus 10 according to one embodiment can execute program code or instructions loaded into one or more memory devices by one or more processors. For example, the multilayer film design apparatus 10 can be embodied in a computing device 50 as described later with respect to Figure 8. In this case, one or more processors may correspond to the processor 510 of the computing device 50, and one or more memory devices may correspond to the memory 530 of the computing device 50. The program code or instructions are executed by one or more processors and can predict the physical properties of a multilayer material, i.e., a multilayer film, using only information on a single-layer film, and design a multilayer film based on the predicted physical properties. In this specification, the term "module" is used to logically distinguish these functions performed by the program code or instructions.

[0033] The multilayer film design apparatus 10 may include all or at least some of the following: a multilayer film modeling module 110, a data acquisition module 120, a feature setting module 130, a model selection module 140, a learning module 150, a prediction module 160, and a design data generation module 162. For example, the learning module 150 may be contained and implemented within the multilayer film design apparatus 10. Alternatively, for example, the learning module 150 may be implemented outside the multilayer film design apparatus 10, and the multilayer film design apparatus 10 may be provided with a machine learning model learned from the learning module 150.

[0034] The multilayer film modeling module 110 can model the laminated structure of a multilayer film to be designed. The multilayer film may include multiple single-layer films. Here, the multiple single-layer films may include one type of single-layer film and other types of single-layer films, and the unique properties of the single-layer films may differ depending on the type. Also, the lamination order of the single-layer films within the multilayer film may vary. The multilayer film modeling module 110 can model a multilayer film by setting parameters such as the type of single-layer film and the lamination order of the single-layer films for which the physical properties of the multilayer film to be predicted are to be made.

[0035] In one embodiment, the multilayer film modeling module 110 can model a multilayer film containing k (where k is an integer of 2 or more) single-layer films. For example, the multilayer film modeling module 110 can model a multilayer film in which the first film layer from the bottom, the second film layer, ..., the kth film layer are sequentially laminated. Features set by the feature setting module 130, described later, can be applied to each of the k single-layer films that make up the multilayer film.

[0036] In one embodiment, the multilayer film modeling module 110 can model a multilayer film containing six single-layer films. For example, the multilayer film modeling module 110 can model a multilayer film in which the first, second, third, fourth, fifth, and sixth film layers are sequentially laminated from the bottom.

[0037] The data acquisition module 120 can collect fall strength data for single-layer films forming a laminated structure modeled by the multilayer film modeling module 110. Here, fall strength may refer to the strength that a material (in this case, a single-layer film) can withstand when subjected to physical impact. Fall strength can be used, for example, as an indicator to measure how much damage a single-layer film will sustain when a weight is dropped from a predetermined height. The multilayer film design apparatus 10 can predict the fall strength of a multilayer film by machine learning based on the fall strength data for each of the multiple single-layer films forming a laminated structure.

[0038] In one embodiment, the data acquisition module 120 can collect drilling strength data for a single-layer film as stress-strain data. The stress-strain data can show the mechanical properties of the single-layer film and, when referred to in conjunction with Figure 2, can be represented in the form of a stress-strain curve. The stress-strain curve can show the stress generated when a force is applied to the single-layer film and how the single-layer film deforms as a result.

[0039] In one embodiment, stress-strain data may include at least one of the following: yield point, yield stress, resilience, strain hardening, failure, strain hardening modulus, flow energy, strain hardening energy, toughness, 1% modulus, and 2% modulus. Here, the yield point indicates the point where the material moves from an elastic deformation region to an inelastic deformation region, and beyond the yield point, the material may not completely return to its original shape even after the applied stress is removed. The yield stress is the stress value at the yield point, and a higher value may mean that the material can withstand a larger stress before deformation occurs. Resilience is the area under the stress-strain curve up to the yield point, and may be a measure of how well the material returns to its original shape after being deformed by a certain stress and then having that stress removed. Strain hardening refers to the phenomenon where a material strengthens as deformation increases beyond the yield point, enabling it to withstand further stress. Fracture can indicate the point at which the material can no longer withstand stress and breaks. The strain hardening coefficient is shown as the slope of the stress-strain curve in the strain hardening section and can be used as an indicator of how quickly the material strengthens. Flow energy is the area under the stress-strain curve from the yield point to strain hardening, and strain hardening energy can indicate the area under the stress-strain curve from strain hardening to failure. Toughness is the entire region under the stress-strain curve and indicates how much energy the material can absorb. The 1% and 2% coefficients can indicate the stress at 1% and 2% deformation, respectively. The values ​​of the 1% and 2% coefficients can be used as measurements indicating the strength within the elastic range of the material.

[0040] In one embodiment, the data acquisition module 120 can collect dart strength data for a single-layer film by a test in accordance with "Method A" of ASTM D1709. Here, the test may evaluate the impact strength or toughness of the single-layer film using a free-falling dart. For more detailed information regarding ASTM D1709, please refer to the ASTM standard document, and therefore, a detailed explanation is omitted herein.

[0041] In one embodiment, the data acquisition module 120 can acquire as data a stress-strain curve constructed based on the results of uniaxial tensile tests performed in the machine direction (MD) and the transverse direction (TD). Here, MD may be the direction in which the film moves along the rollers of the manufacturing machine in the film manufacturing process, and TD is the direction perpendicular to the rollers of the manufacturing machine and may be perpendicular to MD. On the other hand, the uniaxial tensile test may be performed in a manner that measures the resistance force when a single-layer film is stretched in the short-axis direction. In this case, the data acquisition module 120 can acquire data on the stress-strain curve by a test in accordance with ASTM D882. For more detailed information on ASTM D882, please refer to the ASTM standard document and will not provide further details here.

[0042] The feature setting module 130 can perform feature setting according to multiple feature setting modes. The multiple feature setting modes are stored and managed as predetermined values ​​in a memory space accessible by the processor of the multilayer film design apparatus 10. The multilayer film design apparatus 10 can read the values ​​stored in the memory space and determine a feature setting mode to specify a different feature setting method according to the read value.

[0043] In one embodiment, when the feature setting mode is the first feature setting mode, the feature setting module 130 can set a plurality of physical indicators selected from the delamination strength data for a single-layer film and the thickness of the single-layer film as features. In machine learning, a feature is an input variable used to predict the object to be predicted, i.e., the physical properties of a multilayer film, and may represent individual independent variables of the data. Features provide information about the dependent or target variable to be predicted within the dataset and can have a significant impact on the performance of the machine learning model 20. The learning module 150 can train the machine learning model 20 using features. In this case, the features constitute a part of the data input to the machine learning model 20, and the machine learning model 20 can learn patterns based on this data and make predictions. In particular, feature engineering, which involves selecting or transforming features, can be an important process for improving the performance of the machine learning model 20. That is, selecting appropriate features or appropriately transforming features can have a significant impact on the high performance of the machine learning model 20.

[0044] In one embodiment, the feature setting module 130 can set some of the following as features: yield point, yield stress, elasticity, strain hardening, fracture, strain hardening coefficient, flow energy, strain hardening energy, toughness, 1% coefficient, and 2% coefficient, along with the thickness of the single-layer film. In other words, the feature setting module 130 can set some of the stress-strain data collected by the data acquisition module 120, along with the thickness of the single-layer film, as features. When a multilayer film containing k single-layer films is modeled by the multilayer film modeling module 110, the learning module 150 can apply a feature to the first film layer among the k single-layer films that includes some of the stress-strain data collected by the data acquisition module 120, along with the thickness of the single-layer film. The learning module 150 can also apply a feature to the second film layer among the k single-layer films that includes some of the stress-strain data collected by the data acquisition module 120, along with the thickness of the single-layer film. Furthermore, the learning module 150 can train the machine learning model 20 by applying a feature to the k-th film layer that includes only a portion of the stress-strain data collected by the data acquisition module 120 and the thickness of the single-layer film. In this case, the same type of feature may be applied to all k single-layer films for training, or, contrary to the above, one of the k single-layer films may be trained in such a way that a different type of feature is applied to it from the other film layers. This embodiment may be adopted in an environment where computing resources for training and predicting for the machine learning model 20 are limited. In one embodiment, the feature setting module 130 can set the yield point, yield stress, elasticity, strain hardening, fracture, strain hardening coefficient, flow energy, strain hardening energy, toughness, 1% coefficient and 2% coefficient, and the thickness of the single-layer film as features. In other words, the feature setting module 130 can set all the stress-strain data collected by the data acquisition module 120 and the thickness of the single-layer film as features. When a multilayer film containing k single-layer films is modeled by the multilayer film modeling module 110, the learning module 150 can apply a feature to the first film layer among the k single-layer films that includes all the stress-strain data collected by the data acquisition module 120 and the thickness of the single-layer film. The learning module 150 can also apply a feature to the second film layer among the k single-layer films that includes all the stress-strain data collected by the data acquisition module 120 and the thickness of the single-layer film. Furthermore, the learning module 150 can train the machine learning model 20 by applying a feature to the k-th film layer that includes all the stress-strain data collected by the data acquisition module 120 and the thickness of the single-layer film. This embodiment can be adopted in an environment where there are sufficient computing resources to perform training and prediction for the machine learning model 20.

[0045] In one embodiment, the stress-strain data may include first stress-strain data and second stress-strain data. The first stress-strain data may indicate the mechanical properties of the single-layer film in a first direction, and the second stress-strain data may indicate the mechanical properties of the single-layer film in a second direction perpendicular to the first direction. Specifically, the first direction may be MD and the second direction may be TD. The feature setting module 130 can set multiple physical indicators selected from at least one of the first stress-strain data and the second stress-strain data as features.

[0046] In one embodiment, the feature setting module 130 can set as features a portion of the yield point, yield stress, elasticity, strain hardening, fracture, strain hardening coefficient, flow energy, strain hardening energy, toughness, 1% coefficient, and 2% coefficient for the first direction, a portion of the yield point, yield stress, elasticity, strain hardening, fracture, strain hardening coefficient, flow energy, strain hardening energy, toughness, 1% coefficient, and 2% coefficient for the second direction, and the thickness of the single-layer film. In other words, the feature setting module 130 can set as features a selection of only a portion of the first stress-strain data collected by the data acquisition module 120, a selection of only a portion of the second stress-strain data, and the thickness of the single-layer film. When a multilayer film containing k single-layer films is modeled by the multilayer film modeling module 110, the learning module 150 can apply a selection of only a portion of the first stress-strain data collected by the data acquisition module 120, a selection of only a portion of the second stress-strain data, and the thickness of the single-layer film to a feature for the first film layer among the k single-layer films. The learning module 150 can also apply a feature to the second film layer among the k single-layer films that includes a selection of only a portion of the first stress-strain data collected by the data acquisition module 120, a selection of only a portion of the second stress-strain data, and the thickness of the single-layer film. Furthermore, the learning module 150 can train the machine learning model 20 by applying a feature to the kth film layer that includes a selection of only a portion of the first stress-strain data collected by the data acquisition module 120, a selection of only a portion of the second stress-strain data, and the thickness of the single-layer film.

[0047] In one embodiment, the feature setting module 130 can set as features all of the following: yield point, yield stress, elasticity, strain hardening, fracture, strain hardening coefficient, flow energy, strain hardening energy, toughness, 1% coefficient and 2% coefficient for the first direction, yield point, yield stress, elasticity, strain hardening, fracture, strain hardening coefficient, flow energy, strain hardening energy, toughness, 1% coefficient and 2% coefficient for the second direction, and the thickness of the single-layer film. In other words, the feature setting module 130 can set as features all of the first stress-strain data collected by the data acquisition module 120, all of the second stress-strain data, and the thickness of the single-layer film. When a multilayer film containing k single-layer films is modeled by the multilayer film modeling module 110, the learning module 150 can apply all of the first stress-strain data, all of the second stress-strain data, and the thickness of the single-layer film collected by the data acquisition module 120 to the first film layer among the k single-layer films. The learning module 150 can also apply a feature containing all of the first stress-strain data, all of the second stress-strain data, and the thickness of the single-layer film collected by the data acquisition module 120 to the second film layer among the k single-layer films. Furthermore, the learning module 150 can train the machine learning model 20 by applying a feature containing all of the first stress-strain data, all of the second stress-strain data, and the thickness of the single-layer film collected by the data acquisition module 120 to the kth film layer.

[0048] In another embodiment, when the feature setting mode is the second feature setting mode, the feature setting module 130 can also set yield stress, yield strain, necking stress, necking strain, breaking stress, and breaking strain as features. Here, yield stress, yield strain, necking strain, and breaking stress may be features corresponding to the yield stress, yield point, strain hardening, and fracture in the stress-strain data described above. On the other hand, necking stress and breaking strain may be newly specified features analyzed from the stress-strain data. Here, yield stress can represent the stress at the yield point, which is the point where the material begins to deform permanently beyond its elastic limit, in other words, the point where the material moves from an elastic deformation region to an inelastic deformation region, and yield strain can represent the strain corresponding to the yield stress. Necking stress indicates the stress at the point where the cross-section of the material begins to narrow partially (necking point), and generally, the load capacity of the material may decrease sharply after the necking point. Necking strain rate can indicate the strain rate at the necking point. Fracture stress indicates the stress at the point where the material finally fractures, and at the fracture point, the material can no longer withstand the load. Fracture strain rate can indicate the strain rate corresponding to the fracture stress.

[0049] In one embodiment, the feature setting module 130 can set as features at least a portion of the yield stress, yield strain, necking stress, necking strain, fracture stress, and fracture strain for the first and second directions, respectively, and the thickness of the single-layer film. When a multilayer film containing k single-layer films is modeled by the multilayer film modeling module 110, the learning module 150 can apply features to the first film layer among the k single-layer films, including a selection of at least a portion of the yield stress, yield strain, necking stress, necking strain, fracture stress, and fracture strain for the first direction, a selection of at least a portion of the yield stress, yield strain, necking stress, necking strain, fracture stress, and fracture strain for the second direction, and the thickness of the single-layer film. Furthermore, the learning module 150 can apply a feature to the second film layer among the k single-layer films, including at least a selection of yield stress, yield strain, necking point stress, necking point strain, fracture stress, and fracture strain in the first direction, at least a selection of yield stress, yield strain, necking point stress, necking point strain, fracture stress, and fracture strain in the second direction, and the thickness of the single-layer film. Additionally, the learning module 150 can train the machine learning model 20 by applying a feature to the k-th film layer, including at least a selection of yield stress, yield strain, necking point stress, necking point strain, fracture stress, and fracture strain in the first direction, at least a selection of yield stress, yield strain, necking point stress, necking point strain, fracture stress, and fracture strain in the second direction, and the thickness of the single-layer film. In this case, the same type of feature may be applied to all k single-layer films during training, or, contrary to the above, one of the k single-layer films may be trained in such a way that a different type of feature is applied to one of the film layers from the other film layers. In another embodiment, when the feature setting mode is the third feature setting mode, the feature setting module 130 can also set the yield stress, yield strain, and the yield stress / yield strain index as features. Specifically, the feature setting module 130 can set the yield stress, yield strain, and yield stress / yield strain index for the first and second directions, respectively, as well as the thickness of the single-layer film, as features. When a multilayer film containing k single-layer films is modeled by the multilayer film modeling module 110, the learning module 150 can apply features to the first film layer among the k single-layer films, including all indicators for yield stress, yield strain, and yield stress divided by yield strain in the first direction, all indicators for yield stress, yield strain, and yield stress divided by yield strain in the second direction, and the thickness of the single-layer film. The learning module 150 can also apply features to the second film layer among the k single-layer films, including all indicators for yield stress, yield strain, and yield stress divided by yield strain in the first direction, all indicators for yield stress, yield strain, and yield stress divided by yield strain in the second direction, and the thickness of the single-layer film. Furthermore, the learning module 150 can train the machine learning model 20 by applying features to the k-th film layer that include all indicators for the first direction, such as yield stress, yield strain, and yield stress divided by yield strain, as well as all indicators for the second direction, such as yield stress, yield strain, and yield stress divided by yield strain, and the thickness of the single-layer film.

[0050] For example, when the multilayer film modeling module 110 models a multilayer film containing six single-layer films, it can apply seven features to each film layer: all the indicators for yield stress, yield strain, and yield stress divided by yield strain in the first direction, all the indicators for yield stress, yield strain, and yield stress divided by yield strain in the second direction, and the thickness of the single-layer film. Since there are a total of six layers, a total of 42 features can be applied.

[0051] The model selection module 140 can select a machine learning model to be used to predict the physical properties of a multilayer film, namely, its deconing strength. Specifically, the model selection module 140 can select at least one of several supervised learning models capable of regression analysis for predicting the deconing strength of a multilayer film as the machine learning model. Supervised learning is a method of training a model using labeled training data, in which the model learns the relationship between input variables and target variables, thereby enabling it to predict the value of the target variable for new input data. In one embodiment, the model selection module 140 can select a supervised learning model capable of regression analysis from among PLS (partial least squares), linear regression, polynomial regression, ridge regression, lasso regression, logistic regression, decision trees, random forests, gradient boosting, and neural networks. PLS models can be used to model the relationships between multiple independent variables and multiple dependent variables (or target variables), and can identify independent variables that are deemed important for predicting the dependent variable. In linear regression, the model learns the linear relationship between the input variables and the target variable, while in polynomial regression, the model extends linear regression to include higher-order terms of the input variables to learn more complex relationships. Ridge regression and lasso regression can add regularization to linear regression to prevent overfitting and adjust the complexity of the model, and logistic regression, a classification algorithm for solving binary classification problems, can be used in regression analysis because it predicts the outcome probabilistically.Decision trees, random forests, and gradient boosting can learn complex nonlinear relationships and can be used not only for classification but also for regression problems. Neural networks can solve a variety of regression problems, and dip learning in particular can be very effective for learning complex nonlinear relationships. Of course, the supervised learning models capable of regression analysis that the model selection module 140 can select are not limited to those exemplified above, and other types of models not exemplified may be selected considering the specific implementation purpose, implementation environment, required performance, etc.

[0052] The learning module 150 can train the machine learning model 20 using the features set by the feature setting module 130 as independent variables and the deconing strength of the multilayer film as the target variable. This allows the machine learning model 20 to learn the relationship between the features and the target variable, enabling it to predict the physical properties of a multilayer film based on new data, namely, physical property data for a single-layer film. In particular, the learning module 150 can use different learning methods depending on the model selected by the model selection module 140.

[0053] The prediction module 160 can predict the delamination strength of a multilayer film using the machine learning model 20 trained by the learning module 150. In other words, the prediction module 160 can transmit new data, namely physical property data for a single-layer film, to the machine learning model 20 and obtain a prediction result for the delamination strength of a multilayer film.

[0054] Specifically, the model selection module 140 can perform model selection according to multiple model selection modes. The multiple model selection modes are stored and managed as predetermined values ​​in a memory space accessible by the processor of the multilayer film design apparatus 10. The multilayer film design apparatus 10 can read the values ​​stored in the memory space and determine a model selection mode to specify a different model selection method according to the read value.

[0055] In one embodiment, when the model selection mode is the first model selection mode, the model selection module 140 can select the PLS model as the machine learning model to be used to predict the cone strength of a multilayer film. As described above, the features set by the feature setting module 130 according to the feature setting mode can be combined, for example, in the form of a single vector and used as input to the PLS model. The cone strength of the multilayer film is set as the target variable, and the learning module 150 can train the PLS model in a direction that maximizes the correlation between the input features and the target variable. At this time, the learning module 150 optimizes the "n_components" parameter so that the PLS model can adequately explain the data, and the optimal "n_components" value may be determined by cross-validation. Through the PLS model thus trained, features for a new single-layer film can be input to predict the cone strength of a multilayer film.

[0056] In another embodiment, when the model selection mode is the second model selection mode, the model selection module 140 can select a Lasso regression model as the machine learning model used to predict the cone strength of a multilayer film. As described above, the features set by the feature setting module 130 according to the feature setting mode can be combined, for example, in the form of a single vector and used as input to the Lasso regression model. The cone strength of a multilayer film is set as the target variable, and the learning module 150 can train the Lasso regression model in a direction that guides the model to select only important features by L1 normalization (normalization that adds the sum of the absolute values ​​of the regression coefficients as a penalty to the objective function). At this time, the learning module 150 optimizes the "alpha" parameter for adjusting the penalty strength during model training, and the optimal "alpha" value may be determined by cross-validation. Through the Lasso regression model thus trained, features for a new single-layer film can be input to predict the cone strength of a multilayer film.

[0057] In another embodiment, when the model selection mode is the third model selection mode, the model selection module 140 can select a random forest model as the machine learning model used to predict the cone strength of a multilayer film. As described above, the features set by the feature setting module 130 according to the feature setting mode can be combined, for example, in the form of a single vector and used as input to the random forest model. The cone strength of the multilayer film is set as the target variable, and the learning module 150 can be trained by setting "n_estimators", "max_depth", "min_samples_split", etc. as hyperparameters of the random forest model. Through the random forest model thus trained, the cone strength of the multilayer film can be predicted by inputting features for a new single-layer film. The design data generation module 162 can generate design data for a multilayer film by combining predicted values ​​for other properties with predicted values ​​for the puncture strength obtained through the prediction module 160, in order to meet the design requirements for the multilayer film. In one embodiment, the other properties may include at least one of the following for the multilayer film: stiffness, strength, strain rate, tear strength, high-speed puncture strength, sealing initiation temperature (SIT), transparency (haze), moisture permeability, and air permeability.

[0058] The design data generation module 162 can linearly combine predicted values ​​for at least one of the following properties of a multilayer film: stiffness, strength, strain rate, tear strength, high-speed delamination strength, seal initiation temperature, transparency, moisture permeability, and air permeability, with a predicted value for delamination strength. The design requirements for the multilayer film to be designed may include requirements to ensure mechanical properties, thermal properties, optical properties, barrier properties, and chemical properties to secure the desired performance. If the multilayer film designed based on the linear combination of predicted values ​​satisfies the design requirements, the design data generation module 162 can generate design data based on the linear combination. In this case, the design data may include specific predicted values. In contrast, if a multilayer film designed based on a primary combination of predicted values ​​does not meet the design requirements, the design data generation module 162 can redesign the multilayer film by changing the primary combination to another combination. In other words, the design data generation module 162 can secondarily combine predicted values ​​for at least one of the properties of the multilayer film—stiffness, strength, strain rate, tear strength, high-speed delamination strength, seal initiation temperature, transparency, moisture permeability, and air permeability—with the predicted value of delamination strength, so that the combination is different from the primary combination. In this case, the predicted value of delamination strength in the primary combination and the predicted value of delamination strength in the secondary combination may be different values.

[0059] If the multilayer film designed based on the quadratic combination of predicted values ​​meets the design requirements, the design data generation module 162 can generate design data based on the quadratic combination. Contrary to the above, if the multilayer film designed based on the quadratic combination of predicted values ​​does not meet the design requirements, the design data generation module 162 can repeat the process of redesigning the multilayer film by changing the quadratic combination to another combination. According to this embodiment, when designing a multilayer film, the physical properties of the multilayer film can be predicted using only information about the single-layer films that constitute the multilayer film. In particular, by learning and predicting using various indicators of cone strength stress-strain data for single-layer films and the thickness of the single-layer films as features, the cone strength of the multilayer film can be predicted with high accuracy. Furthermore, by differentiating the number of indicators selected as features when performance is prioritized and the number of indicators selected as features to save computing resources, flexible design of the multilayer film design apparatus is possible depending on the purpose and environment of implementation. Ultimately, by considering the prediction results for other properties of the multilayer film together with the prediction results for cone strength, it is possible to design a multilayer film that can satisfy the given design requirements.

[0060] Figure 3 is a flowchart illustrating a design method for a multilayer film according to one embodiment.

[0061] Referring to Figure 3, a multilayer film design method according to one embodiment may include the steps of: modeling the laminated structure of the multilayer film to be designed (S301); collecting cone strength data for single-layer films forming a laminated structure (S302); reading values ​​previously stored in memory space and obtaining a feature setting mode for specifying different feature setting methods according to the read values ​​(S303); setting a plurality of physical indicators selected from cone strength data and the thickness of the single-layer film as features according to the feature setting mode (S304); selecting at least one of a plurality of supervised learning models capable of regression analysis as a machine learning model (S305); training the machine learning model with features as independent variables and the cone strength of the multilayer film as the target variable (S306); predicting the cone strength of the multilayer film using the trained machine learning model (S307); and generating design data for the multilayer film by combining predicted values ​​for other characteristics and predicted values ​​for cone strength to satisfy the design requirements of the multilayer film (S308).

[0062] For more specific details regarding the design method of multilayer films, please refer to and apply the descriptions of the embodiments described herein; therefore, redundant explanations will be omitted below.

[0063] Figure 4 is a block diagram illustrating a design apparatus for multilayer films according to one embodiment.

[0064] Referring to Figure 4, a multilayer film design apparatus 10 according to one embodiment may include all or at least some of the following: a multilayer film modeling module 110, a data acquisition module 120, a feature setting module 130, a model selection module 140, a learning module 150, a prediction module 160, a design data generation module 162, a correlation coefficient analysis module 170, and a PLS (partial least squares) analysis module 180. For example, the learning module 150 may be contained and implemented within the multilayer film design apparatus 10. Contrary to the foregoing, for example, the learning module 150 may be implemented outside the multilayer film design apparatus 10, and the multilayer film design apparatus 10 may be provided with a machine learning model learned from the learning module 150. With respect to the multilayer film modeling module 110, data acquisition module 120, feature setting module 130, model selection module 140, learning module 150, prediction module 160, and design data generation module 162, you can refer to the descriptions of the multilayer film modeling module 110, data acquisition module 120, feature setting module 130, model selection module 140, learning module 150, prediction module 160, and design data generation module 162 described above by referring to Figures 1 and 2, as long as it does not contradict or inconsist with what is described in this embodiment. Therefore, redundant descriptions are omitted here.

[0065] The correlation coefficient analysis module 170 can analyze the correlation coefficients between mechanical properties in stress-strain data, and the PLS analysis module 180 can predict the cone strength value of a multilayer film using PLS analysis according to the correlation coefficients analyzed by the correlation coefficient analysis module 170. PLS analysis can be used to model the relationship between multiple independent variables and multiple dependent variables (or target variables), and can find independent variables that are deemed important for predicting the dependent variables. Specifically, the correlation coefficient analysis module 170 can construct independent and dependent variables from various indicators of cone strength stress-strain data for a single-layer film and find the component with the largest covariance. A new subspace is generated from this, where the relationship between the independent and dependent variables may be most pronounced. The PLS analysis module 180 can repeatedly adjust the variables in proportion to their respective covariances for the new subspace and search for new components until the desired number of components are extracted. In this way, the physical properties of a multilayer film can be predicted together with or instead of the prediction module 160.

[0066] In one embodiment, the correlation coefficient analysis module 170 can be implemented so as to be logically separate from the feature setting module 130, or it can be implemented so as to be logically included in the feature setting module 130. When the correlation coefficient analysis module 170 is implemented so as to be logically included in the feature setting module 130, multiple physical indicators selected from the fall strength data for a single-layer film and the thickness of the single-layer film can be set as features, and the correlation coefficient between mechanical properties can be analyzed on the stress-strain data as a preprocessing step to reduce the data dimension.

[0067] On the other hand, in one embodiment, the PLS analysis module 180 can be implemented so as to be logically separate from the model selection module 140, or it can be implemented so as to be logically included in the model selection module 140. When the PLS analysis module 180 is implemented so as to be logically included in the model selection module 140, when selecting a machine learning model to use to predict the fallout intensity, it is possible to select a model that performs PLS analysis.

[0068] According to this embodiment, by performing a separate analysis on indicators that are highly collinear with each other among various indicators of cone strength stress-strain data for single-layer films, the accuracy of predicting the cone strength of multilayer films can be further improved.

[0069] Figure 5 is a block diagram illustrating a design apparatus for a multilayer film according to one embodiment.

[0070] Referring to Figure 5, the multilayer film design apparatus 10 may include all or at least some of the multilayer film modeling module 110, data acquisition module 120, feature setting module 130, model selection module 140, learning module 150, prediction module 160, design data generation module 162, and model layer setting interface providing module 190. For example, the learning module 150 may be contained and implemented within the multilayer film design apparatus 10. Contrary to the foregoing, for example, the learning module 150 may be implemented outside the multilayer film design apparatus 10, and the multilayer film design apparatus 10 may be provided with a machine learning model learned from the learning module 150. With respect to the multilayer film modeling module 110, data acquisition module 120, feature setting module 130, model selection module 140, learning module 150, prediction module 160, and design data generation module 162, you can refer to the descriptions of the multilayer film modeling module 110, data acquisition module 120, feature setting module 130, model selection module 140, learning module 150, prediction module 160, and design data generation module 162 described above by referring to Figures 1 and 2, as long as it does not contradict or inconsist with what is described in this embodiment. Therefore, redundant descriptions are omitted here.

[0071] The multilayer film modeled by the multilayer film modeling module 110 may be modeled as a multilayer film model. Here, the multilayer film model may be in a form that includes a plurality of empty slots arranged continuously in the vertical direction. Specifically, the multilayer film model may include n slots (where n is a multiple of 6) stacked on top of each other, i.e., model layers. When the multilayer film includes a first single-layer film and a second single-layer film stacked on top of each other, the first single-layer film may correspond to two or more model layers, and the second single-layer film may correspond to two or more other model layers. The data acquisition module 120 can collect cone intensity data for the first single-layer film and the second single-layer film based on such a multilayer film model.

[0072] The model layer setting interface module 190 enables the user to configure such model layers. To this end, the model layer setting interface module 190 displays the model layers on a display device and provides a user interface that allows the user to configure the model layer corresponding to the first single-layer film and the model layer corresponding to the second single-layer film.

[0073] Figures 6 and 7 are schematic diagrams illustrating a user interface according to one embodiment.

[0074] Referring to Figure 6, the model layer setting interface module 190 can present the first to sixth model layers, represented as n (where n is a multiple of 6) slots stacked on top of each other, to the user via a display device. The user can input information through the user interface regarding how many model layers a single-layer film occupies, the stacking order between the single-layer films, and the type and specific characteristics of the single-layer film assigned to a particular model layer. For example, as shown in the drawing, the user can input information that the first and second single-layer films each occupy three model layers, and that the second single-layer film is stacked on top of the first single-layer film, by assigning the first single-layer film to the first to third model layers and the second single-layer film to the fourth to sixth model layers. Although not shown, the user can also input the types of the first and second single-layer films and other specific details.

[0075] Next, referring to Figure 7, the user can input information that the first, second, and third single-layer films each occupy two model layers, the second single-layer film is laminated on the first, third, and third single-layer films, and the third single-layer film is laminated on the second single-layer film, by assigning the first single-layer film to the first or second model layer, the second single-layer film is laminated on the first single-layer film, and the third single-layer film is laminated on the second single-layer film. Although not shown in the figure, the user can also input the types of each of the first, second, and third single-layer films and other specific details. According to this embodiment, the feature setting module 130 can set features based on the same 6-layer model structure whether there are two types of single-layer films constituting the multilayer film or three types of single-layer films. This allows the learning module 150 to train the machine learning model 20 using the same number of features for the model layer structure as independent variables and the cone fall intensity of the multilayer film as the target variable, regardless of the number of types of single-layer films constituting the multilayer film. As a result, a machine learning model 20 independent of the number of types of single-layer films constituting the multilayer film can be applied, which reduces the complexity of implementation and computational cost while maintaining high predictive accuracy compared to the case where different layer structures must be learned for each number of types of single-layer films constituting the multilayer film.

[0076] Figure 8 is a schematic diagram illustrating a computing device according to one embodiment.

[0077] Referring to Figure 8, the multilayer film design apparatus and method according to the embodiment can be implemented using a computing device 50. Such a computing device 50 can be implemented as various forms of electronic equipment, servers, or similar devices, and its functions can be implemented through a combination of software and hardware. The computing device 50 may include at least one of a processor 510, memory 530, user interface input device 540, user interface output device 550, and storage device 560 that communicate via the bus 520. The computing device 50 may further include a network interface 570 that is electrically connected to the network 40. The network interface 570 can transmit or receive signals with other devices via the network 40. The processor 510 can be implemented as various types of arithmetic units, such as an MCU (microcontroller unit), AP (application processor), CPU (central processing unit), GPU (graphic processing unit), NPU (neural processing unit), QPU (quantum processing unit), etc. The processor 510 is a semiconductor device that executes instructions stored in the memory 530 or storage device 560, and can play a core role in the system. The program code and data stored in the memory 530 or storage device 560 instruct the processor 510 to perform specific tasks, thereby enabling the operation of the entire system. The processor 510 can be configured to implement the various functions and methods described above with respect to Figures 1 to 7.

[0078] The memory 530 and storage device 560 may include various forms of volatile or non-volatile storage media for system data storage and access. For example, the memory 530 may include ROM (read-only memory) 531 and RAM (random access memory) 532. In one embodiment, the memory 530 may be integrated into the processor 510, in which case the data transmission speed between the memory 530 and the processor 510 can be very fast. In another embodiment, the memory 530 may be located outside the processor 510, in which case the memory 530 can be connected to the processor 510 through various data buses or interfaces. Such connections are possible through various already known means, such as PCIe (peripheral component interconnect express) interfaces or memory controllers for high-speed data transmission.

[0079] In one embodiment, at least some of the configurations or functions of the multilayer film design apparatus and method according to the embodiment can be embodied in a program or software executed on a computing device 50, and the program or software may be stored on a computer-readable medium. Specifically, the computer-readable medium according to one embodiment may contain a program that causes a computer, which includes a processor 510 that executes programs or instructions stored in memory 530 or storage device 560, to perform steps included in the multilayer film design method according to the embodiment.

[0080] In one embodiment, at least some of the configurations or functions of the multilayer film design apparatus and method according to the embodiment may be implemented using the hardware or circuits of the computing device 50, or by separate hardware or circuits that can be electrically connected to the computing device 50. According to this embodiment, when designing a multilayer film, the physical properties of the multilayer film can be predicted using only information about the single-layer films that constitute the multilayer film. In particular, by learning and predicting various indicators of cone strength stress-strain data for single-layer films and the thickness of the single-layer films as features, the cone strength of the multilayer film can be predicted with high accuracy. Furthermore, by differentiating the number of indicators selected as features when performance is prioritized and the number of indicators selected as features to save computing resources, flexible design of the multilayer film design apparatus is possible depending on the purpose and environment of implementation. Ultimately, by considering the prediction results for other properties of the multilayer film together with the prediction results for cone strength, it is possible to design a multilayer film that can satisfy the given design requirements.

[0081] Although embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto. Various modifications and improvements made by persons with ordinary skill in the art to which the present invention pertains, using the basic concepts of the present invention as defined in the following claims, also fall within the scope of the present invention.

Claims

1. A multilayer film design apparatus that executes program code loaded into one or more memory devices by one or more processors, and designs a multilayer film based on the physical properties of the multilayer film predicted solely from information on a single-layer film, The aforementioned program code is executed, Model the laminated structure of the multilayer film to be designed, The delamination strength data for the single-layer film forming the aforementioned laminated structure is collected as stress-strain data. The design apparatus for the multilayer film reads a value pre-stored in a memory space accessible by the apparatus, and obtains a feature setting mode to specify a different feature setting method according to the read value, Depending on the feature setting mode, a plurality of physical indicators selected from the fall strength data and the thickness of the single-layer film are set as features. Select at least one of several supervised learning models capable of regression analysis as the machine learning model. Using the aforementioned features as independent variables and the delamination strength of the multilayer film as the target variable, the machine learning model is trained to predict the delamination strength of the multilayer film. To satisfy the design requirements for the multilayer film, design data for the multilayer film is generated by combining predicted values ​​for other properties with predicted values ​​for the decontamination strength. A design device for multilayer films.

2. When the feature setting mode includes the first feature setting mode, setting it as a feature means: yield point, yield stress, elasticity ( A multilayer film design apparatus according to claim 1, comprising setting at least some of the following as features: resilience, strain hardening, failure, strain hardening modulus, flow energy, strain hardening energy, toughness, 1% modulus, and 2% modulus, and the thickness of the single-layer film.

3. Selecting the aforementioned machine learning model means This includes selecting one of the following as the machine learning model: a PLS (partial least squares) model, a Lasso regression model, and a random forest model. The machine learning model selected from the aforementioned PLS model, Lasso regression model, and Random Forest model is: A multilayer film design apparatus according to claim 2, which is trained to receive inputs of a vector generated from the yield point, yield stress, elasticity, strain hardening, fracture, strain hardening coefficient, flow energy, strain hardening energy, toughness, 1% coefficient, and 2% coefficient, and the thickness of the single-layer film, and to predict the fall strength of the multilayer film.

4. When the aforementioned feature setting mode includes a second feature setting mode, setting it as the feature means: A multilayer film design apparatus according to claim 1, comprising setting at least a portion of the yield stress, yield strain, necking stress, necking strain, breaking stress, and breaking strain, and the thickness of the single-layer film as features.

5. Selecting the aforementioned machine learning model means This includes selecting one of the following as the machine learning model: a PLS (partial least squares) model, a Lasso regression model, and a random forest model. The machine learning model selected from the aforementioned PLS model, Lasso regression model, and Random Forest model is: A design apparatus for a multilayer film according to claim 4, which receives input of at least a portion of the yield stress, yield strain, necking stress, necking strain, fracture stress, and fracture strain, and a vector generated from the thickness of the single-layer film, and is trained to predict the fall strength of the multilayer film.

6. When the aforementioned feature setting mode includes a third feature setting mode, setting it as a feature means: A multilayer film design apparatus according to claim 1, comprising setting the yield stress, yield strain, and an index obtained by dividing the yield stress by the yield strain (yield stress / yield strain) and the thickness of the single-layer film as features.

7. Selecting the aforementioned machine learning model means This includes selecting one of the following as the machine learning model: a PLS (partial least squares) model, a Lasso regression model, and a random forest model. The machine learning model selected from the aforementioned PLS model, Lasso regression model, and Random Forest model is: A design apparatus for a multilayer film according to claim 6, which receives input from the yield stress, the yield strain ratio, an index obtained by dividing the yield stress by the yield strain ratio, and a vector generated from the thickness of the single-layer film, and is trained to predict the delamination strength of the multilayer film.

8. Generating design data for the aforementioned multilayer film is A predicted value for at least one of the following properties of the multilayer film—rigidity, strength, strain rate, tear strength, high-speed tapering strength, sealing initiation temperature (SIT), haze, moisture permeability, and air permeability—is combined with the predicted value of the tapering strength. If a multilayer film designed based on the aforementioned combination of predicted values ​​satisfies the design requirements, the design data is generated based on the aforementioned combination. A multilayer film design apparatus according to claim 1, comprising changing the combination of predicted values ​​to another combination and redesigning the multilayer film if the multilayer film designed based on the combination of predicted values ​​does not satisfy the design requirements.

9. The aforementioned multilayer film is modeled as a multilayer film model, The multilayer film includes a first single-layer film and a second single-layer film laminated together, The multilayer film model includes n (where n is a multiple of 6) model layers stacked on top of each other. The multilayer film design apparatus according to claim 1, wherein the first single-layer film corresponds to two or more model layers among the model layers, and the second single-layer film corresponds to two or more other model layers among the model layers.

10. The aforementioned program code is executed, and further, A multilayer film design apparatus according to claim 9, which displays the model layer on a display device and provides a user interface that allows the user to set the model layer corresponding to the first single-layer film and the model layer corresponding to the second single-layer film.

11. A method for designing a multilayer film, which involves designing the multilayer film based on the physical properties of the multilayer film predicted solely from information about a single-layer film, performed by a computing device including one or more processors and one or more memory devices, A step of modeling the laminated structure of the multilayer film to be designed; A step of collecting the drop strength data for the single-layer film forming the aforementioned laminated structure as stress-strain data; The steps include: reading a value pre-stored in a memory space accessible by the computing device, and obtaining a feature setting mode to specify a different feature setting method according to the read value; Depending on the feature setting mode, the step of setting a plurality of physical indicators selected from the cone strength data and the thickness of the single-layer film as a feature; The step of selecting at least one machine learning model from among several supervised learning models capable of regression analysis; A step of predicting the deconing intensity of the multilayer film using the machine learning model that has been trained with the features as independent variables and the deconing intensity of the multilayer film as the target variable; and The step of generating design data for the multilayer film by combining predicted values ​​for other properties with predicted values ​​for the fall strength so as to satisfy the design requirements for the multilayer film, Design methods for multilayer films.

12. When the feature setting mode includes the first feature setting mode, the step of setting as a feature is: A method for designing a multilayer film according to claim 11, comprising the step of setting at least some of the following as features: yield point, yield stress, resilience, strain hardening, failure, strain hardening modulus, flow energy, strain hardening energy, toughness, 1% modulus, and 2% modulus, and the thickness of the single-layer film.

13. The step of selecting the aforementioned machine learning model is: The process includes the step of selecting one of the following as the machine learning model: a PLS (partial least squares) model, a Lasso regression model, and a random forest model. The machine learning model selected from the aforementioned PLS model, Lasso regression model, and Random Forest model is: A method for designing a multilayer film according to claim 12, wherein the system is trained to receive inputs of a vector generated from the yield point, the yield stress, the elasticity, the strain hardening, the fracture, the strain hardening coefficient, the flow energy, the strain hardening energy, the toughness, the 1% coefficient, and the 2% coefficient, and the thickness of the single-layer film, in order to predict the fall strength of the multilayer film.

14. When the feature setting mode includes a second feature setting mode, the step of setting as a feature is: A method for designing a multilayer film according to claim 11, comprising the step of setting at least some of the yield stress, yield strain, necking stress, necking strain, breaking stress, and breaking strain, and the thickness of the single-layer film, as features.

15. The step of selecting the aforementioned machine learning model is: The process includes the step of selecting one of the following as the machine learning model: a PLS (partial least squares) model, a Lasso regression model, and a random forest model. The machine learning model selected from the aforementioned PLS model, Lasso regression model, and Random Forest model is: A method for designing a multilayer film according to claim 14, wherein the system is trained to predict the deconing strength of the multilayer film by receiving inputs of at least a portion of the yield stress, the yield strain, the necking stress, the necking strain, the fracture stress, and the fracture strain, and a vector generated from the thickness of the single-layer film.

16. When the feature setting mode includes a third feature setting mode, the step of setting as a feature is: A method for designing a multilayer film according to claim 11, comprising the step of setting the yield stress, yield strain, and an index obtained by dividing the yield stress by the yield strain (yield stress / yield strain) and the thickness of the single-layer film as features.

17. The step of selecting the aforementioned machine learning model is: The process includes the step of selecting one of the following as the machine learning model: a PLS (partial least squares) model, a Lasso regression model, and a random forest model. The machine learning model selected from the aforementioned PLS model, Lasso regression model, and Random Forest model is: A method for designing a multilayer film according to claim 16, wherein the system is trained to predict the delamination strength of the multilayer film by receiving input from the yield stress, the yield strain ratio, an index obtained by dividing the yield stress by the yield strain ratio, and a vector generated from the thickness of the single-layer film.

18. The step of generating design data for the multilayer film is: A step of combining a predicted value for at least one of the following properties of the multilayer film: stiffness, strength, strain rate, tear strength, high-speed tapering strength, sealing initiation temperature (SIT), haze, moisture permeability, and air permeability, with a predicted value for the tapering strength; If a multilayer film designed based on the combination of predicted values ​​satisfies the design requirements, the step of generating the design data based on the combination; and A method for designing a multilayer film according to claim 11, further comprising the step of redesigning the multilayer film by changing the combination of predicted values ​​to another combination if the multilayer film designed based on the combination of predicted values ​​does not satisfy the design requirements.

19. The aforementioned multilayer film is modeled as a multilayer film model, The multilayer film includes a first single-layer film and a second single-layer film laminated together, The multilayer film model includes n (where n is a multiple of 6) model layers stacked on top of each other. The method for designing a multilayer film according to claim 11, wherein the first single-layer film corresponds to two or more model layers among the model layers, and the second single-layer film corresponds to two or more other model layers among the model layers.

20. The method for designing a multilayer film according to claim 19, further comprising the step of displaying the model layer on a display device and providing a user interface that allows the user to set the model layer corresponding to the first single-layer film and the model layer corresponding to the second single-layer film.

Images