Raw material prediction device, inference device, machine learning device, raw material prediction method, inference method, and machine learning method

Abstract

This invention provides a raw material prediction device that enables the appropriate prediction of raw material information from the physical properties of a molded product. [Solution] The information processing device 5, which functions as a raw material prediction device, comprises a data acquisition unit 5020 that acquires input data, and a data generation unit 5021 that generates output data for the input data acquired by the data acquisition unit 5020 by inputting the input data into a learning model 16. The learning model 16 is a trained model that has learned the correlation between input data and output data by machine learning. The input data includes physical property information 14 that shows the physical properties of a molded body obtained by melt-molding solid raw materials, and the output data includes raw material information 10 related to the raw materials.

Description

Technical Field

[0004] ,

[0006] , , , , , ,

[0005] , , , , , ,

[0001] The present invention relates to a raw material prediction device, an inference device, a machine learning device, a raw material prediction method, an inference method, and a machine learning method.

Background Art

[0002] Conventionally, by melt-molding a polymer compound such as polyethylene as a raw material, molded articles of various shapes have been manufactured (see, for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] The polymer compounds used as raw materials for molded articles are diverse, and the physical property values of molded articles obtained by melt-molding from raw materials differ depending on the type of raw material. Therefore, it has been difficult to predict information on raw materials from the physical property values of molded articles obtained by melt-molding.

[0005] The present invention has been made to solve the above problems, and an object thereof is to provide a raw material prediction device, an inference device, a machine learning device, a raw material prediction method, an inference method, and a machine learning method that enable appropriate prediction of information on raw materials from the physical property values of molded articles.

Means for Solving the Problems

[0006] To achieve the above object, a raw material prediction device according to one aspect of the present invention includes: a data acquisition unit that acquires input data; The system includes a data generation unit that inputs the input data acquired by the data acquisition unit into a learning model and generates output data for said input data, The aforementioned learning model, This is a trained model that has learned the correlation between the input data and the output data using machine learning. The aforementioned input data is Includes physical property information indicating the physical properties of a molded body obtained by melting and molding solid raw materials, The output data mentioned above is: This includes raw material information relating to the aforementioned raw materials. [Effects of the Invention]

[0007] According to one aspect of the present invention, a raw material prediction device is used to generate output data containing raw material information based on the input data, which includes physical property information indicating the physical properties of a molded body obtained by melt-molding solid raw materials. This input data is then used to generate output data containing raw material information. Therefore, it is possible to appropriately predict information about raw materials from the physical properties of the molded body.

[0008] Other issues, configurations, and effects will be clarified in the embodiments for carrying out the invention described later. [Brief explanation of the drawing]

[0009] [Figure 1] This is an overall schematic diagram showing an example of the raw material prediction system 1. [Figure 2] This is a block diagram showing an example of an information processing device 5. [Figure 3] This is a data structure diagram showing an example of the information management database 510. [Figure 4A] This is a data structure diagram showing an example of molecular chain information 11 and crystallinity information 12. [Figure 4B] This is a data structure diagram showing an example of raw material information 10, process information 13, and physical property information 14. [Figure 5] This is a functional diagram illustrating an example of the learning function 501. [Figure 6A]It is a diagram showing the first learning data 15-1 and the first learning model 16-1. [Figure 6B] It is a diagram showing the second learning data 15-2 and the second learning model 16-2. [Figure 6C] It is a diagram showing the third learning data 15-3 and the third learning model 16-3. [Figure 6D] It is a diagram showing the fourth learning data 15-4 and the fourth learning model 16-4. [Figure 7] It is a functional explanatory diagram showing an example of the raw material prediction function 502. [Figure 8] It is a hardware configuration diagram showing an example of the computer 900. [Figure 9] It is a flowchart showing an example of the machine learning method by the learning function 501. [Figure 10] It is a flowchart showing an example of the raw material prediction method by the raw material prediction function 502. [Figure 11] It is a screen configuration diagram showing an example of the raw material prediction input screen 17. [Figure 12] It is a screen configuration diagram showing an example of the raw material prediction output screen 18.

Embodiments for Carrying Out the Invention

[0010] Hereinafter, embodiments for carrying out the present invention will be described with reference to the drawings. Hereinafter, the range necessary for the description for achieving the object of the present invention will be schematically shown, and the range necessary for the description of the relevant part of the present invention will be mainly described, and the parts where the description is omitted will be based on the known technology.

[0011] FIG. 1 is an overall schematic diagram showing an example of the raw material prediction system 1. The raw material prediction system 1 is a system for predicting information on raw materials from the physical property values of a molded body obtained by melt-molding a solid raw material.

[0012] The raw material is in solid form, for example, processed into pellet form. The raw material is a polymer compound, and examples include polyolefin resins such as polyethylene and polypropylene, polyester resins such as polyethylene terephthalate, polystyrene resins, polyamide resins, acrylic resins, polyvinyl chloride resins, polyacrylonitrile resins, polyvinyl alcohol resins, and thermoplastic resin materials such as biodegradable resins. The raw material contains many molecular chains of the polymer compound and has a crystalline portion in which the molecular chains are arranged regularly and an amorphous portion in which the molecular chains are arranged irregularly. In this embodiment, the case where the raw material is polyethylene will be described in detail.

[0013] The raw materials may be a mixture of multiple types of raw materials in a predetermined ratio, or recycled materials that have undergone recycling processing. Various information regarding the raw materials is recorded as raw material information 10 and used in the raw material prediction system 1.

[0014] The molded article is a molded piece having a predetermined shape or a film having a predetermined thickness. The molded piece may be, for example, a rod or a plate, and is mainly formed by injection molding or compression molding. The shape of the molded piece is preferably one suitable for analysis or testing, but it may also be a shape similar to the final product, such as a bottle, and may be formed by blow molding or the like. The film may be, for example, a membrane, and is mainly formed by extrusion molding. Furthermore, the molded piece or film may be formed by multiple molding processes.

[0015] The raw material prediction system 1 comprises an analysis device 2, a molding device 3, a testing device 4, an information processing device 5, and a user terminal device 6. Each of the devices 2 to 6 is, for example, composed of a general-purpose or dedicated computer (see Figure 8 below) and connected to a wired or wireless network 7, enabling the mutual transmission and reception of various types of data. The number of each device 2 to 6 and the connection configuration of the network 7 are not limited to the example in Figure 1 and may be changed as appropriate.

[0016] Analytical apparatus 2 is a device used to analyze raw materials. Multiple types of analytical apparatus 2 (2A to 2C) with different analytical methods are used. These multiple types of analytical apparatus 2 include melt analysis apparatus 2A for melt analysis of raw materials, solution analysis apparatus 2B for solution analysis of raw materials, and solid analysis apparatus 2C for solid analysis of raw materials.

[0017] Melt analysis apparatus 2A is an apparatus that analyzes the characteristics of raw materials in a molten state. Examples of melt analysis apparatus 2A include melt flow rate (MFR) analyzers, viscometers, and differential scanning calorimetry (DSC) analyzers. Solution analysis apparatus 2B is an apparatus that analyzes the characteristics of raw materials using a solution obtained by dissolving the raw materials in a solvent. Examples of solution analysis apparatus 2B include gel permeation chromatography (GPC) analyzers and nuclear magnetic resonance (NMR) analyzers. Solid analysis apparatus 2C is an apparatus that analyzes the characteristics of raw materials in a solid state. Examples of solid analysis apparatus 2C include dry densimeters and density gradient tube analyzers.

[0018] The analysis results from the analyzer 2 are recorded as molecular chain information 11 regarding the molecular chains of the polymer compound and crystallinity information 12 regarding the crystalline portion of the polymer compound, and are used in the raw material prediction system 1. The analysis conditions when analyzing the raw materials are set according to standards such as JIS, and the analysis conditions may be recorded in the molecular chain information 11 and crystallinity information 12.

[0019] The molding apparatus 3 is a device that performs a molding process in which raw materials are melt-molded into a molded body. Multiple types of molding apparatus 3 with different molding processes can be used. These multiple types of molding apparatus 3 include a molded piece molding apparatus 3A that melt-moldes raw materials into molded pieces, and a film molding apparatus 3B that melt-moldes raw materials into films.

[0020] Various information regarding the molding process by the molding apparatus 3 is recorded as process information 13 and used by the raw material prediction system 1. At that time, various information regarding the molding process by the molded piece molding apparatus 3A is recorded as molded piece process information 13A. Various information regarding the molding process by the film molding apparatus 3B is recorded as film process information 13B.

[0021] Test apparatus 4 is a device for performing tests on the durability and heat resistance of molded articles. Multiple types of test apparatus 4 with different test methods are used. These multiple types of test apparatus 4 include molded article test apparatus 4A for testing molded articles and film test apparatus 4B for testing films. Molded article test apparatus 4A performs tests on molded articles regarding environmental stress crack resistance (ESCR), flexural modulus, tensile modulus, Charpy impact strength, tensile elongation at break, Vicat softening temperature, strength characteristics, contents characteristics, usability, and sealing performance. Film test apparatus 4B performs tests on films regarding impact burst strength, tensile impact strength, tensile modulus, thermal shrinkage rate, tear strength, loop stiffness, drop characteristics, ease of opening, usability, and line suitability.

[0022] Various information regarding the test results from the test apparatus 4 is recorded as physical property information 14, which indicates the physical properties of the molded body, and is used in the raw material prediction system 1. At that time, various information regarding the test results of the molded piece from the molded piece test apparatus 4A is recorded as molded piece physical property information 14A. Various information regarding the test results of the film from the film molding apparatus 3B is recorded as film physical property information 14B. The test conditions when testing the molded body are set according to standards such as JIS, and the test conditions may be recorded in the molded piece physical property information 14A and the film physical property information 14B.

[0023] The information processing device 5 is equipped with an information management database 510 that can register various types of information 10-14 (raw material information 10, molecular chain information 11, crystallinity information 12, process information 13, and physical property information 14) associated with past analyses of raw materials and tests of molded products obtained by melt molding from those raw materials. The information processing device 5 then generates a learning model 16 by performing machine learning based on the various types of information 10-14 registered in the information management database 510.

[0024] Furthermore, the information processing device 5 uses a learning model 16 to predict information about new raw materials for obtaining a new molded body by melt molding, based on input data including physical property information 14 that indicates the physical properties of the new molded body to be predicted, as raw material information 10. In this case, the information processing device 5 may also predict molecular chain information 11, crystallinity information 12, and process information 13 as raw material information, in addition to raw material information 10.

[0025] The user terminal device 6 is used by users such as workers performing analysis, melt molding, and testing, or administrators of the raw material prediction system 1. The user terminal device 6 has various information management programs and a web browser installed, accepts various input operations, and outputs various information via a display screen and audio. The user terminal device 6 is used, for example, for users to input physical property information 14 as input data, or to output raw material information 10 (output data), which is the prediction result of the information processing device 5.

[0026] (Configuration of Information Processing Device 5) Figure 2 is a block diagram showing an example of an information processing device 5. The information processing device 5 comprises a control unit 50 composed of a processor, a storage unit 51 composed of an HDD, SSD, memory, etc., a communication unit 52 which is a communication interface with a network 7 and external devices, an input unit 53 composed of a keyboard, mouse, etc., and a display unit 54 composed of a display, etc. Note that the input unit 53 and the display unit 54 may be omitted.

[0027] The memory unit 51 stores the information management database 510, the trained model management database 511, and the information processing program 512, as well as the operating system, other programs, data, etc.

[0028] The control unit 50 implements an information management function 500, a learning function 501, and a raw material prediction function 502 by executing an information processing program 512 stored in the memory unit 51. The control unit 50 includes a learning data acquisition unit 5010 and a machine learning unit 5011 as parts that implement the learning function 501. The control unit 50 includes a data acquisition unit 5020, a data generation unit 5021, and an output processing unit 5022 as parts that implement the raw material prediction function 502.

[0029] The following describes the data structure of each function 500-502 and each database 510 and 511.

[0030] (Information management function 500) The control unit 50 of the information processing device 5 implements the information management function 500 using the information management database 510. By providing the information management function 500 to the information processing device 5, the information processing device 5 operates as an information management device.

[0031] Figure 3 is a data structure diagram showing an example of the information management database 510. Figure 4A is a data structure diagram showing an example of molecular chain information 11 and crystallinity information 12. Figure 4B is a data structure diagram showing an example of raw material information 10, process information 13, and physical property information 14.

[0032] The information management database 510 has multiple records for each management ID used to associate various types of information 10-14 handled by the raw material prediction system 1. The management ID is information used to identify raw materials and molded products (such as part numbers and model numbers). Each record has fields that can register, for example, raw material information 10, molecular chain information 11, crystallinity information 12, process information 13, and physical property information 14.

[0033] The information management function 500 registers various types of information 10-14 in the information management database 510 when it receives, for example, raw material information 10 from the user terminal device 6, molecular chain information 11 and crystallinity information 12 from the analyzer 2, process information 13 from the molding device 3, or physical property information 14 from the test device 4. Furthermore, when the information management function 500 receives raw material information 10 as a result of processing by the raw material prediction function 502, it registers that raw material information 10 in the information management database 510.

[0034] The various types of information 10-14 registered in the information management database 510 can be accessed from the user terminal device 6. Furthermore, editing operations such as adding, deleting, and modifying the various types of information 10-14 can be performed on the display screen shown by the information management program or web browser of the user terminal device 6. When the information management function 500 receives the results of editing operations on the various types of information 10-14 from the user terminal device 6, it registers those results in the information management database 510. If the analysis device 2, molding device 3, and testing device 4 do not have communication capabilities, the user can use the user terminal device 6 to perform editing operations to add molecular chain information 11, crystallinity information 12, process information 13, and physical property information 14. The information management function 500 will then register the results of these editing operations in the information management database 510 when it receives them from the user terminal device 6.

[0035] As shown in Figures 3 and 4B, the raw material information 10 includes information on the type of raw material, the mixing ratio of the raw materials, and the degree of degradation of the raw materials. The information on the type of raw material specifies the name of the raw material and the grade name. For example, if registering a type of polyethylene as the raw material name, it can be selected from several options: HDPE, LDPE, and LLDPE. The information on the mixing ratio of the raw materials is the mixing ratio of each raw material included in a mixture of multiple types of raw materials. The information on the degree of degradation of the raw materials represents the degree of degradation of the recycled material and is specified, for example, by the chemiluminescence emission amount or the carbonyl group-derived peak measured by a Fourier transform infrared spectrophotometer (FT-IR). If the raw material is a mixture of multiple types of raw materials, the type of raw material and the degree of degradation may be specified for each type of raw material included in the mixture.

[0036] Molecular chain information 11 includes molten analysis molecular chain information 11A, which shows information obtained by performing a molten analysis on the molecular chains of the raw material using a molten analysis device 2A, as shown in Figures 3 and 4A, and solution analysis molecular chain information 11B, which shows information obtained by performing a solution analysis on the molecular chains of the raw material using a solution analysis device 2B.

[0037] The melt analysis molecular chain information 11A includes information on melt viscosity, etc. This information on melt viscosity includes, for example, analysis results from an MFR analyzer or viscometer, such as viscosity values ​​for various shear rates.

[0038] Solution analysis molecular chain information 11B includes information on molecular weight, molecular weight distribution, and branching structure. Information on molecular weight and molecular weight distribution is based on analysis using a GPC analyzer. Information on branching structure is based on analysis using an NMR analyzer.

[0039] Crystallinity information 12 includes melt analysis crystallinity information 12A, which is information obtained by performing a melt analysis on the crystalline portion of the raw material using a melt analysis device 2A, as shown in Figures 3 and 4A, and solid analysis crystallinity information 12B, which is information obtained by performing a solid analysis on the crystalline portion of the raw material using a solid analysis device 2C.

[0040] Melt analysis crystallinity information 12A includes information on crystallinity, crystallization rate, crystal thickness, or melting point. This information is based on analysis results from a DSC analyzer.

[0041] Solid state analysis crystallinity information 12B includes information on density, etc. The density information is the result of analysis using a dry densimeter or a density gradient tube analyzer.

[0042] Process information 13 includes, as shown in Figures 3 and 4B, molded piece process information 13A relating to the molding process when raw materials are melt-molded into molded pieces by the molded piece molding apparatus 3A, and film process information 13B relating to the molding process when raw materials are melt-molded into films by the film molding apparatus 3B.

[0043] The molded piece process information 13A consists of the setting conditions for each part when the molded piece molding apparatus 3A performs the molding process. The film process information 13B consists of the setting conditions for each part when the film molding apparatus 3B performs the molding process.

[0044] As shown in Figures 3 and 4B, the physical property information 14 includes molded piece physical property information 14A, which shows the physical properties of the molded piece when the raw material is melt-molded into a molded piece, and film physical property information 14B, which shows the physical properties of the film when the raw material is melt-molded into a film.

[0045] The molded piece physical property information 14A is obtained either as a test result from the molded piece testing device 4A or as a prediction result from the learning model 16. The film physical property information 14B is obtained either as a test result from the film testing device 4B or as a prediction result from the learning model 16.

[0046] The molded piece properties information 14A include environmental stress crack resistance, flexural modulus, tensile modulus, Charpy impact strength, tensile elongation at break, Vicat softening temperature, strength characteristics, contents characteristics, usability, and sealing properties. The film properties information 14B include impact burst strength, tensile impact strength, tensile modulus, thermal shrinkage coefficient, tear strength, loop stiffness, drop characteristics, ease of opening, usability, and line suitability.

[0047] Furthermore, each value included in the various information items 10-14 may be registered as a numerical value, as a step value selected from multiple step values, or as a choice selected from multiple options. In this case, each value included in the various information items 10-14 may be a specific value, or a value having a specific range such as an upper or lower limit. Also, if the raw material is a mixture of multiple types of raw materials, molecular chain information 11 and crystallinity information 12 for each type of raw material included in the mixture may be registered separately, or molecular chain information 11 and crystallinity information 12 for the mixture may be registered.

[0048] (Learning function 501) Figure 5 is a functional diagram showing an example of the learning function 501. Figure 6A shows the first training data 15-1 and the first learning model 16-1. Figure 6B shows the second training data 15-2 and the second learning model 16-2. Figure 6C shows the third training data 15-3 and the third learning model 16-3. Figure 6D shows the fourth training data 15-4 and the fourth learning model 16-4. The training data acquisition unit 5010 and the machine learning unit 5011 of the information processing device 5 realize the learning function 501 using the information management database 510 and the trained model management database 511 (trained model storage unit). By having the learning function 501, the information processing device 5 operates as a machine learning device.

[0049] The trained model 16 stored in the trained model management database 511 is a trained model that has learned the correlation between input data including physical property information 14 and output data including at least raw material information 10, using machine learning, as shown in Figure 5. In other words, the trained model 16 outputs output data in response to input data. The output data may include only raw material information 10, or, as shown in Figure 5, may include raw material information 10 along with molecular chain information 11, crystallinity information 12, and process information 13, or it may include at least one of molecular chain information 11, crystallinity information 12, and process information 13.

[0050] The physical property information 14 included in the input data consists of at least one of the following: molded piece physical property information 14A and film physical property information 14B. In this case, the molded piece physical property information 14A consists of at least one of the following: environmental stress crack resistance, flexural modulus, tensile modulus, Charpy impact strength, tensile elongation at break, Vicat softening temperature, strength characteristics, contents characteristics, usability, and sealing properties. The film physical property information 14B consists of at least one of the following: impact burst strength, tensile impact strength, tensile modulus, thermal shrinkage coefficient, tear strength, loop stiffness, drop characteristics, ease of opening, usability, and line suitability.

[0051] The raw material information 10 included in the output data consists of at least one of the following: information about the type of raw material, information about the mixing ratio of the raw materials, and information about the degree of deterioration of the raw materials. The details of the raw material information 10 included in the output data are the same as those shown in Figure 4B.

[0052] The molecular chain information 11 included in the output data consists of at least one of the following: melt analysis molecular chain information 11A and solution analysis molecular chain information 11B. In this case, melt analysis molecular chain information 11A is information regarding melt viscosity. Solution analysis molecular chain information 11B consists of at least one of the following: information regarding molecular weight, information regarding molecular weight distribution, and information regarding branched structure. The details of the melt analysis molecular chain information 11A and solution analysis molecular chain information 11B included in the output data are the same as those shown in Figure 4A.

[0053] The crystallinity information 12 included in the output data is at least one of the melt analysis crystallinity information 12A and the solid analysis crystallinity information 12B. In this case, the melt analysis crystallinity information 12A is information regarding the degree of crystallinity, crystallization rate, crystal thickness, or melting point. The solid analysis crystallinity information 12B is information regarding density. The details of the melt analysis crystallinity information 12A and the solid analysis crystallinity information 12B included in the output data are the same as those shown in Figure 4A.

[0054] Furthermore, if the raw material is a mixture of multiple types of raw materials, the output data may include molecular chain information 11 and crystallinity information 12 for each raw material contained in the mixture, or it may include molecular chain information 11 and crystallinity information 12 for the mixture as a whole.

[0055] The process information 13 included in the output data is at least one of molded piece process information 13A and film process information 13B. In this case, the molded piece process information 13A is the setting conditions for each part when the molded piece molding apparatus 3A performs the molding process. The film process information 13B is the setting conditions for each part when the film molding apparatus 3B performs the molding process. The details of the molded piece process information 13A and film process information 13B included in the input data are the same as those shown in Figure 4B.

[0056] In this embodiment, we will describe a case in which the trained model management database 511 stores, as a specific example of a trained model 16, the first to fourth trained models 16-1 to 16-4, which were machine-learned using the first to fourth training data 15-1 to 15-4 shown in Figures 6A to 6D.

[0057] The first learning model 16-1 takes molded piece physical property information 14A as input data and outputs at least raw material information 10 as output data. In Figure 6A, the output data is shown to include molded piece process information 13A in addition to raw material information 10, but molded piece process information 13A does not have to be included, and at least one of molecular chain information 11 and crystallinity information 12 may also be included.

[0058] The second learning model 16-2 takes film property information 14B as input data and outputs at least raw material information 10 as output data. In Figure 6A, the output data includes film process information 13B in addition to raw material information 10, but film process information 13B is not required to be included, and at least one of molecular chain information 11 and crystallinity information 12 may also be included.

[0059] The third learning model 16-3 takes molded piece physical property information 14A as input data and outputs at least raw material information 10 as output data. In Figure 6C, the output data is shown to include molded piece process information 13A in addition to raw material information 10, but molded piece process information 13A does not have to be included, and at least one of molecular chain information 11 and crystallinity information 12 may also be included.

[0060] The fourth learning model 16-4 takes film property information 14B as input data and outputs at least raw material information 10 as output data. In Figure 6D, the output data includes film process information 13B in addition to raw material information 10, but film process information 13B is not required to be included, and at least one of molecular chain information 11 and crystallinity information 12 may also be included.

[0061] The learning model 16 (the first to fourth learning models 16-1 to 16-4) employs, for example, a neural network structure, and as shown in Figure 5, comprises an input layer 160, a hidden layer 161, and an output layer 162. Synapses (not shown) connect each neuron between each layer, and each synapse is associated with a weight. The weight parameter set, consisting of the weights of each synapse, is adjusted by a machine learning algorithm such as backpropagation.

[0062] The input layer 160 has a number of neurons corresponding to the input data, with each value being input to each neuron. The output layer 162 has a number of neurons corresponding to the output data, and the output data is output as the inference result.

[0063] The training data acquisition unit 5010 acquires multiple sets of training data 15 (the first to fourth training data sets 15-1 to 15-4), each set consisting of input data including physical property information 14 and output data including at least raw material information 10. The training data 15 (the first to fourth training data sets 15-1 to 15-4) is used as training data, validation data, and test data in supervised learning. The output data included in the training data 15 (the first to fourth training data sets 15-1 to 15-4) is used as correct labels in supervised learning.

[0064] For example, the learning data acquisition unit 5010 acquires learning data 15 (first to fourth learning data 15-1 to 15-4) by referring to various pieces of information 10 to 14 registered in the information management database 510 or by receiving input operations from the user terminal device 6. If information corresponding to the learning data 15 (first to fourth learning data 15-1 to 15-4) is stored in an external system (such as a raw material management system or a molded product management system), the learning data acquisition unit 5010 may acquire the learning data 15 (first to fourth learning data 15-1 to 15-4) from the external system.

[0065] The machine learning unit 5011 uses multiple sets of training data 15 (the first to fourth training data 15-1 to 15-4) acquired by the training data acquisition unit 5010 to perform machine learning on each of the training models 16 (the first to fourth training models 16-1 to 16-4). The machine learning unit 5011 then trains each of the training models 16 (the first to fourth training models 16-1 to 16-4) to learn the correlation between the input data and the output data, thereby generating trained training models 16 (the first to fourth training models 16-1 to 16-4).

[0066] Furthermore, when performing machine learning, the machine learning unit 5011 can employ any method, such as online learning, batch learning, or mini-batch learning. In addition, the machine learning unit 5011 may perform predetermined preprocessing on the input data input to each learning model 16 (the first to fourth learning models 16-1 to 16-4), or may perform predetermined postprocessing on the output data output from each learning model 16 (the first to fourth learning models 16-1 to 16-4).

[0067] The trained model management database 511 stores the trained models 16 (first to fourth trained models 16-1 to 16-4) (specifically, the adjusted weight parameter sets) generated by the machine learning unit 5011. The trained models 16 (first to fourth trained models 16-1 to 16-4) stored in the trained model management database 511 may be provided to other systems via the network 7, recording media, etc.

[0068] In this embodiment, the data configuration of the training data 15 (first to fourth training data 15-1 to 15-4) and the training model 16 (first to fourth training models 16-1 to 16-4) was described as being as shown in Figures 5 and 6A to 6D. However, multiple data configurations with different conditions may be adopted, for example, such as differences in machine learning methods, input data, or output data. In such cases, the training data acquisition unit 5010 acquires multiple types of training data corresponding to the multiple data configurations with different conditions, and the machine learning unit 5011 performs machine learning using each of these training data and stores the trained training model 16 in the trained model management database 511.

[0069] (Raw material forecasting function 502) Figure 7 is a functional diagram illustrating an example of the raw material prediction function 502. The data acquisition unit 5020, data generation unit 5021, and output processing unit 5022 of the information processing device 5 implement the raw material prediction function 502 using the trained model management database 511. By equipping the information processing device 5 with the raw material prediction function 502, the information processing device 5 operates as a raw material prediction device.

[0070] The data acquisition unit 5020 acquires input data, including physical property information 14, for the new molded body to be predicted. For example, the data acquisition unit 5020 acquires input data by referring to the physical property information 14 registered in the information management database 510, receiving the physical property information 14 from the test apparatus 4, or accepting input operations for the physical property information 14 from the user terminal device 6. In doing so, the data acquisition unit 5020 acquires input data according to the data structure of the learning model 16.

[0071] The data generation unit 5021 inputs the input data acquired by the data acquisition unit 5020 into the learning model 16, thereby generating output data for the input data.

[0072] The learning models 16 (first to fourth learning models 16-1 to 16-4) used by the data generation unit 5021 are the learned learning models 16 (first to fourth learning models 16-1 to 16-4) that have been trained and stored in the trained model management database 511. If the first to fourth learning models 16-1 to 16-4 shown in Figures 6A to 6D are stored in the trained model management database 511 as learning models 16, the data generation unit 5021 can use the first to fourth learning models 16-1 to 16-4 selectively or in parallel. In this case, the data generation unit 5021 should input input data corresponding to the data structure of the input data in the first to fourth learning models 16-1 to 16-4 to each of the first to fourth learning models 16-1 to 16-4.

[0073] The data generation unit 5021 may perform predetermined preprocessing on the input data to be input to the learning model 16 (the first to fourth learning models 16-1 to 16-4). The data generation unit 5021 may also perform predetermined postprocessing on the output data output from the learning model 16 (the first to fourth learning models 16-1 to 16-4).

[0074] The output processing unit 5022 performs output processing to output the output data (raw material information 10, etc.) generated by the data generation unit 5021. For example, as part of the output processing, the output processing unit 5022 may send display information to the user terminal device 6 for displaying the raw material information 10, etc. on the user terminal device 6, or it may register the raw material information 10, etc. in the information management database 510.

[0075] Figure 8 is a hardware configuration diagram showing an example of the computer 900 that constitutes each device. Each device 2 to 6 in the raw material prediction system 1 is composed of a general-purpose or dedicated computer 900.

[0076] As shown in Figure 8, the computer 900 comprises, as its main components, a bus 910, a processor 912, memory 914, an input device 916, an output device 917, a display device 918, a storage device 920, a communication interface unit 922, an external device interface unit 924, an I / O device interface unit 926, and a media input / output unit 928. Note that the above components may be omitted as appropriate depending on the intended use of the computer 900.

[0077] The processor 912 consists of one or more arithmetic processing units (CPU (Central Processing Unit), MPU (Micro-Processing Unit), DSP (Digital Signal Processor), GPU (Graphics Processing Unit), NPU (Neural Processing Unit), etc.) and operates as a control unit that oversees the entire computer 900. The memory 914 stores various data and programs 930 and consists of volatile memory (DRAM, SRAM, etc.) that functions as main memory, and non-volatile memory (ROM), flash memory, etc.

[0078] The input device 916 consists of, for example, a keyboard, mouse, numeric keypad, or electronic pen, and functions as an input unit. The output device 917 consists of, for example, a sound (voice) output device or a vibration device, and functions as an output unit. The display device 918 consists of, for example, a liquid crystal display, an organic EL display, electronic paper, or a projector, and functions as an output unit. The input device 916 and the display device 918 may be configured as an integrated unit, such as a touch panel display. The storage device 920 consists of, for example, an HDD or SSD, and functions as a storage unit. The storage device 920 stores various data necessary for the execution of the operating system and program 930.

[0079] The communication I / F unit 922 is connected by wire or wireless to a network 940 such as the Internet or an intranet (which may be the same as network 7 in Figure 1) and functions as a communication unit that sends and receives data with other computers according to a predetermined communication standard. The external device I / F unit 924 is connected by wire or wireless to external devices 950 such as cameras, printers, scanners, and reader / writers and functions as a communication unit that sends and receives data with external devices 950 according to a predetermined communication standard. The I / O device I / F unit 926 is connected to I / O devices 960 such as various sensors and actuators and functions as a communication unit that sends and receives various signals and data with the I / O devices 960, for example, detection signals from sensors and control signals to actuators. The media input / output unit 928 consists of, for example, a drive device such as a DVD drive or CD drive, a memory card slot, and a USB connector, and reads and writes data to media (non-temporary storage media) 970 such as DVDs, CDs, memory cards, and USB memory.

[0080] In the computer 900 having the above configuration, the processor 912 calls and executes the program 930 stored in the storage device 920 in the memory 914, and controls various parts of the computer 900 via the bus 910. The program 930 may also be stored in the memory 914 instead of the storage device 920. The program 930 may be recorded on the media 970 in an installable or executable file format and provided to the computer 900 via the media input / output unit 928. The program 930 may also be provided to the computer 900 by downloading it via the network 940 through the communication interface unit 922. Furthermore, the computer 900 may implement various functions realized by the processor 912 executing the program 930 using hardware such as an FPGA (Field-Programmable Gate Array) or ASIC (Application Specific Integrated Circuit).

[0081] Computer 900 is an electronic device of any form, consisting of, for example, a stationary computer or a portable computer. Computer 900 may be a client computer, a server computer, a cloud computer, or an embedded computer such as a control panel or controller (including microcontrollers, programmable logic controllers, and sequencers).

[0082] (Operation of Raw Material Prediction System 1) The following describes the operation of the raw material prediction system 1, specifically the learning function 501 and the raw material prediction function 502, which are implemented by the information processing device 5.

[0083] (Machine learning methods) Figure 9 is a flowchart illustrating an example of a machine learning method using the learning function 501. Below, we will describe the case in which a learning model 16 (one of the first to fourth learning models 16-1 to 16-4) is generated using multiple sets of training data 15 (for example, any of the first to fourth training data 15-1 to 15-4 shown in Figures 6A to 6D).

[0084] First, in step S100, the training data acquisition unit 5010 acquires a desired number of training data 15 as preparation for starting machine learning, and temporarily stores the acquired training data 15 in the storage unit 51.

[0085] Next, in step S110, the machine learning unit 5011 prepares a pre-training model 16 in order to start machine learning. The pre-training model 16 prepared here is, for example, a neural network model, in which the weights of each synapse are set to initial values.

[0086] Next, in step S120, the machine learning unit 5011 randomly selects, for example, one set of training data 15 from the multiple sets of training data 15 stored in the memory unit 51.

[0087] Next, in step S130, the machine learning unit 5011 inputs the physical property information 14 (input data) contained in a set of training data 15 to the input layer 160 of the prepared pre-training (or training) learning model 16. As a result, raw material information 10, etc. (output data) is output from the output layer 162 of the learning model 16 as an inference result, but this output data is generated by the pre-training (or training) learning model 16. Therefore, in the pre-training (or training) state, the output data output as an inference result shows information different from the raw material information 10, etc. (ground truth labels) contained in the training data 15.

[0088] Next, in step S140, the machine learning unit 5011 compares the raw material information 10 etc. (ground truth labels) included in the set of training data 15 acquired in step S120 with the raw material information 10 etc. (output data) output as an inference result from the output layer 162 in step S130, and performs machine learning by adjusting the weight of each synapse (backpropagation). In this way, the machine learning unit 5011 trains the learning model 16 to recognize the correlation between the physical property information 14 and at least the raw material information 10.

[0089] Next, in step S150, the machine learning unit 5011 determines whether predetermined learning termination conditions have been met, for example, based on the evaluation value of the error function, which is based on the raw material information 10 etc. (correct labels) included in the training data 15 and the raw material information 10 etc. output as inference results (output data), or based on the remaining number of untrained training data 15 stored in the memory unit 51.

[0090] In step S150, if the machine learning unit 5011 determines that the learning termination condition has not been met and that machine learning should continue (No in step S150), it returns to step S120 and repeats steps S120 to S140 multiple times using the untrained training data 15 on the learning model 16 that is currently being trained. On the other hand, in step S150, if the machine learning unit 5011 determines that the learning termination condition has been met and that machine learning should be terminated (Yes in step S150), it proceeds to step S160.

[0091] Then, in step S160, the machine learning unit 5011 stores the trained model 16 (set weight parameters) generated by adjusting the weights associated with each synapse in the trained model management database 511, and the series of machine learning methods shown in Figure 9 is completed. In the machine learning method, step S100 corresponds to the training data acquisition process, steps S110 to S150 are the machine learning process, and step S160 is the trained model storage process.

[0092] As described above, the learning function 501 and machine learning method according to this embodiment can provide a learning model 16 that generates output data including at least raw material information 10 in response to input data including physical property information 14.

[0093] (Method for predicting raw materials) Figure 10 is a flowchart illustrating an example of a raw material prediction method using the raw material prediction function 502. In the following explanation, it is assumed that the trained model management database 511 stores the first to fourth trained models 16-1 to 16-4 shown in Figures 6A to 6D as trained models 16 trained by the learning function 501.

[0094] First, in step S200, the user terminal device 6 executes the information management program and, in cooperation with the raw material prediction function 502 of the information processing device 5, displays the raw material prediction input screen 17. On the raw material prediction input screen 17, it accepts input operations from the user regarding the new molded body to be predicted. The user terminal device 6 then transmits input operation information indicating the result of the user's input operations to the information processing device 5. The user terminal device 6 may also accept input operations to specify physical property information 14 or to specify information to be predicted other than raw material information 10 (for example, any of molecular chain information 11, crystallinity information 12, and process information 13) and transmit them to the information processing device 5.

[0095] Figure 11 is a screen configuration diagram showing an example of the raw material prediction input screen 17. The raw material prediction input screen 17 includes a physical property specification field 170 for specifying various information about the new molded product to be predicted, and a prediction target specification field 171 for specifying the information to be predicted.

[0096] In the physical property specification field 170, for example, by specifying a management ID registered in the information management database 510 or by entering physical property values, the physical property information 14 of the new molded product to be predicted is specified. In the example in Figure 11, the physical property specification field 170 is provided with a specification field for specifying a management ID registered in the information management database 510, an input field for entering molded piece physical property information 14A, and an input field for entering film physical property information 14B. In addition, the input field for molded piece physical property information 14A and the input field for film physical property information 14B are provided with input boxes for entering each physical property value.

[0097] In the prediction target specification field 171, the information to be predicted can be specified using checkboxes. In the example in Figure 11, it is shown that the "Molded Piece Physical Property Information 14A" entered in the physical property value specification field 170 has been specified to be predicted along with the "Raw Material Information 10" and "Process Information 13".

[0098] Next, in step S210, when the data acquisition unit 5020 receives the input operation information transmitted in step S200, it acquires input data to be input to the learning model 16 based on the input operation information. In the example of the raw material prediction input screen 17 shown in Figure 11, the data acquisition unit 5020 acquires molded piece physical property information 14A based on the value entered in the input box of the prediction target specification field 171.

[0099] Next, in step S220, the data generation unit 5021 inputs the input data (physical property information 14) acquired in step S210 into the learning model 16 to generate output data (raw material information 10, etc.) for the input data. In the example of the raw material prediction input screen 17 shown in Figure 11, "molded piece physical property information 14A" is specified as input data, and "raw material information 10" and "process information 13" are specified as information to be predicted, so the data generation unit 5021 selects the first learning model 16-1. Then, the data generation unit 5021 inputs the molded piece physical property information 14A of the molded body to be predicted, specified via the raw material prediction input screen 17, into the first learning model 16-1, and outputs raw material information 10 and molded piece process information 13A as output data, thereby predicting raw material information 10 and molded piece process information 13A.

[0100] Next, in step S230, the output processing unit 5022 performs output processing to output raw material information 10 and the like as output data generated in step S220. For example, if display information for displaying the output data is sent to the user terminal device 6, in step S240, the user terminal device 6 displays the raw material prediction output screen 18 based on that display information, thereby presenting the raw material information 10 and the like included in the output data to the user.

[0101] Figure 12 is a screen configuration diagram showing an example of the raw material prediction output screen 18. The raw material prediction output screen 18 includes a physical property value display field 180 that displays physical property information 14 as input data based on the input operations entered on the raw material prediction input screen 17, and a prediction result display field 181 that displays raw material information 10 etc. as output data generated by the data generation unit 5021.

[0102] In the raw material prediction output screen 18 shown in Figure 12, the prediction results for "raw material information 10" and "molded piece process information 13A" are displayed in the prediction result display field 181 as output data generated by the data generation unit 5021.

[0103] The series of raw material prediction methods shown in Figure 10 is completed as described above. In the above raw material prediction method, step S210 corresponds to the data acquisition step, step S220 to the data generation step, and step S230 to the output processing step. The series of raw material prediction methods can be executed at any time as long as the information processing device 5 is able to acquire the input data.

[0104] As described above, according to the raw material prediction function 502 and raw material prediction method of the information processing device 5 according to this embodiment, by inputting input data including physical property information 14 indicating the physical properties of a molded body obtained by melt-molding solid raw materials into the learning model 16, output data including raw material information 10 relating to the raw materials is generated from the input data. Therefore, it is possible to appropriately predict information about the raw materials from the physical properties of the molded body.

[0105] (Other embodiments) The present invention is not limited to the embodiments described above, and can be implemented with various modifications without departing from the spirit of the invention. All such modifications are included in the technical concept of the present invention.

[0106] In the above embodiment, the multiple functions of the information processing device 5 were described as being implemented by a single device, but each function may be distributed among multiple devices (computers) to be implemented by multiple devices. For example, the information management device that implements the information management function 500, the machine learning device that implements the learning function 501, and the raw material prediction device that implements the raw material prediction function 502 may be configured as different devices.

[0107] In the above embodiment, the information processing device 5 was described as having a communication unit 52 that transmits and receives various types of data to and from the user terminal device 6 via the communication unit 52. However, the information processing device 5 may also operate as a standalone device.

[0108] In the above embodiment, the case in which the information processing device 5 operates according to the flowcharts shown in Figures 9 and 10 has been described. However, some of the steps (parts) may be omitted, or other steps may be added. In that case, the omitted steps (parts) may be executed by an external system.

[0109] In the above embodiment, the learning function 501 of the information processing device 5 generates the first to fourth learning models 16-1 to 16-4, and the raw material prediction function 502 generates output data (raw material information 10, etc.) from input data (physical property information 14) using the first to fourth learning models 16-1 to 16-4. In contrast, the learning model 16 generated by the learning function 501 and used by the raw material prediction function 502 may be any one of the first to fourth learning models 16-1 to 16-4.

[0110] In the above embodiment, a case in which a neural network is used as the learning model 16 for realizing machine learning by the learning function 501 was described, but other machine learning models may also be used. Examples of other machine learning models include tree-type models such as decision trees and regression trees, ensemble learning such as bagging and boosting, neural network types such as recurrent neural networks, convolutional neural networks, and LSTM (including deep learning), hierarchical clustering, non-hierarchical clustering, clustering types such as k-nearest neighbors and k-means, multivariate analysis such as principal component analysis, factor analysis, and logistic regression, and support vector machines.

[0111] In the above embodiment, the case was described in which the raw material prediction function 502 of the information processing device 5 generates output data (raw material information 10, etc.) from input data (physical property information 14) using the first to fourth learning models 16-1 to 16-4. In contrast, the raw material prediction function 502 may generate output data (raw material information 10, etc.) from input data (physical property information 14) using known information 10 to 14 registered in the information management database 510 (storage unit) without using the first to fourth learning models 16-1 to 16-4.

[0112] Specifically, the data generation unit 5021 searches the information management database 510, which stores multiple known physical property information 14 and known raw material information 10 indicating raw materials from which molded products having the physical property values ​​indicated by the known physical property information 14 can be obtained, for known physical property information 14 that matches or is similar to the physical property information 14 included in the input data acquired by the data acquisition unit 5020. Based on the known raw material information 10 associated with the known physical property information 14 extracted as a search result, the data generation unit 5021 generates output data that includes the known raw material information 10.

[0113] As a search method for finding known physical property information 14 that matches or is similar to the physical property information 14 contained in the input data, for example, the physical property information 14 can be converted into features (e.g., feature vectors, etc.), and by comparing the features with each other, known physical property information 14 that satisfy a predetermined degree of match or similarity with respect to the input data can be extracted as search results. In this case, the output data may include not only known raw material information 10 associated with the known physical property information 14 extracted as a search result, but also molecular chain information 11, crystallinity information 12, or process information 13 associated with the known physical property information 14.

[0114] Furthermore, the process by which the data generation unit 5021 inputs input data into the first to fourth learning models 16-1 to 16-4 to generate output data (model input processing) and the process of searching for known information 10 to 14 registered in the information management database 510 based on the input data to generate output data (database search processing) may be performed selectively according to predetermined selection conditions or in parallel. For example, the data generation unit 5021 may generate output data by the database search processing if the number of data registered in the information management database 510 exceeds a predetermined threshold value, and generate output data by the model input processing otherwise. In addition, the data generation unit 5021 may search for known physical property information 14 registered in the information management database 510 based on the input data, and generate output data by the database search processing if the degree of match or similarity to the extracted known physical property information 14 exceeds a predetermined threshold value, and generate output data by the model input processing otherwise.

[0115] (Inference device, inference method, and inference program) The present invention can be provided not only in the form of an information processing device 5 (information processing method or information processing program) according to the above embodiment that functions as a raw material prediction device, but also in the form of an inference device (inference method or inference program) used to infer raw material information 10. In that case, the inference device (inference method or inference program) may include a memory and a processor, the processor of which may execute a series of processes. The series of processes includes a data acquisition process (data acquisition step) that acquires input data including physical property information 14, and an inference process (inference step) that, once the input data has been acquired by the data acquisition process, infers output data including at least raw material information 10 from the input data.

[0116] The various aspects of this disclosure are summarized below as an appendix.

[0117] (Note 1) A data acquisition unit that acquires input data, The system includes a data generation unit that inputs the input data acquired by the data acquisition unit into a learning model and generates output data for said input data, The aforementioned learning model, This is a trained model that has learned the correlation between the input data and the output data using machine learning. The aforementioned input data is Includes physical property information indicating the physical properties of a molded body obtained by melting and molding solid raw materials, The output data is, Including raw material information relating to the aforementioned raw materials, Raw material prediction device.

[0118] (Note 2) The physical property information included in the input data is, This is molded piece property information indicating the physical properties when the aforementioned raw materials are melt-molded into a molded piece having a predetermined shape as the molded body. The raw material prediction device described in Appendix 1.

[0119] (Note 3) The aforementioned molded piece physical properties information is, Environmental stress crack resistance, Flexural modulus, Tensile modulus, Charpy impact strength, Tensile elongation at fracture, and, At least one of the Vicat softening temperatures, The raw material prediction device described in Appendix 2.

[0120] (Note 4) The aforementioned molded piece physical properties information is, Strength characteristics, Contents characteristics, Usability, and At least one of the following is airtight: The raw material prediction device described in Appendix 2.

[0121] (Note 5) The physical property information included in the input data is, This is film property information indicating the physical properties when the aforementioned raw materials are melt-molded into a film having a predetermined thickness as the molded body. A raw material prediction device as described in any one of the appendices 1 to 4.

[0122] (Note 6) The aforementioned film physical property information is, Impact burst strength, Tensile impact strength, Tensile modulus, Thermal shrinkage rate, Tear strength, and, At least one of the loop stiffnesses, The raw material prediction device described in Appendix 5.

[0123] (Note 7) The aforementioned film physical property information is, Fall characteristics, unsealability, Preservability, Usability, and At least one of the line suitability criteria, The raw material prediction device described in Appendix 5.

[0124] (Note 8) The raw material information included in the output data is, Information regarding the types of raw materials mentioned above, Information regarding the blending ratio of the aforementioned raw materials, and At least one piece of information relating to the degree of deterioration of the aforementioned raw materials, A raw material prediction device as described in any one of the appendices 1 to 7.

[0125] (Note 9) The output data is, Molecular chain information relating to the molecular chains of the polymer compounds that serve as raw materials, and This further includes at least one piece of crystallinity information relating to the crystalline portion of the polymer compound, A raw material prediction device as described in any one of the appendices 1 to 8.

[0126] (Note 10) The molecular chain information included in the output data is, The information obtained by performing a melt analysis on the solid raw material, relating to the molecular chains, is shown as melt analysis molecular chain information, and At least one of the solution analysis molecular chain information, which represents information obtained by performing a solution analysis on the solid raw material relating to the molecular chain, The raw material prediction device described in Appendix 9.

[0127] (Note 11) The aforementioned melt analysis molecular chain information is, This is information regarding melt viscosity. The molecular chain information for the aforementioned solution analysis is, Information regarding molecular weight, Information regarding molecular weight distribution, and At least one piece of information regarding the branching structure, The raw material prediction device described in Appendix 10.

[0128] (Note 12) The crystallinity information included in the output data is, The melt analysis crystallinity information, which indicates information obtained by performing a melt analysis on the crystalline portion of the solid raw material, and At least one piece of solid analysis crystallinity information, which indicates information obtained by performing solid analysis on the crystalline portion of the solid raw material, A raw material prediction device as described in any one of the appendices 9 to 11.

[0129] (Note 13) The aforementioned melt analysis crystallinity information is, This information pertains to the degree of crystallinity, crystallization rate, crystal thickness, or melting point. The aforementioned solid analysis crystallinity information is, This is information about density. The raw material prediction device described in Appendix 12.

[0130] (Note 14) The output data is, The process information further includes process information relating to the molding process when melting and molding the solid raw material, A raw material prediction device as described in any one of the appendices 1 to 13.

[0131] (Note 15) The process information included in the output data is, Information relating to the molding process when the aforementioned raw materials are melt-molded into a molded piece having a predetermined shape as the molded body, and At least one piece of film process information relating to the molding process when the aforementioned raw materials are melt-molded into a film having a predetermined thickness as the molded body, That is, The raw material prediction device described in Appendix 14.

[0132] (Note 16) The data generation unit, From a storage unit that stores multiple known physical property information and known raw material information indicating the raw materials from which the molded body having the physical property values ​​indicated by the known physical property information can be obtained, the system searches for known physical property information that matches or is similar to the physical property information included in the input data acquired by the data acquisition unit, and generates the output data based on the known raw material information associated with the known physical property information extracted as a search result. A raw material prediction device as described in any one of the appendices 1 through 15. [Explanation of Symbols]

[0133] 1…Raw material prediction system, 2…Analytical device, 2A…Melting analyzer, 2B…Solution analyzer 2C...solid analysis device, 3...forming device, 3A...molded piece forming device, 3B...film forming device, 4...Testing equipment, 4A...Molded piece testing equipment, 4B...Film testing equipment, 5... Information processing device, 6... User terminal device, 10…Raw material information, 11…Molecular chain information, 11A…Melting analysis molecular chain information, 11B…Molecular chain information from solution analysis, 12…Crystalline information from fusion analysis, 12B...Solid analysis crystallinity information, 13...Process information, 13A...Molded piece process information, 13B...Film process information, 14...Physical property information, 14A...Molded piece physical property information, 14B...Film physical properties information, 15, 15-1~15-4...Training data, 16, 16-1~16-4... Learning models, 50...Control unit, 51...Storage unit, 52...Communication unit, 53...Input unit, 54...Display unit, 500... Information management function, 501... Learning function, 502... Raw material prediction function, 510... Information management database, 511... Model management database, 512... Information processing program, 5010...Training data acquisition unit, 5011...Machine learning unit, 5020...Data acquisition unit, 5021...Data generation unit, 5022...Output processing unit

Claims

1. A data acquisition unit that acquires input data, The system includes a data generation unit that inputs the input data acquired by the data acquisition unit into a learning model and generates output data for said input data, The aforementioned learning model, This is a trained model that has learned the correlation between the input data and the output data using machine learning. The aforementioned input data is Includes physical property information indicating the physical properties of a molded body obtained by melting and molding solid raw materials, The output data is, Including raw material information relating to the aforementioned raw materials, Raw material prediction device.

2. The physical property information included in the input data is, This is molded piece property information indicating the physical properties when the aforementioned raw materials are melt-molded into a molded piece having a predetermined shape as the molded body. The raw material prediction device according to claim 1.

3. The aforementioned molded piece physical properties information is, Environmental stress crack resistance, Flexural modulus, Tensile modulus, Charpy impact strength, Tensile elongation at fracture, and, At least one of the Vicat softening temperatures, The raw material prediction device according to claim 2.

4. The aforementioned molded piece physical properties information is, Strength characteristics, Contents characteristics, Usability, and At least one of the following is airtight: The raw material prediction device according to claim 2.

5. The physical property information included in the input data is, This is film property information indicating the physical properties when the aforementioned raw materials are melt-molded into a film having a predetermined thickness as the molded body. The raw material prediction device according to claim 1.

6. The aforementioned film physical property information is, Impact burst strength, Tensile impact strength, Tensile modulus, Thermal shrinkage rate, Tear strength, and, At least one of the loop stiffnesses, The raw material prediction device according to claim 5.

7. The aforementioned film physical property information is, Fall characteristics, unsealability, Preservability, Usability, and At least one of the line suitability criteria, The raw material prediction device according to claim 5.

8. The raw material information included in the output data is, Information regarding the types of raw materials mentioned above, Information regarding the blending ratio of the aforementioned raw materials, and At least one piece of information relating to the degree of deterioration of the aforementioned raw materials, The raw material prediction device according to claim 1.

9. The output data is, Molecular chain information relating to the molecular chains of the polymer compounds that serve as raw materials, and This further includes at least one piece of crystallinity information relating to the crystalline portion of the polymer compound, The raw material prediction device according to claim 1.

10. The molecular chain information included in the output data is, The information obtained by performing a melt analysis on the solid raw material, relating to the molecular chains, is shown as melt analysis molecular chain information, and At least one of the solution analysis molecular chain information, which represents information obtained by performing a solution analysis on the solid raw material relating to the molecular chain, The raw material prediction device according to claim 9.

11. The aforementioned melt analysis molecular chain information is, This is information regarding melt viscosity. The molecular chain information for the aforementioned solution analysis is, Information regarding molecular weight, Information regarding molecular weight distribution, and At least one piece of information regarding the branching structure, The raw material prediction device according to claim 10.

12. The crystallinity information included in the output data is, The melt analysis crystallinity information, which indicates information obtained by performing a melt analysis on the crystalline portion of the solid raw material, and At least one piece of solid analysis crystallinity information, which indicates information obtained by performing solid analysis on the crystalline portion of the solid raw material, The raw material prediction device according to claim 9.

13. The aforementioned melt analysis crystallinity information is, This information pertains to the degree of crystallinity, crystallization rate, crystal thickness, or melting point. The aforementioned solid analysis crystallinity information is, This is information about density. The raw material prediction device according to claim 12.

14. The output data is, The process information further includes process information relating to the molding process when melting and molding the solid raw material, The raw material prediction device according to claim 1.

15. The process information included in the output data is, Information relating to the molding process when the aforementioned raw materials are melt-molded into a molded piece having a predetermined shape as the molded body, and At least one piece of film process information relating to the molding process when the aforementioned raw materials are melt-molded into a film having a predetermined thickness as the molded body, That is, The raw material prediction device according to claim 14.

16. The data generation unit, From a storage unit that stores multiple known physical property information and known raw material information indicating the raw materials from which the molded body having the physical property values ​​indicated by the known physical property information can be obtained, the system searches for known physical property information that matches or is similar to the physical property information included in the input data acquired by the data acquisition unit, and generates the output data based on the known raw material information associated with the known physical property information extracted as a search result. The raw material prediction device according to claim 1.

17. An inference device comprising memory and a processor, The aforementioned processor, A data acquisition process that acquires input data containing physical property information indicating the physical properties of a molded body obtained by melting and molding solid raw materials, When the input data is acquired through the data acquisition process, an inference process is performed on the input data to infer output data including raw material information relating to the raw materials. Reasoning device.

18. A training data acquisition unit that acquires multiple sets of training data consisting of input data and output data, A machine learning unit uses multiple sets of the training data acquired by the training data acquisition unit to train a learning model on the correlation between the input data and the output data using machine learning. The machine learning unit then stores the learned model in which the correlation has been learned, and the machine learning unit stores the learned model in a learned model. The aforementioned input data is Includes physical property information indicating the physical properties of a molded body obtained by melting and molding solid raw materials, The output data is, Including raw material information relating to the aforementioned raw materials, Machine learning device.

19. A method for predicting raw materials performed by a computer, The data acquisition process involves obtaining input data, The system includes a data generation step which generates output data for the input data obtained in the data acquisition step by inputting the input data into a learning model, The aforementioned learning model, This is a trained model that has learned the correlation between the input data and the output data using machine learning. The aforementioned input data is Includes physical property information indicating the physical properties of a molded body obtained by melting and molding solid raw materials, The output data is, Including raw material information relating to the aforementioned raw materials, Methods for predicting raw material costs.

20. An inference method performed by an inference device comprising memory and a processor, The aforementioned processor, A data acquisition process that acquires input data containing physical property information indicating the physical properties of a molded body obtained by melting and molding solid raw materials, When the input data is acquired through the data acquisition process, an inference process is performed on the input data to infer output data including raw material information relating to the raw materials. Reasoning method.

21. A machine learning method performed by a computer, The training data acquisition process involves acquiring multiple sets of training data consisting of input data and output data, A machine learning step in which a learning model learns the correlation between the input data and the output data using machine learning, using multiple sets of the training data acquired in the training data acquisition step, The system includes a trained model storage step, which stores the trained model, which has learned the correlation relationship through the machine learning step, in a trained model storage unit. The aforementioned input data is Includes physical property information indicating the physical properties of a molded body obtained by melting and molding solid raw materials, The output data is, Including raw material information relating to the aforementioned raw materials, Machine learning methods.

Images