Method and apparatus for predicting characteristics of polymer composites

Abstract

According to various embodiments of the present disclosure, a method and apparatus for predicting characteristics of a polymer composite may be provided. A method of predicting characteristics of a polymer composite includes the steps of: inputting a formulation including two or more materials including at least one polymer and a mix proportion of each of the two or more materials; predicting the physical property of the polymer composite material based on the formula based on the formula and a physical property prediction model; and outputting the physical properties of the polymer composite material.

Description

Technical Field

[0001] The present disclosure relates to a method and apparatus for predicting characteristics of polymer composite materials. Background Art

[0002] Polymer composites are widely used in various industrial fields. In order to predict the physical properties based on the formulation during the development and synthesis of polymer composites, a combination of experimental, simulation and data-based methods is used.

[0003] Current technologies still use trial-and-error methods, for example, by trying various material ratios and process conditions in the laboratory to collect physical property data, inferring material properties based on the physical property data, and improving the method to predict the physical properties of polymer composites generated based on the formula.

[0004] Therefore, attempts are being made to use models trained based on the development of artificial intelligence to predict the physical properties of polymer composites based on formulations, but the training of commercially available artificial intelligence models is costly and time-consuming, and in particular, they have the limitation of being unable to consider the countless variables and interactions of the materials that make up the formulation. Summary of the Invention

[0005] (1) Technical issues to be resolved

[0006] The present disclosure may provide a method and apparatus for predicting characteristics of polymer composite materials.

[0007] The technical problems to be solved by various embodiments are not limited to the technical problems mentioned. It is obvious to those skilled in the art that other technical problems not mentioned can be clearly understood from the following description.

[0008] (2) Technical solution

[0009] According to one embodiment, a method for predicting characteristics of a polymer composite material can be provided, comprising the following steps: inputting a formula comprising two or more materials including at least one polymer and a mix ratio of each of the two or more materials; predicting physical properties of a polymer composite material based on the formula based on the formula and a physical property prediction model; and outputting the physical properties of the polymer composite material.

[0010] The training of the physical property prediction model can be performed by the following steps: obtaining multiple training formulas and physical properties of multiple training polymer composites based on each formula in the multiple training formulas using a data set, wherein the training formula includes two or more training materials containing at least one polymer and a mixing ratio of each material in the two or more training materials; and performing training of the physical property prediction model to predict the physical properties of the training polymer composite based on each formula in the multiple training formulas based on the physical properties of the training materials included in each formula in the multiple training formulas and the physical properties of the multiple training polymer composites.

[0011] Among them, the step of performing training of the physical property prediction model may include the following steps: extracting a pre-set first training specific physical property from each formula of the multiple training formulas, that is, extracting a pre-set second training specific physical property from the physical properties of each material in at least one training specific material, and extracting a pre-set second training specific physical property from the physical properties of the training polymer composite material; and for each formula of the multiple training formulas, performing training of the physical property prediction model based on the multiple training formulas including the multiple first training specific physical properties and the second training specific physical property, the specific materials including at least one material selected from polymers, talc and polyolefin elastomers.

[0012] Among them, the step of predicting the physical properties of the polymer composite material may include the following steps: extracting at least one specific physical property from the multiple physical properties of each material in at least one specific material among the two or more materials; and inputting the at least one specific physical property of the two or more materials and the at least one specific material into the physical property prediction model to predict the physical properties of the polymer composite material.

[0013] According to another embodiment, a method for predicting characteristics of a polymer composite material may be provided, comprising the following steps: inputting a formula comprising two or more materials including at least one polymer and a mix ratio of each of the two or more materials; predicting at least one property of each of the two or more materials based on the two or more materials and a physical property prediction model; predicting physical properties of a polymer composite material based on the formula based on at least one property of each of the two or more materials and the physical property prediction model; and outputting physical properties of the polymer composite material.

[0014] The training of the physical property prediction model can be performed by the following steps: obtaining multiple training formulas and physical properties of multiple training polymer composite materials based on each formula in the multiple training formulas using a data set, wherein the training formula includes two or more training materials containing at least one polymer and a mixing ratio of each material in the two or more training materials; and performing training of the physical property prediction model based on the physical properties of the training materials included in each formula in the multiple training formulas, the attributes of the training materials and the physical properties of the multiple training polymer composite materials.

[0015] Among them, the step of performing training of the physical property prediction model may include the following steps: performing training to infer the properties of the training material included in each of the multiple training formulas based on the physical properties of the training material included in each of the multiple training formulas; and performing training to predict the physical properties of the training polymer composite material based on each of the multiple training formulas based on the properties of the training material included in each of the multiple training formulas.

[0016] Among them, the step of performing training of the physical property prediction model may further include the following steps: extracting a preset first training specific physical property from each formula of the multiple training formulas, that is, extracting a preset second training specific physical property from the physical properties of each material in at least one training specific material; extracting a preset second training specific physical property from the physical properties of the training polymer composite material; in the step of performing training to infer the properties of the training material, performing training to infer the properties of the training material included in each formula of the multiple training formulas based on the first training specific physical property included in each formula of the multiple training formulas; in the step of performing training to predict the physical properties of the training polymer composite material, performing training to predict the second training specific physical property of the training polymer composite material based on the properties of the training material included in each formula of the multiple training formulas, wherein the specific material includes at least one material selected from polymer, talc and polyolefin elastomer.

[0017] Among them, the step of predicting the physical properties of the polymer composite material may include the following steps: extracting at least one specific physical property from the multiple physical properties of each material in at least one specific material among the two or more materials; and inputting the at least one specific physical property of the two or more materials and the at least one specific material into the physical property prediction model to predict the physical properties of the polymer composite material.

[0018] According to another embodiment, a device for predicting the characteristics of a polymer composite material can be provided, comprising: an information input unit, which inputs a formula including two or more materials containing at least one polymer and a mixing ratio of each of the two or more materials; an information prediction unit, which predicts the physical properties of a polymer composite material based on the formula based on the formula and a physical property prediction model; and a result output unit, which outputs the physical properties of the polymer composite material.

[0019] Wherein, the device for predicting the characteristics of polymer composite materials may further include a model training unit, which uses a data set to obtain multiple training formulas and physical properties of multiple training polymer composite materials based on each formula in the multiple training formulas, wherein the training formula includes two or more training materials containing at least one polymer and the mixing ratio of each material in the two or more training materials. The model training unit performs training of the physical property prediction model to predict the physical properties of the training polymer composite material based on each formula in the multiple training formulas based on the physical properties of the training materials included in each formula in the multiple training formulas and the physical properties of the multiple training polymer composite materials. The physical property prediction model is trained by the model training unit.

[0020] In which, the model training unit can extract from each of the multiple training formulas, that is, it can extract a pre-set first training specific property from the physical properties of each material in at least one training specific material, and extract a pre-set second training specific property from the physical properties of the training polymer composite material. The model training unit can perform training of the physical property prediction model for each of the multiple training formulas based on the multiple training formulas including the multiple first training specific properties and the second training specific property. The specific material may include at least one material selected from polymers, talc and polyolefin elastomers.

[0021] In which, the information prediction unit can extract at least one specific physical property from the multiple physical properties of each material in at least one specific material among the two or more materials, and the information prediction unit can input the at least one specific physical property of the two or more materials and the at least one specific material into the physical property prediction model to predict the physical properties of the polymer composite material.

[0022] According to another embodiment, a device for predicting the characteristics of a polymer composite material can be provided, comprising: an information input unit, which inputs a formula including two or more materials containing at least one polymer and the mixing ratio of each of the two or more materials; an information prediction unit, which predicts at least one property of each of the two or more materials based on the two or more materials and a physical property prediction model, and predicts the physical properties of the polymer composite material based on the formula based on at least one property of each of the two or more materials and the physical property prediction model; and a result output unit, which outputs the physical properties of the polymer composite material.

[0023] The physical property prediction model may include a physical property-property prediction model for predicting the at least one property and a property-property prediction model for predicting the physical property of the polymer composite material.

[0024] In which, the device for predicting the characteristics of polymer composite materials may further include a model training unit, which uses a data set to obtain multiple training formulas and physical properties of multiple training polymer composite materials based on each formula in the multiple training formulas, wherein the training formulas include two or more training materials containing at least one polymer and the mixing ratio of each material in the two or more training materials. The model training unit performs training of the physical property prediction model based on the physical properties of the training materials included in each formula in the multiple training formulas, the attributes of the training materials and the physical properties of the multiple training polymer composite materials, and the physical property prediction model is trained by the model training unit.

[0025] In which, the model training unit can train the physical property-attribute prediction model to infer the attributes of the training material included in each of the multiple training formulas based on the physical properties of the training material included in each of the multiple training formulas, and the model training unit can train the attribute-physical property prediction model to predict the physical properties of the training polymer composite material based on each of the multiple training formulas based on the attributes of the training material included in each of the multiple training formulas.

[0026] In which, the model training unit can extract from each of the multiple training formulas, that is, extract a pre-set first training specific physical property from the physical properties of each material in at least one training specific material, and extract a pre-set second training specific physical property from the physical properties of the training polymer composite material. The model training unit can train the physical property-attribute prediction model to infer the properties of the training material included in each of the multiple training formulas based on the first training specific physical property included in each of the multiple training formulas. The model training unit can train the attribute-physical property prediction model to predict the second training specific physical property of the training polymer composite material based on each of the multiple training formulas based on the properties of the training material included in each of the multiple training formulas. The specific material may include at least one material selected from polymers, talc and polyolefin elastomers.

[0027] In which, the information prediction unit can extract at least one specific physical property from the multiple physical properties of each material in at least one specific material among the two or more materials, and the information prediction unit can input the at least one specific physical property of the two or more materials and the at least one specific material into the physical property prediction model to predict the physical properties of the polymer composite material.

[0028] (3) Beneficial effects

[0029] According to various embodiments, the method and apparatus for predicting characteristics of a polymer composite material can provide an environment for quickly determining changes in the physical properties of a polymer composite material generated during synthesis by simply changing the materials and the material ratio.

[0030] According to various embodiments, the method and apparatus for predicting the characteristics of polymer composite materials can predict the physical properties of the polymer composite material generated during synthesis based on the mix ratio of materials included in the recipe without performing the material synthesis and physical property measurement process of the polymer composite material, thereby greatly reducing the time and cost required for material development. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 is a diagram showing a configuration of an apparatus according to one embodiment.

[0032] Figure 2 is a flow chart illustrating an operational flow of predicting characteristics of a polymer composite material in an apparatus according to one embodiment.

[0033] Figure 3is a diagram schematically illustrating the operation of a physical property prediction model in an apparatus according to one embodiment, in which physical properties of materials included in a recipe are input to output physical properties of a polymer composite material.

[0034] Figure 4 3 is a flowchart illustrating the operational process of training a physical property prediction model based on a deep learning algorithm in an apparatus according to one embodiment.

[0035] Figure 5 is a flow chart illustrating an operational flow of predicting characteristics of a polymer composite material in an apparatus according to one embodiment.

[0036] Figure 6 is a diagram schematically illustrating the operation of a physical property prediction model in an apparatus according to one embodiment, wherein physical properties of materials included in a recipe are input to predict properties of the materials, and the properties of the materials are input to output physical properties of a polymer composite material.

[0037] Figure 7 3 is a flowchart illustrating the operational process of training a physical property prediction model based on a deep learning algorithm in an apparatus according to one embodiment.

[0038] Figure 8 3 is a flowchart illustrating a detailed operation process of training a physical property prediction model based on a deep learning algorithm in an apparatus according to one embodiment.

[0039] Description of reference numerals:

[0040] 100: Installation

[0041] 110: Processing Department

[0042] 120: Storage

[0043] 111: Information Acquisition Department

[0044] 113: Information Forecasting Department

[0045] 115: Result output unit

[0046] 117: Model Training Department

[0047] 119: Physical Property Extraction Department DETAILED DESCRIPTION

[0048] The following describes the embodiments in detail with reference to the accompanying drawings. However, the embodiments may be modified in various ways, and the scope of the present invention is not limited or restricted by these embodiments. It should be understood that any modifications, equivalents, or alternatives of the embodiments are included in the scope of the invention.

[0049] When a component is described as being “connected,” “coupled,” or “coupled” to another component, it should be understood that the component may be directly connected or coupled to another component, but other components may also be “connected,” “coupled,” or “coupled” between the components.

[0050] In addition, when describing each component of the embodiment, "or" can refer to each component, can refer to two or more components in the component, or can refer to all components. For example, the expression "a, b, or c" should be understood to mean any one of "a", "b", "c", "a and b", "a and c", "b and c", or "a, b, and c".

[0051] Components included in any embodiment and components with the same functions are described in other embodiments using the same names. Unless otherwise stated, the description of any embodiment is applicable to other embodiments, and detailed description is omitted within the scope of repetition.

[0052] The "data" processed by a device or a device can be represented by the term "information." Here, information can be used as a concept that includes data.

[0053] The present disclosure relates to a method and apparatus for predicting characteristics of a polymer composite material. More specifically, a method and apparatus for predicting physical properties of a polymer composite material generated (or synthesized) based on a material input recipe may be described.

[0054] According to various embodiments, the operation of predicting the characteristics of a polymer composite material can be performed based on at least one deep learning algorithm. More specifically, the operation of predicting the characteristics of a polymer composite material can be based on at least one deep learning model. When a recipe is input, the physical properties of the predicted polymer composite material can be predicted based on the mix ratio and physical properties of the materials.

[0055] According to one embodiment, the deep learning model may include a physical property prediction model, which is trained to predict the physical properties of a polymer composite material generated based on the physical properties of various materials including at least one polymer and the mixing ratio of the materials.

[0056] Here, the polymer composite material may be at least a portion of various polymer compositions such as polymer blends, polymer copolymers, polymer nanocomposites, polymer interpenetrating networks (IPNs), or polymer metal composites, and may be a material generated using at least one polymer.

[0057] In addition, the polymer as the material of the polymer composite material may include at least one polymer among polypropylene (PP), polyethylene (PE), and polyethylene terephthalate (PET).

[0058] Here, as described above, the polymer may include not only normal (or virgin) polymers but also at least one of various recycled polymers such as recycled polypropylene (rPP), recycled polyethylene (rPE), and recycled polyethylene terephthalate (rPET).

[0059] The preferred embodiments are described below with reference to the accompanying drawings. The purpose of the accompanying drawings is to explain the technical ideas of the present disclosure in detail together with the detailed description. The present disclosure should not be interpreted as being limited to the matters shown in the drawings.

[0060] Figure 1 : is a diagram showing the configuration of an apparatus according to one embodiment. More specifically, Figure 1 As shown, the detailed configuration of the device may be distinguished by function and represented in the form of a block diagram.

[0061] First, refer to Figure 1 In order to predict the physical properties of a polymer composite material based on an input formula using an input formula, the device 100 may include: an information acquisition unit 111, which acquires a formula for synthesizing the polymer composite material, and acquires a formula including multiple materials and the physical properties of each material and the material ratio from the acquired formula; an information prediction unit 113, which predicts the physical properties of the polymer composite material based on the acquired physical properties and the ratio of the materials; and a result output unit 115, which outputs the predicted physical properties of the polymer composite material.

[0062] Here, the information prediction unit 113 can predict the physical properties of the polymer composite material based on the physical properties of the materials and the mixing ratio of the materials obtained from the recipe using the trained physical property prediction model 121 .

[0063] In addition, the apparatus 100 may further include a model training unit 117 for training the physical property prediction model 121 .

[0064] According to various embodiments, the apparatus 100 may include a storage unit 120 for storing a physical property prediction model 121. The storage unit 120 may store at least one data set 123 used for training the physical property prediction model 121.

[0065] Below, refer to Figures 2 to 8 Can be described based on Figure 1 Operation of the apparatus 100, wherein the apparatus 100 can predict characteristics of a polymer composite.

[0066] According to one embodiment, referring to Figures 2 to 4 , describing operations of predicting physical properties of a polymer composite material based on physical properties of materials included in a recipe using a physical property prediction model and operations of training the physical property prediction model.

[0067] to this end, Figure 2 is a flow chart illustrating an operational flow of predicting characteristics of a polymer composite material in an apparatus according to one embodiment. Figure 3 is a diagram schematically illustrating the operation of a physical property prediction model in an apparatus according to one embodiment, wherein physical properties of materials included in a recipe are input to output physical properties of a polymer composite material. Figure 4 3 is a flowchart illustrating the operational process of training a physical property prediction model based on a deep learning algorithm in an apparatus according to one embodiment.

[0068] First, refer to Figure 2 In step 201 , the information acquisition unit 111 may input (or acquire) a formula including two or more materials including at least one polymer and a mixing ratio of each of the two or more materials.

[0069] According to one embodiment, the information acquiring unit 111 may acquire the specific recipe through at least one input unit (not shown) and / or a communication unit (not shown) constituting the apparatus 100 .

[0070] According to one embodiment, a recipe may include at least one polymer, a reinforcing agent (e.g., talc), an additive (e.g., polyolefin ether), and other materials used to generate a polymer composite material, and identification information corresponding to each material.

[0071] At this time, when the recipe includes identification information corresponding to each material, the information acquisition part 111 may acquire material information corresponding to each identification information from the storage part 120 .

[0072] To this end, the storage unit 120 may store material information of the polymer composite material, identification information corresponding to the material, and physical property information of each material in a data table.

[0073] According to one embodiment, the physical properties of each material stored in the storage unit 120 or the recipe may include at least one of melt index, tensile strength, tensile break, tear strength, yield strength, flexural modulus, impact strength, elongation at break, heat distortion temperature, air permeability, and shrinkage, and corresponding physical property values.

[0074] More specifically, impact strength may include dart impact strength and / or IZOD (Izod beam) impact strength. Furthermore, at least some of tear strength, yield strength, tensile strength, and elongation at break may include machine direction properties and / or transverse direction properties.

[0075] When describing an embodiment of the present invention, the expression "physical properties of a specific material" may include the physical properties (or physical property items) and physical property values of the specific material, for example, the physical property items (e.g., tensile strength) and physical property values (tensile strength value of the specific polypropylene) of a specific material (e.g., a specific polypropylene).

[0076] In addition, according to one embodiment, the mix ratio is the relative weight ratio of each material (eg, weight ratio sum (WRS)), and the content of each material can be expressed as a percentage (%).

[0077] In addition, the recipe may include physical property information of each material included in the recipe. However, the physical property of each material is stored in the storage unit 120 , and the information acquisition unit 111 may acquire the physical property of each material from the storage unit 120 .

[0078] Reference Figure 3The materials, physical properties and mixing ratios of the materials obtained from the specific formula are described in more detail. The information acquisition unit 111 can obtain various materials including material 1, material 2 and material 3 for forming the polymer composite material from the obtained specific formula.

[0079] At this time, the information acquisition unit 111 may acquire identification codes corresponding to the materials from the recipe, and may acquire material information corresponding to each identification code from the storage unit 120 .

[0080] Reference Figure 3 , shows three materials included in the recipe, such as material 1, material 2, and material 3, but is not limited thereto, and the recipe may include more or less of at least one material.

[0081] Furthermore, the information acquisition unit 111 may acquire the physical properties of each material included in the recipe.

[0082] For example, the information acquisition unit 111 can obtain physical properties including physical property value 1 of physical property 1, physical property value 2 of physical property 2, and physical property value 3 of physical property 3 from material 1, and can obtain physical properties including physical property value 4 of physical property 4, physical property value 5 of physical property 5, and physical property value 6 of physical property 6 from material 2, and can obtain physical properties including physical property value 7 of physical property 7, physical property value 8 of physical property 8, and physical property value 9 of physical property 9 from material 3.

[0083] Reference Figure 3 , showing that the physical properties of each material included in the formula include three physical properties, but is not limited thereto, and the physical properties of each material may include at least one more or less physical property.

[0084] Here, the description of physical properties 1 to 9 is intended to describe the physical properties of the material, and two or more physical properties from physical properties 1 to 9 may show the same physical property. Similarly, two or more mix ratios from mix ratio 1 to mix ratio 3 may show the same mix ratio.

[0085] In addition, the information acquisition unit 111 may acquire the mixing ratio of each material included in the recipe.

[0086] For example, the information acquisition unit 111 may acquire the mix ratio 1 of material 1, the mix ratio 2 of material 2, and the mix ratio 3 of material 3. More specifically, the mix ratios of the materials included in the recipe may include the proportion of each material. For example, when all materials are counted as 100%, the mix ratio of material 1 is 20%, the mix ratio of material 2 is 30%, and the mix ratio of material 3 is 50%.

[0087] In step 203 , the information acquisition unit 113 may predict the physical properties of the polymer composite material based on the recipe based on the recipe and the physical property prediction model.

[0088] According to one embodiment, the information prediction unit 113 can process multiple physical properties of each of two or more materials obtained based on the recipe as input to the physical property prediction model 121, and can obtain physical properties of the polymer composite material based on the recipe as output of the physical property prediction model 121.

[0089] Reference Figure 3 Describe in more detail, according to Figures 2 to 4 The physical property prediction model of the embodiment may be a physical property-physical property prediction model (hereinafter referred to as physical property prediction model 301), such as Figure 3 As shown, the physical property-physical property prediction model is trained to predict the physical properties of a polymer composite material based on a formulation based on the physical properties of the materials included in the formulation.

[0090] The information prediction unit 113 can process the materials included in the formula (for example, material 1, material 2 and material 3), the physical properties of the materials (for example, the physical properties of material 1 including physical property value 1 of physical property 1, physical property value 2 of physical property 2 and physical property value 3 of physical property 3, the physical properties of material 2 including physical property value 4 of physical property 4, physical property value 5 of physical property 5 and physical property value 6 of physical property 6, and the physical properties of material 3 including physical property value 7 of physical property 7, physical property value 8 of physical property 8 and physical property value 9 of physical property 9), and the physical ratios of the materials (for example, mix ratio 1 of material 1, mix ratio 2 of material 2 and mix ratio 3 of material 3) as inputs of the physical property prediction model 301, and obtain specific physical properties of the polymer composite material based on the formula (for example, physical properties including specific property value 1 of specific property 1, specific property value 2 of specific property 2 and specific property value 3 of specific property 3).

[0091] Here, the specific physical property may include at least one of melt index, tensile strength, tensile break, tear strength, yield strength, flexural modulus, impact strength, elongation at break, heat distortion temperature, air permeability, and shrinkage.

[0092] Reference Figure 3 The physical properties of the polymer composite material output by the physical property prediction model 301 are described as outputting three specific physical properties, but are not limited thereto. The physical properties of the polymer composite material output by the physical property prediction model 301 may include more or less of at least one specific property.

[0093] In execution according to Figure 2 and Figure 3 In an embodiment, the trained physical property prediction model 301 of step 203 may be a deep learning model that is trained to output specific physical properties of the polymer composite material based on the formulation based on the physical properties of the materials included in the formulation.

[0094] In this regard, refer to Figure 4 The training of the physical property prediction model 301 may be described in detail.

[0095] First, the physical property prediction model 301 can be constructed based on at least one of various deep learning algorithms for processing structured data, such as TabNet, XGBoost, LightGBM, CatBoost, and Deep Neural Network (DNN).

[0096] The physical property prediction model 301 may be a deep learning model configured to input a formula including multiple materials, the physical properties of each material, and the mixing ratio of the multiple materials, to output the physical properties of the polymer composite material corresponding to the formula.

[0097] In step 401, the model training unit 117 can obtain multiple training formulas and physical properties of multiple training polymer composite materials based on each formula in the multiple training formulas using a data set, wherein the training formulas include two or more training materials containing at least one polymer and a mixing ratio of each material in the two or more training materials.

[0098] More specifically, the dataset may include multiple training formulations and physical properties of training polymer composites based on each of the multiple training formulations. Each of the multiple formulations may include information about two or more materials, physical properties of each of the two or more materials, and a mix ratio of each of the two or more materials.

[0099] For example, a data set including information about physical properties of a specific training formulation and a specific training polymer composite material based on the specific training formulation is described. The data set may include training materials included in the specific training formulation (e.g., material 4, material 5, and material 6), physical properties of the training materials (e.g., properties of material 4 including property value 10 of property 10, property value 11 of property 11, and property value 12 of property 12, properties of material 5 including property value 13 of property 13, property value 14 of property 14, and property value 15 of property 15, and properties of material 6 including property value 16 of property 16, property value 17 of property 17, and property value 18 of property 18), and mix ratios of the training materials (e.g., mix ratio 4 of material 4, mix ratio 5 of material 5, and mix ratio 6 of material 6).

[0100] A specific training polymer composite based on a specific training formula is a polymer composite that has been determined to be capable of being synthesized using a specific training formula, and may include physical properties possessed by the specific training polymer composite (for example, properties including property value 19 of property 19, property value 20 of property 20, and property value 21 of property 21).

[0101] Here, the number of materials included in the training formula, the number of physical properties included in each material, and the number of physical properties of the training polymer composite are described as three each, but are not limited thereto and may vary depending on the setting.

[0102] In addition, the expressions of physical properties 10 to physical properties 18 are expressions used to describe the physical properties of the material. Two or more physical properties among physical properties 10 to physical properties 18 can show the same physical properties, and at least one physical property among physical properties 10 to physical properties 18 can show the same physical properties as at least one physical property among physical properties 1 to physical properties 9 and at least one physical property among physical properties 19 to physical properties 21.

[0103] Likewise, two or more mixes among mixes 4 to 6 may show mixes of the same proportion, and at least one mix among mixes 4 to 6 may have the same value as at least one mix among mixes 1 to 3.

[0104] Here, the information acquisition unit 111 may receive the dataset 123 and / or the data contained in the dataset (eg, a recipe and physical properties of a polymer composite material corresponding to the recipe) through an input unit or a communication unit.

[0105] Here, the physical properties of the polymer composite material may include at least a portion of melt index, tensile strength, tensile break, tear strength, yield strength, flexural modulus, impact strength, elongation at break, heat distortion temperature, air permeability, and shrinkage.

[0106] In step 403, the model training unit 117 may perform training of the physical property prediction model to predict the physical properties of the training polymer composite material based on each of the multiple training formulas based on the physical properties of the training materials included in each of the multiple training formulas and the physical properties of the multiple training polymer composite materials.

[0107] More specifically, the model training unit 117 can perform training of the physical property prediction model 301 to output the physical properties of the polymer composite material based on the training materials included in each of multiple training formulas, the physical properties of the training materials, and the physical properties of multiple training polymer composite materials when a formula is input.

[0108] When the model training unit 117 trains the physical property prediction model 301 , the physical property prediction model 301 may be trained based on the specific physical properties of at least a portion of the specific materials included in the training formulation and the specific physical properties of the training polymer composite material.

[0109] More specifically, the model training unit 117 can extract from each of the multiple training formulas, that is, extract the pre-set first training specific physical property from the physical properties of each material in at least one training specific material, and extract the pre-set second training specific physical property from the physical properties of the training polymer composite material.

[0110] To this end, the apparatus 100 may further include a physical property extraction unit 119 for extracting specific physical properties from the physical properties of the materials included in the recipe and / or the physical properties of the polymer composite material.

[0111] Here, the specific material may include a polymer, talc, or a polyolefin elastomer.

[0112] Here, the model training unit 117 and / or the physical property extraction unit 119 can extract the first training specific physical property or the second training specific physical property including at least a part of melt index, tensile strength, tensile break, tear strength, yield strength, flexural modulus, impact strength, elongation at break, heat distortion temperature, air permeability and shrinkage.

[0113] Here, the second specific physical property for training may include at least one specific physical property of the polymer composite material predicted by the trained physical property prediction model 121 .

[0114] Furthermore, the model training unit 117 and / or the physical property extraction unit 119 may extract different first specific physical properties for training based on the type of specific training material.

[0115] For example, when the specific training material is polypropylene (PP), the model training unit 117 and / or the physical property extraction unit 119 may extract at least one of melt index, flexural modulus, tensile strength, impact strength, shrinkage, and heat deformation temperature as the first feature.

[0116] When the specific training material is polyethylene (PE), the model training unit 117 and / or the physical property extraction unit 119 may extract at least one physical property of impact strength, tear strength, yield strength, tensile strength, and elongation at break as the second feature.

[0117] Furthermore, when the specific training material is polyethylene terephthalate (PET), the model training unit 117 and / or the physical property extraction unit 119 may extract at least one physical property of air permeability, tensile strength, tear strength, and elongation at break as the third feature.

[0118] Thereafter, the model training unit 117 may train the physical property prediction model for each of the plurality of training formulations based on the plurality of training formulations including the plurality of first training specific physical properties and the second training specific physical properties.

[0119] More specifically, the model training unit 117 can perform training of the physical property prediction model 301 to set the first training specific physical property and / or the second training specific physical property as the area of interest, input the specific training formula and the first training specific physical property of the specific training formula, to output the physical properties including the second training specific physical property of the specific polymer composite material corresponding to the specific training formula.

[0120] As described above, the information prediction unit 113 may perform an operation of predicting the physical properties of the polymer composite material based on the recipe using the trained physical property prediction model 301 .

[0121] Refer again Figure 2 The operation of predicting the physical properties of a polymer composite material based on an input recipe using the trained physical property prediction model 301 is described in detail.

[0122] For example, in step 203 , the information prediction unit 113 may process the two or more materials included in the recipe, the physical properties of each material, and the mixing ratio of each material as inputs to the physical property prediction model 301 , and obtain the physical properties of the polymer composite material from the physical property prediction model 301 .

[0123] According to various embodiments, the information prediction unit 113 may extract at least one specific physical property from multiple physical properties of each of at least one specific material among two or more materials included in the input recipe, and process it as an input to the physical property prediction model 301 .

[0124] At this time, the information prediction unit 113 may input the two or more materials and the at least one specific physical property of the at least one specific material into the physical property prediction model 301 to predict the physical properties of the polymer composite material.

[0125] At step 205, the result output unit 115 may output the physical properties of the polymer composite material. More specifically, the result output unit 115 may generate and output result information including the recipe obtained at step 201 and specific physical properties of the polymer composite material obtained from the physical property prediction model 301 based on the obtained recipe.

[0126] According to various embodiments, the result output unit 115 may include at least a portion of the chemical formula and / or at least a portion of the chemical structure of the polymer composite material predicted by the trained physical property prediction model 121 based on the recipe in the result information.

[0127] In addition, the result output unit 115 can display the portion corresponding to the predicted specific physical property of the polymer composite material through at least a portion of the chemical formula and / or at least a portion of the chemical structure. In this case, the result output unit 115 can highlight or display the portion corresponding to the specific physical property of the polymer composite material through color change.

[0128] According to various embodiments, the result output unit 115 may transmit the result information to other pre-set devices through the communication unit 130 .

[0129] The result output unit 115 performs the operation of step 205 to end Figure 2 Example of .

[0130] According to various embodiments, the physical properties of the polymer composite material may be predicted based on the physical properties of the materials included in the formulation, rather than being limited to the above-described physical properties. Alternatively, the physical properties of the polymer composite material may be predicted based on the properties of the materials.

[0131] In this regard, refer to Figures 5 to 8 The operation of predicting the physical properties of the polymer composite material based on the physical properties of the materials included in the recipe and the attributes of the materials using the physical property prediction model and the operation of training the physical property prediction model can be described.

[0132] to this end, Figure 5 is a flow chart illustrating an operational flow of predicting characteristics of a polymer composite material in an apparatus according to one embodiment. Figure 6 is a diagram schematically illustrating the operation of a physical property prediction model in an apparatus according to one embodiment, wherein physical properties of materials included in a recipe are input to predict properties of the materials, and the properties of the materials are input to output physical properties of a polymer composite material. Figure 7 3 is a flowchart illustrating the operational process of training a physical property prediction model based on a deep learning algorithm in an apparatus according to one embodiment. Figure 8 3 is a flowchart illustrating a detailed operation process of training a physical property prediction model based on a deep learning algorithm in an apparatus according to one embodiment.

[0133] First, refer to Figure 5 In step 501 , the information acquisition unit 111 may input (or acquire) a formula including two or more materials including at least one polymer and a mixing ratio of each of the two or more materials.

[0134] Unless otherwise specified, step 501 can be performed with Figure 2 At least a portion of step 201 is performed identically or similarly, and a detailed description thereof may be omitted.

[0135] Reference Figure 3 The materials, physical properties and mixing ratios of the materials obtained from a specific formula are described in detail, and the description of the operation of the information acquisition unit 111 obtaining various materials, physical properties and mixing ratios of the physical properties including material 1, material 2 and material 3 for forming a polymer composite material from a specific formula can be omitted.

[0136] Thereafter, the information prediction unit 113 may predict the properties of the material based on the physical properties of the material acquired from the recipe, and may predict the physical properties of the polymer composite material based on the properties of the material.

[0137] For this purpose, refer to Figure 6 Described in more detail, the physical property prediction model 121 may include a physical property-property prediction model (hereinafter referred to as the first prediction model 601) that inputs the physical properties of a material to predict the properties of the material and a property-property prediction model (hereinafter referred to as the second prediction model 603) that predicts specific physical properties of the polymer composite material based on the properties of the material.

[0138] In step 503 , the information prediction unit 113 may predict at least one property of each of the two or more materials based on the two or more materials and the first prediction model 601 .

[0139] According to one embodiment, the information prediction unit 113 can process multiple physical properties of each of the two or more materials obtained based on the recipe as input to the first prediction model 601, and obtain multiple attributes of each of the two or more materials from the first prediction model 601 as output of the first prediction model 601.

[0140] The information prediction unit 113 may include the materials included in the recipe (for example, material 1, material 2, and material 3), the physical properties of the materials (for example, the physical properties of material 1 including physical property value 1, physical property value 2 of physical property 2, and physical property value 3), the physical properties of material 2 including physical property value 4 of physical property 4, physical property value 5 of physical property 5, and physical property value 6 of physical property 6, and the physical properties of material 3 including physical property value 7 of physical property 7, physical property value 8 of physical property 8, and physical property value 9 of physical property 9), and the mixing ratio of the materials (for example, Mix 1 of material 1, mix 2 of material 2, and mix 3 of material 3) are processed as inputs of the first prediction model 601, and properties of the materials are obtained (for example, properties of material 1 including property value 1 of property 1, property value 2 of property 2, and property value 3 of property 3, properties of material 2 including property value 4 of property 4, property value 5 of property 5, and property value 6 of property 6, and properties of material 3 including property value 7 of property 7, property value 8 of property 8, and property value 9 of property 9).

[0141] Here, the material properties may include at least one of syndiotactic PP content or molecular weight, atactic PP content or molecular weight, rubber content or molecular weight, reinforcing material type or content, polyethylene (PE) content or molecular weight, copolymer type or content, long-chain branch content, acid content, density, and crystallinity.

[0142] Reference Figure 6 The properties of each material output by the first prediction model 601 are described as outputting three properties, but are not limited thereto. The properties of each material output by the first prediction model 601 may include at least one more or less property.

[0143] In step 505 , the information prediction unit 113 may predict the physical properties of the polymer composite material based on the recipe based on at least one attribute of each of the two or more materials and the physical property prediction model.

[0144] According to one embodiment, the information prediction unit 113 may process the material properties and material mix ratio obtained from the first prediction model 601 as inputs to the second prediction model 603 , and obtain physical properties of the polymer composite material based on the recipe as output from the second prediction model 603 .

[0145] The information prediction unit 113 can process the properties of the materials (for example, the properties of material 1 including attribute value 1 of attribute 1, attribute value 2 of attribute 2, and attribute value 3 of attribute 3, the properties of material 2 including attribute value 4 of attribute 4, attribute value 5 of attribute 5, and attribute value 6 of attribute 6, and the properties of material 3 including attribute value 7 of attribute 7, attribute value 8 of attribute 8, and attribute value 9 of attribute 9) as inputs of the second prediction model 603, and obtain specific physical properties of the polymer composite material based on the formula (for example, properties including specific property value 1 of specific property 1, specific property value 2 of specific property 2, and specific property value 3 of specific property 3).

[0146] Here, the specific physical property may include at least one of melt index, tensile strength, tensile break, tear strength, yield strength, flexural modulus, impact strength, elongation at break, heat distortion temperature, air permeability, and shrinkage.

[0147] Reference Figure 6 The physical properties of the polymer composite material output by the second prediction model 603 are described as outputting three specific physical properties, but are not limited thereto. The physical properties of the polymer composite material output by the second prediction model 603 may include more or less of at least one specific property.

[0148] In execution according to Figure 5 and Figure 6 In an embodiment, the first prediction model 601 trained in step 503 can be a deep learning model trained to output the properties of the material based on the physical properties of the materials included in the recipe, and the second prediction model 603 trained in step 505 can be a deep learning model that outputs specific physical properties of the polymer composite material based on the recipe based on the properties of the material.

[0149] In this regard, refer to Figure 7 and Figure 8 The training of the first prediction model 601 and the second prediction model 603 may be described in detail.

[0150] First, the first prediction model 601 and the second prediction model 603 can be constructed based on at least a portion of various deep learning algorithms for processing structured data, such as TabNet, XGBoost, LightGBM, CatBoost, and Deep Neural Network (DNN).

[0151] In step 701, the model training unit 117 can obtain multiple training formulas and physical properties of multiple training polymer composite materials based on each formula in the multiple training formulas using a data set, wherein the training formulas include two or more training materials containing at least one polymer and a mixing ratio of each material in the two or more training materials.

[0152] More specifically, the dataset may include multiple training formulations and physical properties of training polymer composite materials based on each of the multiple training formulations. Here, each of the multiple formulations may include information about two or more materials, physical properties of each of the two or more materials, attributes of each of the two or more materials, and a mix ratio of each of the two or more materials.

[0153] For example, a data set including information about physical properties of a specific training formulation and a specific training polymer composite material based on the specific training formulation is described. The data set may include training materials included in the specific training formulation (e.g., material 4, material 5, and material 6), physical properties of the training materials (e.g., physical properties including property value 10 of property 10, property value 11 of property 11, and property value 12 of property 12 of material 4, physical properties including property value 13, property value 14 of property 14, and property value 15 of property 15 of material 5, and physical properties including property value 16, property value 17 of property 18, and property value 19 of property 20). The properties of the training materials include property value 17 of property 17 and property value 18 of property 18), properties of the training materials (for example, properties of material 4 including property value 10 of property 10, property value 11 of property 11 and property value 12 of property 12, properties of material 5 including property value 13 of property 13, property value 14 of property 14 and property value 15 of property 15, and properties of material 6 including property value 16 of property 16, property value 17 of property 17 and property value 18 of property 18), and mix ratios of the training materials (for example, mix ratio 4 of material 4, mix ratio 5 of material 5, and mix ratio 6 of material 6).

[0154] A specific training polymer composite based on a specific training formula is a polymer composite that has been determined to be capable of being synthesized using a specific training formula, and may include physical properties possessed by the specific training polymer composite (for example, properties including property value 19 of property 19, property value 20 of property 20, and property value 21 of property 21).

[0155] Here, the number of materials included in the training formula, the number of physical properties included in each material, and the number of physical properties of the training polymer composite are described as three each, but are not limited thereto and may vary depending on the setting.

[0156] In addition, the expressions of physical properties 10 to physical properties 18 are expressions used to describe the physical properties of the material. Two or more physical properties among physical properties 10 to physical properties 18 can show the same physical properties, and at least one physical property among physical properties 10 to physical properties 18 can show the same physical properties as at least one physical property among physical properties 1 to physical properties 9 and at least one physical property among physical properties 19 to physical properties 21.

[0157] Likewise, two or more mixes among mixes 4 to 6 may show mixes of the same proportion, and at least one mix among mixes 4 to 6 may have the same value as at least one mix among mixes 1 to 3.

[0158] Here, the information acquisition unit 111 may receive the dataset 123 and / or the data contained in the dataset (eg, a recipe and physical properties of a polymer composite material corresponding to the recipe) through an input unit or a communication unit.

[0159] In step 703 , the model training unit 117 may train the physical property prediction model based on the physical properties of the training materials included in each of the plurality of training formulations, the attributes of the training materials, and the physical properties of the plurality of training polymer composite materials.

[0160] In this regard, refer to Figure 8 The operations of training the first prediction model 601 and the second prediction model 603 may be described in more detail. Figure 8 At least a portion of step 801 and / or step 803 may be Figure 7 Executed in step 703.

[0161] In step 801 , the model training unit 117 may perform training of the first prediction model 601 to infer properties of the training material included in each of the multiple training recipes based on the physical properties of the training material included in each of the multiple training recipes.

[0162] More specifically, the model training unit 117 may train the first prediction model 601 to output properties of the material included in a recipe when a recipe is input, based on the training materials and physical properties of the training materials included in each of the plurality of training recipes.

[0163] When the model training unit 117 trains the first prediction model 601 , the first prediction model 601 may be trained based on specific physical properties of at least a portion of specific materials included in the training recipe.

[0164] More specifically, the model training unit 117 can extract from each of the multiple training formulas, that is, extract the pre-set first training specific physical property from the physical properties of each material in at least one training specific material, and extract the pre-set second training specific physical property from the physical properties of the training polymer composite material.

[0165] To this end, the apparatus 100 may further include a physical property extraction unit 119 for extracting specific physical properties from the physical properties of the materials included in the recipe and / or the physical properties of the polymer composite material.

[0166] Here, the specific material may include a polymer, talc, or a polyolefin elastomer.

[0167] Here, the model training unit 117 and / or the physical property extraction unit 119 can extract the first training specific physical property or the second training specific physical property including at least a part of melt index, tensile strength, tensile break, tear strength, yield strength, flexural modulus, impact strength, elongation at break, heat distortion temperature, air permeability and shrinkage.

[0168] Here, the second specific physical property for training may include at least one specific physical property of the polymer composite material to be predicted by the trained physical property prediction model 121 .

[0169] Furthermore, the model training unit 117 and / or the physical property extraction unit 119 may extract different first specific physical properties for training based on the type of specific training material.

[0170] For example, when the specific training material is polypropylene (PP), the model training unit 117 and / or the physical property extraction unit 119 may extract at least one physical property of melt index, flexural modulus, tensile strength, impact strength, shrinkage, and heat deformation temperature as the first feature.

[0171] When the specific training material is polyethylene (PE), the model training unit 117 and / or the physical property extraction unit 119 may extract at least one physical property of impact strength, tear strength, yield strength, tensile strength, and elongation at break as the second feature.

[0172] Furthermore, when the specific training material is polyethylene terephthalate (PET), the model training unit 117 and / or the physical property extraction unit 119 may extract at least one physical property of air permeability, tensile strength, tear strength, and elongation at break as the third feature.

[0173] Thereafter, the model training unit 117 may perform training of the first prediction model 601 to infer the properties of the training material included in each of the plurality of training recipes based on the first training specific physical property included in each of the plurality of training recipes.

[0174] More specifically, the model training unit 117 can perform training of the first prediction model 601 to set the first training specific property as the focus area, input the specific training formula and the first training specific property of the specific training formula, and output the properties of at least a portion of the material including the specific material in the materials included in the specific training formula.

[0175] In step 803 , the model training unit 117 may perform training of the second prediction model 603 to predict the physical properties of the training polymer composite material based on each of the plurality of training formulas based on the properties of the training material included in each of the plurality of training formulas.

[0176] More specifically, the model training unit 117 can perform training of the second prediction model 603 to output the physical properties of the polymer composite material based on the formula when the physical properties of at least a portion of the materials included in a formula are input based on the training materials and properties of the training materials included in each formula of multiple training formulas.

[0177] When the model training unit 117 performs training of the second prediction model 603 , the second prediction model 603 may be trained based on properties of the plurality of materials and at least a portion of the specific materials included in the training recipe.

[0178] Thereafter, the model training unit 117 may perform training of the second prediction model 603 to predict the second training specific physical properties of the training polymer composite material based on each of the multiple training formulas based on the properties of the training material included in each of the multiple training formulas and the pre-set second training specific physical properties in the physical properties of the training polymer composite material.

[0179] More specifically, the model training unit 117 can perform training of the second prediction model 603 to set the second training specific property as the area of interest, input the specific training formula, the properties of at least a portion of the materials included in the specific training formula including the specific material, and the second training specific property of the specific training formula to output the physical properties including the second training specific property of the specific polymer composite material corresponding to the specific training formula.

[0180] As described above, the information prediction unit 113 may perform an operation of predicting the physical properties of the polymer composite material based on the recipe using the physical property prediction model 121 including the trained first prediction model 601 and the second prediction model 603 .

[0181] Refer again Figure 5 The operation of predicting the physical properties of the polymer composite material based on the input recipe using the trained first prediction model 601 and / or second prediction model 603 is described in detail.

[0182] For example, in step 503, the information prediction unit 113 may process two or more materials included in the recipe, the physical properties of each material, and the mixing ratio of each material as inputs to the first prediction model 601, and obtain the properties of at least part of the materials included in the recipe from the first prediction model 601.

[0183] Here, the information prediction unit 113 may extract at least one specific physical property from a plurality of physical properties of each of at least one specific material among the two or more materials included in the input recipe, and process the extracted specific physical property as an input to the first prediction model 601 .

[0184] At this time, the information prediction unit 113 may input the two or more materials and the at least one specific physical property of the at least one specific material into the first prediction model 601 to predict the properties of at least a portion of the materials included in the recipe.

[0185] Thereafter, as described above, the information prediction unit 113 may perform the operation of step 505 based on the properties of the material acquired from the first prediction model 601 and the input recipe.

[0186] At step 507, the result output unit 115 may output the physical properties of the polymer composite material. More specifically, the result output unit 115 may generate and output result information that includes the recipe acquired in step 201 and specific physical properties of the polymer composite material acquired from the physical property prediction model 301 based on the acquired recipe. Here, the result output unit 115 may further include predicted properties of the materials included in the recipe in the result information.

[0187] According to various embodiments, the result output unit 115 may include at least a portion of the chemical formula and / or at least a portion of the chemical structure of the polymer composite material predicted by the trained physical property prediction model 121 based on the recipe in the result information.

[0188] In addition, the result output unit 115 can display the portion corresponding to the predicted specific physical property of the polymer composite material through at least a portion of the chemical formula and / or at least a portion of the chemical structure. In this case, the result output unit 115 can highlight or display the portion corresponding to the specific physical property of the polymer composite material through color change.

[0189] According to various embodiments, the result output unit 115 may transmit the result information to other pre-set devices through the communication unit 130 .

[0190] The result output unit 115 performs the operation of step 507 to end Figure 5 Example of .

[0191] As described above, the method and apparatus for predicting characteristics of a polymer composite material can provide an environment in which changes in the physical properties of a polymer composite material generated during synthesis can be quickly determined simply by changing the materials and the material ratio.

[0192] According to various embodiments, the method and apparatus for predicting the characteristics of polymer composite materials can predict the physical properties of the polymer composite material generated during synthesis based on the mix ratio of materials included in the recipe without performing the material synthesis and physical property measurement process of the polymer composite material, thereby greatly reducing the time and cost required for material development.

[0193] According to the above description, two or more of the information acquisition unit 111, the information prediction unit 113, the result output unit 115, the model training unit 117, and the physical property extraction unit 119 can be configured as a single module, or each can be configured as a separate module. In this case, each module can include at least one processor.

[0194] According to various embodiments, functions of various embodiments described as being performed by the device 100 are operations processed by the processing section 110 of the device 100 and may be performed by being organically connected with components of the device 100 or other devices connected thereto.

[0195] For this reason, Figure 1 As shown, at least a portion of the information acquisition unit 111 , the information prediction unit 113 , the result output unit 115 , the model training unit 117 and the physical property extraction unit 119 may be included in the processing unit 110 .

[0196] The processing unit 110 may include at least one processor, and may process data received from a battery connected to the apparatus 100 by at least one program (an application, a tool, a plug-in, software, etc., hereinafter referred to as a battery diagnostic program).

[0197] The storage unit 120 may store various data processed by at least one component of the apparatus 100 (e.g., the processing unit 110 or the communication unit). For example, the data may include a program for processing control instructions, data processed by the program, or input data and output data related thereto.

[0198] The communication unit (not shown) may support establishing a wired communication channel within the device 100 and / or between the device 100 and at least one other device (eg, a user device or a server), establishing a wireless communication channel, and performing communication through the established communication channel.

[0199] The input / output portion may include or be connected to at least a part of an input portion (not shown) for inputting data such as a keyboard, mouse, touchpad, etc. and an output portion (not shown) for outputting data such as a display portion (e.g., a monitor), a speaker, a driving portion, etc.

[0200] According to various embodiments of the present invention, the device 100 or user device may include at least a portion of the functions within the scope of all information communication devices such as mobile communication terminals, multimedia terminals, wired terminals, fixed terminals and Internet Protocol (IP) terminals.

[0201] The apparatus 100 is an apparatus for processing control instructions, and may be configured to include at least a portion of the functions of a workstation or a large-capacity database, or be connected to a workstation or a large-capacity database through communication.

[0202] As described above, although the embodiments are described with limited drawings, a person skilled in the art may make various technical modifications and changes based on the various embodiments.

[0203] For example, even if the described techniques are performed in an order different from that of the described methods, or the described systems, structures, devices, circuits, and other components are combined or combined in a form different from that of the described methods, or are replaced or substituted with other components or equivalents, appropriate results can still be achieved.

[0204] In particular, when describing with reference to a flowchart, although it is described as including a plurality of steps and the plurality of steps are sequentially executed in a specified order, it is not necessarily limited to the described order.

[0205] In other words, changing or deleting at least a portion of the steps in the flowchart, or adding at least one step can also be applied as an embodiment, or executing one or more steps in the flowchart in parallel can also be applied as an embodiment. In other words, the steps are not necessarily limited to being operated in a chronological order, and this should also be included in the embodiments of the present disclosure.

[0206] Therefore, other embodiments, other examples, and equivalents thereof should also fall within the scope of the appended claims.

Claims

1. A method for predicting characteristics of a polymer composite material, comprising the steps of: Inputting a recipe including two or more materials including at least one polymer and a mix ratio of each of the two or more materials; Predicting the physical properties of a polymer composite material based on the formulation based on the formulation and the physical property prediction model; as well as The physical properties of the polymer composite are output.

2. The method for predicting characteristics of a polymer composite material according to claim 1, wherein: The training of the physical property prediction model is performed by the following steps: Acquiring physical properties of a plurality of training formulations and a plurality of training polymer composite materials based on each of the plurality of training formulations using a data set, wherein the training formulations include two or more training materials comprising at least one polymer and a mixing ratio of each of the two or more training materials; as well as The physical property prediction model is trained to predict the physical properties of the training polymer composite material based on each of the plurality of training formulations based on the physical properties of the training materials included in each of the plurality of training formulations and the physical properties of the plurality of training polymer composite materials.

3. The method for predicting characteristics of a polymer composite material according to claim 2, wherein: The step of executing the training of the physical property prediction model comprises the following steps: Extracting a preset first specific physical property for training from each of the plurality of training formulas, that is, extracting a preset second specific physical property for training from the physical properties of each of the at least one specific training material, and extracting a preset second specific physical property for training from the physical properties of the training polymer composite material; and For each of the plurality of training formulations, training of the physical property prediction model is performed based on the plurality of training formulations including the plurality of first training specific physical properties and the second training specific physical property. The specific material includes at least one of a polymer, talc and a polyolefin elastomer.

4. The method for predicting characteristics of a polymer composite material according to claim 1, wherein: The step of predicting the physical properties of the polymer composite material comprises the following steps: extracting at least one specific physical property from a plurality of physical properties possessed by each of at least one specific material among the two or more materials; and The two or more materials and the at least one specific physical property of the at least one specific material are input into the physical property prediction model to predict the physical properties of the polymer composite material.

5. A method for predicting characteristics of a polymer composite material, comprising the steps of: Inputting a recipe including two or more materials including at least one polymer and a mix ratio of each of the two or more materials; Predicting at least one property of each of the two or more materials based on the two or more materials and the physical property prediction model; Predicting physical properties of a polymer composite material based on the formulation based on at least one attribute of each of the two or more materials and the physical property prediction model; as well as The physical properties of the polymer composite are output.

6. The method for predicting characteristics of a polymer composite material according to claim 5, wherein: The training of the physical property prediction model is performed by the following steps: Acquiring physical properties of a plurality of training formulations and a plurality of training polymer composite materials based on each of the plurality of training formulations using a data set, wherein the training formulations include two or more training materials comprising at least one polymer and a mix ratio of each of the two or more training materials; and The physical property prediction model is trained based on the physical properties of the training material included in each of the plurality of training formulations, the attributes of the training material, and the physical properties of the plurality of training polymer composites.

7. The method for predicting characteristics of a polymer composite material according to claim 6, wherein: The step of executing the training of the physical property prediction model comprises the following steps: performing training to infer properties of the training material included in each of the plurality of training formulations based on physical properties of the training material included in each of the plurality of training formulations; and Training is performed to predict physical properties of a training polymer composite based on each of the plurality of training formulations based on properties of the training material included in each of the plurality of training formulations.

8. The method for predicting characteristics of a polymer composite material according to claim 7, wherein: The step of performing the training of the physical property prediction model further comprises the following steps: Extracting a preset first specific physical property for training from each of the plurality of training formulas, that is, extracting a preset first specific physical property for training from the physical properties of each material in at least one specific material for training, and extracting a preset second specific physical property for training from the physical properties of the polymer composite material for training, In the step of performing training to infer the property of the training material, the training is performed to infer the property of the training material included in each of the plurality of training formulas based on the first training specific physical property included in each of the plurality of training formulas. In the step of performing training to predict the physical property of the training polymer composite material, the training is performed to predict the second training specific physical property of the training polymer composite material based on each of the plurality of training formulations based on the properties of the training material included in each of the plurality of training formulations. The specific material includes at least one of a polymer, talc and a polyolefin elastomer.

9. The method for predicting characteristics of a polymer composite material according to claim 5, wherein: The step of predicting the physical properties of the polymer composite material comprises the following steps: extracting at least one specific physical property from a plurality of physical properties possessed by each of at least one specific material among the two or more materials; and The two or more materials and the at least one specific physical property of the at least one specific material are input into the physical property prediction model to predict the physical properties of the polymer composite material.

10. An apparatus for predicting characteristics of a polymer composite material, comprising: an information input unit for inputting a formula including two or more materials including at least one polymer and a mixing ratio of each of the two or more materials; an information prediction unit, configured to predict the physical properties of a polymer composite material based on the formula based on the formula and a physical property prediction model; as well as The result output unit outputs the physical properties of the polymer composite material.

11. The device for predicting characteristics of a polymer composite material according to claim 10, further comprising a model training unit, The model training unit obtains a plurality of training formulas and physical properties of a plurality of training polymer composite materials based on each of the plurality of training formulas using a data set, wherein the training formulas include two or more training materials containing at least one polymer and a mixing ratio of each of the two or more training materials. The model training unit performs training of the physical property prediction model to predict the physical properties of the training polymer composite material based on each of the plurality of training formulations based on the physical properties of the training materials included in each of the plurality of training formulations and the physical properties of the plurality of training polymer composite materials. The physical property prediction model is trained by the model training unit.

12. The device for predicting characteristics of a polymer composite material according to claim 11, wherein: The model training unit extracts a preset first specific physical property for training from each of the plurality of training formulas, that is, extracts a preset second specific physical property for training from the physical properties of each of the at least one specific material for training, and extracts a preset second specific physical property for training from the physical properties of the polymer composite material for training. The model training unit performs training of the physical property prediction model for each of the plurality of training formulations based on the plurality of training formulations including the plurality of first training specific physical properties and the second training specific physical property. The specific material includes at least one of a polymer, talc and a polyolefin elastomer.

13. The apparatus for predicting characteristics of a polymer composite material according to claim 10, wherein: The information prediction unit extracts at least one specific physical property from a plurality of physical properties possessed by each of at least one specific material among the two or more materials. The information prediction unit inputs the at least one specific physical property of the two or more materials and the at least one specific material into the physical property prediction model to predict the physical properties of the polymer composite material.

14. An apparatus for predicting characteristics of a polymer composite material, comprising: an information input unit for inputting a formula including two or more materials including at least one polymer and a mixing ratio of each of the two or more materials; an information prediction unit, configured to predict at least one property of each of the two or more materials based on the two or more materials and a physical property prediction model, and to predict physical properties of a polymer composite material based on the recipe based on the at least one property of each of the two or more materials and the physical property prediction model; as well as The result output unit outputs the physical properties of the polymer composite material.

15. The apparatus for predicting characteristics of a polymer composite material according to claim 14, wherein: The physical property prediction model includes a physical property-property prediction model for predicting the at least one property and a property-property prediction model for predicting the physical property of the polymer composite material.

16. The apparatus for predicting characteristics of a polymer composite material according to claim 15, further comprising a model training unit, The model training unit obtains a plurality of training formulas and physical properties of a plurality of training polymer composite materials based on each of the plurality of training formulas using a data set, wherein the training formulas include two or more training materials containing at least one polymer and a mixing ratio of each of the two or more training materials. The model training unit performs training of the physical property prediction model based on the physical properties of the training material included in each of the plurality of training formulations, the attributes of the training material, and the physical properties of the plurality of training polymer composite materials. The physical property prediction model is trained by the model training unit.

17. The apparatus for predicting characteristics of a polymer composite material according to claim 16, wherein: The model training unit trains the physical property-attribute prediction model to infer the attribute of the training material included in each of the plurality of training recipes based on the physical property of the training material included in each of the plurality of training recipes. The model training unit trains the attribute-physical property prediction model to predict physical properties of a training polymer composite material based on each of the plurality of training formulations based on the attributes of the training material included in each of the plurality of training formulations.

18. The apparatus for predicting characteristics of a polymer composite material according to claim 17, wherein: The model training unit extracts a preset first specific physical property for training from each of the plurality of training formulas, that is, extracts a preset second specific physical property for training from the physical properties of each of the at least one specific material for training, and extracts a preset second specific physical property for training from the physical properties of the polymer composite material for training. The model training unit trains the physical property-attribute prediction model to infer the attribute of the training material included in each of the plurality of training recipes based on the first training specific physical property included in each of the plurality of training recipes. The model training unit trains the attribute-physical property prediction model to predict the second training specific physical property of the training polymer composite material based on each of the plurality of training formulations based on the attribute of the training material included in each of the plurality of training formulations. The specific material includes at least one of a polymer, talc and a polyolefin elastomer.

19. The apparatus for predicting characteristics of a polymer composite material according to claim 14, wherein: The information prediction unit extracts at least one specific physical property from a plurality of physical properties possessed by each of at least one specific material among the two or more materials. The information prediction unit inputs the at least one specific physical property of the two or more materials and the at least one specific material into the physical property prediction model to predict the physical properties of the polymer composite material.

Images