Material design device, material design method and material design program

Abstract

A material design device is provided that can quickly derive an optimal solution for design conditions of a polymer that satisfies desired required physical properties. The material design device 1 comprises: a mix ratio range input unit 11 for setting a specified range of polymer design conditions; a comprehensive analysis point generation unit 12 for generating a plurality of comprehensive analysis points within the specified range inputted by the mix ratio range input unit 11; a comprehensive analysis point-polymer property storage unit 14 for storing a data set in which polymer property values ​​calculated by inputting the comprehensive analysis points generated by the comprehensive analysis point generation unit 12 into a learning model 13 and associating them with each of the comprehensive analysis points; a required physical property input unit 21 for setting a specified range of required physical properties of a design target polymer material; and a filtering unit 22 for selecting, from the data set stored in the comprehensive analysis point-polymer property storage unit 14, comprehensive analysis points that satisfy the required physical properties inputted by the required physical property input unit 21.

Description

Technical Field

[0001] The present disclosure relates to a material design device, a material design method, and a material design program. Background Art

[0002] Conventionally, polymer materials have been designed by repeatedly trial-producing while adjusting the composition based on the experience of material developers.

[0003] However, in most cases, trial production based on the material developer's experience requires repeated trial production until the optimal design is achieved, which takes a lot of time and effort. In addition, the material developer often conducts local condition exploration near the design conditions he or she has previously used, rather than focusing on the search for optimal design conditions over a wide range.

[0004] For example, Patent Document 1 proposes a material design support system that, when a material designer specifies a desired material function, presents material candidates that satisfy that function. The material design support system disclosed in Patent Document 1 utilizes a machine learning system, such as a neural network, as a tool for performing inverse problem analysis to derive design conditions for materials that satisfy the desired material function (physical properties), thereby making it possible to more efficiently predict material candidates.

[0005] Prior art literature

[0006] Patent Literature

[0007] Patent Document 1: Japanese Patent Application Laid-Open No. 2010-198561 Summary of the Invention

[0008] Problems to be solved by the invention

[0009] However, the material design support system disclosed in Patent Document 1 does not consider the number of substances to be compounded or the order of compounding. In addition, the polymer material itself is treated as the compounded material, making it unsuitable for material design of the polymer itself.

[0010] An object of the present disclosure is to provide a material design apparatus, a material design method, and a material design program for a polymer material that simultaneously satisfies a plurality of desired material properties.

[0011] Methods for solving problems

[0012] The present disclosure has the following configuration.

[0013] [1] A material design device for designing a polymer produced by polymerizing a plurality of monomers, comprising:

[0014] The learning model has learned the correspondence between the input information of the monomer mixing ratio and the output information of the polymer physical property values.

[0015] A mixing ratio range input unit for inputting a mixing ratio range of at least one monomer;

[0016] a required physical property input unit for inputting a required range of at least one physical property value of the polymer;

[0017] a comprehensive analysis point generating unit for generating a comprehensive analysis point of a polymer polymerized using a plurality of monomers, the plurality of monomers including at least one monomer having an input mixing ratio range within a mixing ratio range;

[0018] a comprehensive analysis point-polymer property value storage unit that inputs the generated comprehensive analysis point into the learned model to calculate the polymer property value, creates a data set that associates the comprehensive analysis point with the calculated polymer property value, and stores the created data set; and

[0019] The filtering unit selects polymers whose physical property values ​​are within a required range inputted through the required physical property input unit from the data set.

[0020] [2] The material design device according to [1],

[0021] In the above-mentioned mixing ratio range input part, the number of monomers used in polymerization is input to limit the number of monomers used in polymerization.

[0022] Generates comprehensive analysis points for polymers produced using a limited number of monomers.

[0023] [3] The material design device according to [1] or [2],

[0024] At least one monomer required for polymerization of the monomers having the input mixing ratio range is input into the mixing ratio range input section.

[0025] Generates comprehensive analytical points for polymers obtained by polymerizing multiple monomers including essential monomers within a certain blending ratio range.

[0026] [4] A material design device for designing a graft polymer obtained by two-stage polymerization of a plurality of monomers, comprising:

[0027] The learning model has learned the correspondence between the input information of the monomer mixing ratio and the output information of the polymer physical property values;

[0028] A mixing ratio range input unit for inputting a mixing ratio range of at least one monomer;

[0029] a required physical property input unit for inputting a required range of at least one physical property value of the polymer;

[0030] a first-stage comprehensive analysis point generating unit for selecting at least one first-stage monomer to be used in the first-stage polymerization from the monomers for which a mixing ratio range is input, and generating a comprehensive analysis point of a main-chain polymer polymerized using a plurality of monomers, the plurality of monomers including the at least one first-stage monomer within the mixing ratio range;

[0031] The second-stage monomer proposal section proposes at least one monomer to be used for the second-stage polymerization of the main chain polymer based on the comprehensive analysis points of the first stage;

[0032] a comprehensive analytical point generating unit for polymerizing a graft polymer from the second-stage monomer and the main chain polymer to generate a comprehensive analytical point for the graft polymer;

[0033] a comprehensive analysis point-polymer property value storage unit that inputs the comprehensive analysis point into the learned model to calculate the physical property values ​​of the graft polymer, creates a data set that associates the comprehensive analysis point with the calculated physical property values ​​of the graft polymer, and stores the created data set; and

[0034] The filtering unit selects, from the data set, graft polymers within the required range of the physical property values ​​inputted by the required physical property input unit.

[0035] [5] The material design device according to [4], wherein the number of monomers used for polymerization is input in the mixing ratio range input unit to limit the number of monomers used for polymerization.

[0036] A comprehensive analysis of the generation of graft polymers using a limited number of monomers.

[0037] [6] The material design device according to [4] or [5], wherein at least one monomer required for polymerization of the monomers for which the mixing ratio range is input is input into the mixing ratio range input unit,

[0038] In the above-mentioned integrated analysis point generating section, an integrated analysis point of a graft polymer obtained by polymerizing a plurality of monomers including an essential monomer within a blending ratio range is generated.

[0039] [7] The material design device according to [4] or [5], wherein at least one monomer required for the first stage polymerization of the monomers for which the mixing ratio range is input is input into the mixing ratio range input unit,

[0040] In the first-stage comprehensive analysis point generating unit, a comprehensive analysis point of a main chain polymer obtained by polymerizing a plurality of monomers including an essential monomer within a blending ratio range is generated.

[0041] [8] A material design method for designing a polymer formed by polymerizing a plurality of monomers, comprising:

[0042] a model creation step of creating a learned model that has learned the correspondence between input information on monomer mixing ratios and output information on polymer physical property values;

[0043] a design condition setting step of inputting a mixing ratio range of at least one monomer and a required range of at least one physical property value of the polymer;

[0044] a comprehensive analysis point generating step of generating a comprehensive analysis point of a polymer polymerized within a mixing ratio range using at least one monomer having a mixing ratio range input in the design condition setting step;

[0045] a data set creation step of inputting the comprehensive analysis points generated in the comprehensive analysis point generation step into the learned model to calculate polymer property values, creating a data set in which the comprehensive analysis points are associated with the calculated polymer property values, and storing the created data set in a comprehensive analysis point-polymer property value storage unit; and

[0046] In the filtering step, polymers whose physical property values ​​are within the required ranges input in the design condition setting step are selected from the data set.

[0047] [9] A material design program for designing a polymer composed of a plurality of monomers, which is used to enable a computer to implement the following functions:

[0048] A model creation function that creates a learned model that has learned the correspondence between input information on monomer mix ratios and output information on polymer physical property values;

[0049] A design condition setting function is provided, which inputs the mixing ratio range of at least one monomer and the required range of at least one physical property value of the above-mentioned polymer;

[0050] A comprehensive analysis point generation function for generating a comprehensive analysis point of a polymer polymerized within a range of a mixing ratio using at least one monomer having a mixing ratio range input by the above-mentioned design condition setting function;

[0051] a data set creation function for inputting the comprehensive analysis points generated by the comprehensive analysis point generation function into the learned model to calculate polymer property values, creating a data set that associates the comprehensive analysis points with the calculated polymer property values, and storing the created data set in a comprehensive analysis point-polymer property value storage unit; and

[0052] The filtering function selects polymers within the required range of the physical property values ​​input by the above-mentioned design condition setting function from the data set.

[0053]

[10] The material design device according to [4],

[0054] When a specific monomer is selected as the first-stage monomer, the learned model learns the monomer and the monomer constituting the main-chain polymer as different substances.

[0055]

[11] A material design method for designing a graft polymer obtained by two-stage polymerization of a plurality of monomers, comprising:

[0056] Creating a learning model that learns the correspondence between input information of monomer mix ratios and output information of polymer physical property values;

[0057] Enter the mix ratio range of at least one monomer;

[0058] Enter the required range of at least one physical property value of the polymer;

[0059] Selecting at least one first-stage monomer to be used in the first-stage polymerization from the monomers having a mixing ratio range input thereto, and generating a comprehensive analysis point of a main chain polymer polymerized using a plurality of monomers, the plurality of monomers including the at least one first-stage monomer within the mixing ratio range;

[0060] Based on the comprehensive analysis points of the first stage, propose at least one monomer used for the second stage polymerization with the main chain polymer;

[0061] A graft polymer is polymerized from the second-stage monomer and the main chain polymer to generate a comprehensive analytical point of the graft polymer;

[0062] inputting the comprehensive analysis points into the learned model to calculate the physical property values ​​of the graft polymer, creating a data set in which the comprehensive analysis points are associated with the calculated physical property values ​​of the graft polymer, and storing the created data set; and

[0063] A graft polymer within a required range of the physical property value inputted by the required physical property input unit is selected from the data set.

[0064]

[12] A non-transitory computer-readable storage medium storing a material design program for designing a graft polymer obtained by two-stage polymerization of a plurality of monomers, which is used to cause a computer to implement the following functions:

[0065] Creating a learning model that learns the correspondence between input information of monomer mix ratios and output information of polymer physical property values;

[0066] Enter the mix ratio range of at least one monomer;

[0067] Enter the required range of at least one physical property value of the polymer;

[0068] Selecting at least one first-stage monomer to be used in the first-stage polymerization from the monomers having a mixing ratio range input thereto, and generating a comprehensive analysis point of a main-chain polymer polymerized using a plurality of monomers, wherein the plurality of monomers include at least one first-stage monomer within the mixing ratio range;

[0069] Based on the comprehensive analysis points of the first stage, propose at least one monomer used for the second stage polymerization with the main chain polymer;

[0070] A graft polymer is polymerized from the second-stage monomer and the main chain polymer to generate a comprehensive analytical point of the graft polymer;

[0071] inputting the comprehensive analysis points into the learned model to calculate the physical property values ​​of the graft polymer, creating a data set in which the comprehensive analysis points are associated with the calculated physical property values ​​of the graft polymer, and storing the created data set; and

[0072] A graft polymer within a required range of the physical property value inputted by the required physical property input unit is selected from the data set.

[0073] Effects of the Invention

[0074] According to the present disclosure, it is possible to provide a material design apparatus, a material design method, and a material design program capable of designing a polymer satisfying desired physical properties in a short time. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 This is a block diagram showing an example of a schematic configuration of a material design device according to an embodiment.

[0076] Figure 2 This is a block diagram showing an example of a schematic configuration of a material design device according to an embodiment.

[0077] Figure 3 This is a diagram showing an example of an input screen of the mixing ratio range input section.

[0078] Figure 4 This is a diagram showing an example of an input screen of the mixing ratio range input section.

[0079] Figure 5 This is a diagram showing an example of an input screen of the mixing ratio range input section.

[0080] Figure 6 This is a diagram showing an example of an input screen of the mixing ratio range input section.

[0081] Figure 7 This is a diagram showing an example of an input screen of the mixing ratio range input section.

[0082] Figure 8It is a diagram showing an example of an input screen of the required physical property input section.

[0083] Figure 9 This is a diagram showing an example of a data set stored in the comprehensive analysis point-polymer property value storage unit.

[0084] Figure 10 A diagram showing an example of an output screen of the information display unit.

[0085] Figure 11 This is a block diagram showing an example of the hardware configuration of a material design device.

[0086] Figure 12 This is an example of a flowchart of the processing performed by the forward problem analysis unit and the inverse problem analysis unit. DETAILED DESCRIPTION

[0087] Below, with reference to the attached Figure 1 To facilitate understanding of the present invention, identical components are denoted by identical reference numerals in the accompanying drawings as much as possible, and redundant descriptions are omitted.

[0088] Reference Figures 1 to 11 The structure of the material design device 1 according to the embodiment will be described. Figure 1 1 is a block diagram showing an example of a schematic configuration of a material design apparatus 1 according to an embodiment. The material design apparatus 1 is an apparatus for designing a polymer obtained by polymerizing a plurality of monomers.

[0089] like Figure 1 As shown, the material design device 1 includes a forward problem analysis unit 10, an inverse problem analysis unit 20, and a GUI (Graphical User Interface) 30. The forward problem analysis unit 10 uses the learning completion model 13 to generate a comprehensive analysis point of a polymer polymerized using multiple monomers whose mix ratio range includes at least one monomer whose mix ratio range is input by the material designer. The inverse problem analysis unit 20 selects a data set that meets the required range of physical property values ​​input by the material designer from a data set created based on the comprehensive analysis point generated by the forward problem analysis unit 10. The GUI 30 is an interface that displays the output results of the forward problem analysis unit 10 and the inverse problem analysis unit 20 and has functions such as providing prompts to the material designer.

[0090] The forward problem analysis unit 10 includes a mix ratio range input unit 11 , a comprehensive analysis point generation unit 12 , a learned model 13 , and a comprehensive analysis point-polymer property value storage unit 14 .

[0091] The mixing ratio range input unit 11 inputs the mixing ratio range of at least one monomer constituting the polymer to be designed. The mixing ratio range input unit 11 can display a list of monomers to be analyzed on the GUI 30, for example, to facilitate input by the material designer.

[0092] The items of the mixing ratio range input unit 11 include the mixing ratio of the monomers. Figure 3 1 is a diagram showing an example of an input screen 11A of the mixing ratio range input unit 11. The input screen 11A displays a pre-prepared monomer list, and allows the user to input the minimum and maximum values ​​of the mixing ratio range for each monomer.

[0093] The items of the mixing ratio range input unit 11 may further include evaluation conditions such as the amount of the polymerization initiator and the polymerization inhibitor, the solvent used for polymerization, and the post-baking temperature and time.

[0094] In the comprehensive analysis point generation unit 12, the mixing ratio range input unit 11 generates a comprehensive analysis point for a polymer polymerized using multiple monomers whose mixing ratio ranges include at least one monomer for which the mixing ratio range was input. For example, when a mixing ratio range for three monomers (Monomer 1, Monomer 2, and Monomer 3) is input and these three monomers are selected for polymerization, the factors that ultimately determine the mixing ratio are first randomly determined. For example, when the mixing ratio of Monomer 3 is finally determined, multiple mixing ratio combinations of Monomer 1 and Monomer 2 are first created within the input mixing ratio ranges, either randomly or with a predetermined step. Next, the mixing ratio of Monomer 3 is determined so that the sum of the mixing ratios is 1 (= 100% for the total number of monomers constituting the polymer). Among the mixing ratio combinations that include Monomer 3, a mixing ratio combination that falls within the mixing ratio range for Monomer 3 input in the mixing ratio input unit 11 is selected, and a comprehensive analysis point for the polymer polymerized using the selected mixing ratio combination is generated.

[0095] The number of monomers used in polymerization can be limited by inputting the number of monomers used in polymerization in the mixing ratio range input unit 11. If the number of monomers used in polymerization is limited, the comprehensive analysis point generation unit 12 generates a comprehensive analysis point for the polymer polymerized using the limited number of monomers.

[0096] Figure 4 This figure shows an example of an input screen 11B of the mixing ratio range input unit 11 when the number of monomers used for polymerization is input and the number of monomers used for polymerization is limited. In the input screen 11B, "3" is input as the number of monomers used for polymerization, and the number of monomers used for polymerization is limited to "3".

[0097] When the number of monomers used in polymerization is limited, the comprehensive analysis point generation unit 12 generates a comprehensive analysis point for a polymer polymerized using the limited number of monomers within the mix ratio range entered in the mix ratio range input unit 11. For example, if the number of monomers is "3," three monomers are selected whose maximum value in the mix ratio range entered in the mix ratio range input unit 11 is not 0. An example of comprehensive analysis point generation is shown for the case where monomers 1, 2, and 3 are selected from monomers for which a mix ratio range has been entered. First, the monomer for which the mix ratio is finally determined is randomly selected. For example, when the mix ratio of monomer 3 is finally determined, multiple mix ratios are first calculated within the mix ratio range entered for monomers 1 and 2, using random or predetermined steps, and all mix ratio combinations of the calculated mix ratios are created. Next, the mix ratio of monomer 3 is determined so that the sum of the mix ratios is 1 (= 100% for the total number of monomers constituting the polymer). Among the prepared mixing ratio combinations, a comprehensive analysis point of a polymer polymerized with a mixing ratio combination within the mixing ratio range of the monomer 3 inputted into the mixing ratio range input unit 11 is generated.

[0098] At least one monomer required for polymerization of monomers having an input mixing ratio range can be input into the mixing ratio range input unit 11. Once the required monomer is input, the comprehensive analysis point generation unit 12 generates a comprehensive analysis point for a polymer obtained by polymerization of a plurality of monomers including the required monomer within the mixing ratio range.

[0099] Figure 5 The figure shows an example of an input screen 11C for inputting at least one monomer required for polymerization from monomers for which a mixing ratio range has been input, in the mixing ratio range input unit 11. In the example of the input screen 11C, monomer 1 and monomer 3 are input as required monomers.

[0100] When the essential monomer is input, the comprehensive analysis point generating unit 12 generates a comprehensive analysis point of a polymer polymerized from a plurality of monomers included in the blending ratio range in which the monomer 1 and the monomer 3 are input as essential monomers.

[0101] The learned model 13 is a model that has been formulated by learning the correspondence between input information including the polymer mixing ratio range and output information including the polymer physical property values ​​through machine learning. For example, a supervised learning model such as a neural network or a genetic algorithm can be used as the learned model 13.

[0102] The comprehensive analysis point-polymer property value storage unit 14 inputs the comprehensive analysis points generated by the comprehensive analysis point generation unit 12 into the learned model 13 to calculate polymer property values, creates a data set in which the comprehensive analysis points and the calculated polymer property values ​​are associated, and stores the created data set. Figure 9 1 is a diagram showing an example of a data set stored in the comprehensive analysis point-polymer property value storage unit 14. Figure 9 As shown, the monomer mix ratio, which serves as input to the learned model 13, and the polymer physical property values, which serve as output, are recorded as a single data set on the same row. Each row of the data set stored in the comprehensive analysis point-polymer physical property value storage unit 14 corresponds to the physical property values ​​of each polymer at the generated comprehensive analysis point. Furthermore, this data set may include information on the composition, such as the amount of polymerization initiator and inhibitor, the solvent used for polymerization, and evaluation conditions, such as post-baking temperature and time.

[0103] In this way, the forward problem analysis unit 10 is constructed so that the material designer only needs to input the mixing ratio range of at least one monomer, and can automatically generate a data set of comprehensive analysis points of polymers within the mixing ratio range of all monomers and polymer property values ​​of the polymer generated by the mixing ratio.

[0104] The inverse problem analyzing unit 20 includes a required physical property input unit 21 and a filtering unit 22. The inverse problem analyzing unit 20 also includes the aforementioned integrated analysis point-polymer physical property value storage unit 14.

[0105] The required range of at least one physical property value of the design target polymer is input in the required physical property input unit 21. The required physical property input unit 21 can prompt input by, for example, displaying an input screen for the required range of physical property values ​​on the GUI 30.

[0106] Figure 8 This figure shows an example of an input screen 21A for the required physical property input unit 21. Items of physical property values ​​entered as required physical properties include, for example, viscosity, glass transition point, color difference, molecular weight, acid value, developability (speed, development portability, development residue), adhesion, transparency, sensitivity, solvent resistance (e.g., NMP resistance), dispersibility, heat resistance (thermal component phenomenon), and heat resistance (yellowing color difference). On input screen 21A, the minimum and maximum values ​​within the required range can be entered for each physical property value.

[0107] The filtering unit 22 selects comprehensive analysis points that satisfy the required range of the polymer property values ​​input in the required property input unit 21 from the data set stored in the comprehensive analysis point-polymer property storage unit 14 .

[0108] In this manner, the inverse problem analyzing unit 20 is configured such that the material designer only needs to input a required range of at least one physical property value to obtain the mixing ratios of multiple monomers required for polymerizing polymers that all meet the input required range of physical property values.

[0109] The GUI 30 includes an information display unit 31. The information display unit 31 displays the outputs of the forward problem analysis unit 10 and the inverse problem analysis unit 20. For example, the information display unit 31 displays the monomer mix ratio and physical property values ​​of the data set selected in the inverse problem analysis unit 20.

[0110] Reference Figure 2 , the structure of the material design device 2 involved in the embodiment is explained.

[0111] Figure 2 A schematic configuration of a material design device 2 is shown as another example of the configuration of the material design device according to the embodiment of the present invention. Figure 2 The material design device 2 shown is a material design device for designing a graft polymer formed by two-stage polymerization of multiple monomers. It includes a forward problem analysis unit 40, an inverse problem analysis unit 20, and a GUI 30, which are similar to the components included in the material design device 1 described above. The following description focuses on the forward problem analysis unit 40, which has a different configuration from the forward problem analysis unit 10 included in the material design device 1.

[0112] The forward problem analysis unit 40 of the material design device 2 includes: a mixing ratio range input unit 41, a first-stage comprehensive analysis point generation unit 42, a second-stage monomer proposal unit 43, a comprehensive comprehensive analysis point generation unit 44, a learning completion model 45, and a comprehensive analysis point-polymer property value storage unit 46.

[0113] The mixing ratio range input unit 41 inputs the mixing ratio range of at least one monomer constituting the polymer of the design object. The mixing ratio range input unit 41 can be input by displaying a list of monomers to be analyzed in the GUI 30, for example. As the input screen of the mixing ratio range input unit 41, the above-mentioned Figure 3 11A.

[0114] In the first-stage comprehensive analysis point generation unit 42, at least one first-stage monomer used for the first-stage polymerization is selected from the monomers with a mixing ratio range input, and a comprehensive analysis point of a main-chain polymer polymerized using multiple monomers with a mixing ratio range including at least one first-stage monomer is generated. For example, when monomer 1 and monomer 2 are selected as first-stage monomers from the monomers with a mixing ratio range input, first, multiple mixing ratios are calculated with random or specified steps within the mixing ratio ranges of monomer 1 and monomer 2, and a first-stage comprehensive analysis point is generated with information that monomer 1 and monomer 2 are monomers constituting the main-chain polymer for the calculated mixing ratio combination. The information of the monomers constituting the main-chain polymer is assigned, for example, by changing the name of monomer 1 to monomer 1 (main chain). In this case, the learning completion model 45 uses a learning completion model that has learned monomer 1 and monomer 1 (main chain) as different substances.

[0115] The second-stage monomer proposal unit 43 proposes, based on the first-stage comprehensive analysis point generated by the first-stage comprehensive analysis point generation unit 42, at least one second-stage monomer that reacts with the main-chain polymer to form a graft polymer, and a second-stage monomer blending ratio such that the sum of the blending ratios is 1 (= the total number of monomers constituting the graft polymer is 100%). The second-stage monomers and their blending ratios can be proposed using a learned model that has learned the relationship between predetermined rules, the first-stage comprehensive analysis point, and the composition of producible polymers.

[0116] The comprehensive analysis point generation unit 44 uses the second-stage monomer proposal unit 43 to generate a comprehensive analysis point for the graft polymer obtained by polymerization with the main chain polymer using the proposed second-stage monomers and the proposed mixing ratios. For example, if monomers 3 and 4 are proposed as second-stage monomers, the factors that will ultimately determine the mixing ratio are first randomly determined. For example, when ultimately determining the mixing ratio of monomer 4, the mixing ratio of monomer 3 is first determined using a random or predetermined step. At this time, if a mixing ratio range for monomer 3 is input into the mixing ratio range input unit 41, the mixing ratio range for monomer 3 is calculated to be within the input mixing ratio range. Next, the mixing ratio of monomer 4 is determined so that the sum of the mixing ratios of monomers 1 to 4 is 1 (= 100% for the total number of monomers constituting the graft polymer). For the generated mixing ratio combinations, a comprehensive analysis point is generated for the graft polymer polymerized using the mixing ratio combination of monomers 3 and 4 within the mixing ratio range input into the mixing ratio range input unit 41.

[0117] The learned model 45 may be the same as the learned model 13. When both the first-stage monomer and the second-stage monomer are available for a specific monomer, learned models learned as different substances are used.

[0118] The total number of monomers used in the polymerization can be limited by inputting the number of monomers constituting the graft polymer into the blending ratio range input unit 41. If the total number of monomers used in the polymerization is limited, the overall comprehensive analysis point generation unit 44 generates an overall comprehensive analysis point for the graft polymer polymerized using the limited number of monomers.

[0119] At least one monomer required for the first stage polymerization of the monomers for which the mixing ratio range is input can be input into the mixing ratio range input unit 41. When the essential monomer required for the first stage polymerization is input, the first stage comprehensive analysis point generation unit 42 generates a first stage comprehensive analysis point for a main chain polymer polymerized from a plurality of monomers including the essential monomer within the mixing ratio range.

[0120] Figure 6This figure shows an example of an input screen 41A for inputting at least one monomer required for the first stage polymerization of monomers for which a mixing ratio range has been input, in the mixing ratio range input section 41. The example of input screen 41A allows input of the minimum and maximum mixing ratio values ​​for each monomer, as well as the required monomers for the first stage polymerization. In the example of input screen 41A, input of monomers 1 and 3 as required monomers for the first stage polymerization is performed.

[0121] When the essential monomers are input, the first-stage comprehensive analysis point generating unit 42 generates a first-stage comprehensive analysis point of a main chain polymer polymerized from a plurality of monomers included in the blending ratio range in which the essential monomers monomer 1 and monomer 3 are input in the first-stage polymerization.

[0122] The number of first-stage monomers used in the first-stage polymerization can be limited by inputting the number of first-stage monomers used in the first-stage polymerization in the mixing ratio range input unit 41. If the number of monomers used in the first-stage polymerization is limited, the first-stage comprehensive analysis point generation unit 42 generates a comprehensive analysis point for the main chain polymer polymerized using the limited number of monomers.

[0123] Figure 7 This figure shows an example of an input screen 41B of the mixing ratio range input unit 41. The number of monomers used in the first stage polymerization can be input on the input screen 41B. In the example of the input screen 41B, "2" is input as the number of first-stage monomers used in the first-stage polymerization.

[0124] The comprehensive analysis point-polymer property value storage unit 46 inputs the comprehensive analysis point generated by the comprehensive analysis point generation unit 44 into the learned model 45 to calculate the graft polymer property value, creates a data set that associates the comprehensive analysis point with the calculated polymer property value, and stores the created data set.

[0125] In this way, the forward problem analyzing unit 40 is configured so that the material designer only needs to input the mixing ratio range of at least one monomer to automatically generate a data set of comprehensive analysis points and physical property values ​​of the graft polymer within the mixing ratio range of all monomers.

[0126] The inverse problem analyzing unit 20 included in the material design apparatus 2 includes a required property input unit 21 and a filtering unit 22. The inverse problem analyzing unit 20 also includes the comprehensive analysis point-polymer property value storage unit 46.

[0127] The filter unit 22 selects data sets that meet the required range of polymer property values ​​inputted in the required property input unit 21 from the comprehensive analysis point-polymer property storage unit 46. The selected data sets are limited to polymers that meet the required range of the input property values ​​and may include all structural isomers.

[0128] The inverse problem analyzing unit 20 and the GUI 30 included in the material design apparatus 2 have the same configuration as those included in the material design apparatus 1 described above.

[0129] Figure 11 FIG. 1 is a block diagram showing the hardware structure of the material design device 1. Figure 11 As shown, the material design device 1 can be physically configured as a computer system including a CPU (Central Processing Unit) 101, a RAM (Random Access Memory) 102 and a ROM (Read Only Memory) 103 as main storage devices, an input device 104 such as a keyboard and a mouse, an output device 105 such as a display, and an auxiliary storage device 106 such as a hard disk. Furthermore, the material design device 2 can also have the same hardware configuration as the material design device 1.

[0130] Figure 1 The functions of the material design device 1 shown are realized by loading the specified computer software (material design program) into the hardware such as CPU 101 and RAM 102, and by operating the input device 104 and output device 105 under the control of CPU 101, while reading and writing data in RAM 102 and auxiliary storage device 106. That is, by executing the material design program of this embodiment on the computer, the material design device 1 is realized as Figure 1 The mixing ratio range input unit 11, the comprehensive analysis point generation unit 12, the required physical property input unit 21, the filter unit 22 and the information display unit 31 function. It is also possible to realize a model creation function of making a learning model 13 that obtains the correspondence between the input information of the monomer mixing ratio and the output information of the polymer physical property value through machine learning; and a data set creation function of inputting the comprehensive analysis point generated by the comprehensive analysis point generation function into the learning model 13 and associating the polymer physical property value calculated with each point of the comprehensive analysis point in the comprehensive analysis point-polymer physical property value storage unit 14. In addition, Figure 1 The comprehensive analysis point-polymer property value storage unit 14 shown can be realized by a part of the storage device (RAM 102, ROM 103, auxiliary storage device 106, etc.) provided in the computer. Figure 1The GUI 30 shown can be realized by using the output device 105 and the input device 104 included in the computer.

[0131] The material design program of this embodiment is stored, for example, in a storage device included in a computer. Alternatively, a portion or all of the material design program may be transmitted via a transmission medium such as a communication line, received by a communication module included in the computer, and recorded (including installed). Alternatively, a portion or all of the material design program may be stored on a portable storage medium such as a CD-ROM, DVD-ROM, or flash memory, and recorded (including installed) in the computer.

[0132] Reference Figure 12 , a material design method using the material design device 1 involved in the embodiment is described.

[0133] In addition, Figure 12 Prior to the analysis process, a process (model creation step) is performed to create a learned model 13. The learned model 13 learns the correspondence between input information including the polymer mixing ratio range and output information including the polymer physical property values ​​through machine learning. The model creation step can be performed by the material design device 1, or it can be performed by another device, with the material design device 1 using the configuration of the learned model 13 created by the other device.

[0134] In step S101, the input of the mixing ratio range and the required range of the physical property value is performed. The mixing ratio range of at least one monomer constituting the polymer to be designed is input through the mixing ratio range input unit 11, and the required range of at least one physical property value of the polymer is input through the required physical property input unit 21 (design condition setting step). For example, the mixing ratio range input unit 11 Figure 3 The input screen 11A shown is displayed on the GUI 30, allowing the material designer to input the mixing ratio range. Figure 8 The input screen 21A shown is displayed on the GUI 30 , and allows the material designer to input a required range of polymer property values.

[0135] In step S102 , the comprehensive analysis point generating unit 12 generates a comprehensive analysis point of a polymer polymerized using a plurality of monomers whose blending ratio range includes at least one monomer whose blending ratio range is input in step S101 (comprehensive analysis point generating step).

[0136] In steps S103 to S106, the comprehensive analysis points generated in step S102 are input into the learned model 13 through the forward problem analysis unit 10 to calculate the polymer property values, and a data set is created by associating the comprehensive analysis points with the calculated polymer property values. The created data set is stored in the comprehensive analysis point-polymer property value storage unit 14 (data set creation step).

[0137] First, in step S103, a comprehensive analysis point is selected. In step S104, the selected comprehensive analysis point is input into the learned model 13 to calculate polymer properties. Furthermore, in step S105, the input comprehensive analysis point and the output material property values ​​of the learned model 13 are correlated. The processing of steps S103 to S105 generates a data set. This data set is stored in the comprehensive analysis point-polymer property storage unit 14.

[0138] In step S106, it is determined whether all comprehensive analysis points have been calculated. If not all comprehensive analysis points have been calculated (false in step S106), the process returns to step S103 and the selection of comprehensive analysis points is repeated. If all comprehensive analysis points have been calculated (true in step S106), data set generation is terminated and the process proceeds to step S107.

[0139] In step S107 , an inverse problem analysis process is performed by selecting a data set that satisfies the required range of polymer property values ​​inputted from the required property input unit 21 from the comprehensive analysis point-polymer property storage unit 14 (filtering step).

[0140] In step S108, the information display unit 31 displays the monomer mixing ratio that satisfies the required range of the polymer physical property value input in step S101 on the GUI 30 based on the data set selected in step S107. The information display unit 31 displays, for example, Figure 10 The output screen 31A shown as an example is displayed on the GUI 30 .

[0141] The effects of this embodiment will be described. The material design device 1 of this embodiment includes, as a forward problem analysis unit 10, a mix ratio range input unit 11 for inputting a mix ratio range of at least one monomer; a comprehensive analysis point generation unit 12 for generating a comprehensive analysis point for a polymer polymerized using multiple monomers, including at least one monomer for which the mix ratio range is input, within the mix ratio range; a comprehensive analysis point-polymer property value storage unit 14 for inputting the comprehensive analysis point generated by the comprehensive analysis point generation unit 12 into a learned model 13 to calculate polymer property values, creating a dataset associating the comprehensive analysis point with the calculated polymer property values, and storing the created dataset. Furthermore, as an inverse problem analysis unit 20, the device includes a required property input unit 21 for inputting a required range of polymer property values, and a filtering unit 22 for selecting datasets that meet the required range of property values ​​input by the required property input unit 21.

[0142] Thus, in this embodiment, during the forward problem analysis, a dataset used in the inverse problem analysis is created and stored in the comprehensive analysis point-polymer property value storage unit 14. Furthermore, during the inverse problem analysis, the dataset stored in the comprehensive analysis point-polymer property value storage unit 14 is referenced to select a dataset that meets the required range of polymer property values. In other words, during the inverse problem analysis, no numerical calculations such as simulations or model calculations are performed; instead, only the dataset stored in the comprehensive analysis point-polymer property value storage unit 14 is searched. This significantly reduces computational costs and allows the optimal solution for the monomer mix ratio required to polymerize a polymer that meets the required range of desired property values ​​to be derived in a short period of time.

[0143] In addition, when using the inverse problem analysis and the inverse problem analysis application machine learning system for the previous simulation, when multiple physical properties are required, the calculation is performed in a way that the optimal solution is gradually obtained while adjusting each physical property in sequence, rather than summarizing the exploration of candidate materials and implementing it in a way that satisfies multiple physical properties at the same time. In many cases, the physical properties of multiple materials are mutually exclusive, and it takes a long time to find the optimal solution by repeated trial and error until the optimal solution is obtained, thereby finding the optimal solution that satisfies the design conditions of the desired material properties. In contrast, in this embodiment, if the output (polymer physical property value) of the learning completion model 13 is set to multiple, and multiple polymer physical property value items are created in the comprehensive analysis point-polymer physical property value storage unit 14, it is possible to summarize the exploration of the monomer mix ratio and implement it in a way that satisfies multiple polymer physical property values ​​at the same time when analyzing the inverse problem. As a result, even when the required range of multiple physical property values ​​is set, the time required to derive the optimal solution can be greatly reduced compared to the previous method.

[0144] Furthermore, the data set stored in the comprehensive analysis point / polymer property value storage unit 14 is information derived from a large number of comprehensive analysis points automatically generated during forward problem analysis. Therefore, the steps for each item in the mix ratio range and the required range of physical property values ​​are sufficiently small, resulting in high resolution. Therefore, even during inverse problem analysis, design conditions that satisfy the required range of physical property values ​​can be predicted with high accuracy.

[0145] The present embodiment has been described above with reference to specific examples. However, the present disclosure is not limited to these specific examples. In these specific examples, examples to which those skilled in the art have applied appropriate design changes are included within the scope of the present disclosure as long as they have the features of the present disclosure. The various elements and their configurations, conditions, shapes, etc. possessed by the above-mentioned specific examples are not limited to the illustrated cases and can be appropriately changed. The various elements possessed by the above-mentioned specific examples can be appropriately changed in combination as long as no technical contradictions are generated.

[0146] This application claims the benefit of priority based on Japanese Patent Application No. 2019-163105, filed on September 6, 2019, and incorporates the entire contents of No. 2019-163105 into this application.

[0147] Explanation of symbols

[0148] 1 Material Design Device

[0149] 2 Material Design Device

[0150] 10 Forward Problem Analysis Department

[0151] 20 Inversion Problem Analysis Department

[0152] 30 GUI

[0153] 11. Mix ratio range input section

[0154] 12 Comprehensive analysis point generation unit

[0155] 13 Learning Completion Model

[0156] 14 Comprehensive Analysis Point - Polymer Property Value Storage

[0157] 21 Required physical property input

[0158] 22 Filtration Unit

[0159] 31 Information display unit

[0160] 40 Forward Problem Analysis Department

[0161] 41 Mix ratio range input section

[0162] 42 Phase 1 Comprehensive Analysis Point Generation Unit

[0163] 43 Phase 2 Monomer Proposal

[0164] 44 Overall comprehensive analysis point generation unit

[0165] 45 Learning Completion Model

[0166] 46 Comprehensive Analysis Point - Polymer Property Value Storage

Claims

1. A material design device for designing a graft polymer obtained by two-stage polymerization of a plurality of monomers, comprising: The learning model has learned the correspondence between the input information of the monomer mixing ratio and the output information of the polymer physical property values; A mixing ratio range input unit for inputting a mixing ratio range of at least one monomer; a required physical property input unit for inputting a required range of at least one physical property value of the polymer; a first-stage comprehensive analysis point generating unit for selecting at least one first-stage monomer to be used in the first-stage polymerization from the monomers for which a mixing ratio range is input, and generating a comprehensive analysis point of a main-chain polymer polymerized using a plurality of monomers, the plurality of monomers including the at least one first-stage monomer within the mixing ratio range; A second-stage monomer proposal section proposes at least one monomer to be used for the second-stage polymerization with the main chain polymer based on the comprehensive analysis points of the first stage; a comprehensive analytical point generating unit for polymerizing a graft polymer from a second-stage monomer and the main chain polymer to generate a comprehensive analytical point for the graft polymer; a comprehensive analysis point-polymer property value storage unit that inputs the comprehensive analysis point into the learned model to calculate the physical property values ​​of the graft polymer, creates a data set that associates the comprehensive analysis point with the calculated physical property values ​​of the graft polymer, and stores the created data set; as well as The filtering unit selects, from the data set, a graft polymer within a required range of the physical property value input by the required physical property input unit.

2. The material design device according to claim 1, In the mixing ratio range input part, the number of monomers used in polymerization is input to limit the number of monomers used in polymerization. A comprehensive analysis of the generation of graft polymers using a limited number of monomers.

3. The material design device according to claim 1 or 2, At least one monomer required for polymerization of the monomers having the input mixing ratio range is input into the mixing ratio range input unit. The integrated comprehensive analysis point generating unit generates a comprehensive analysis point of a graft polymer obtained by polymerizing a plurality of monomers including an essential monomer within a blending ratio range.

4. The material design device according to claim 1 or 2, At least one monomer required for the first stage polymerization of the monomers for which the mixing ratio range is input is input into the mixing ratio range input unit. In the first-stage comprehensive analysis point generating unit, a comprehensive analysis point of a main chain polymer obtained by polymerizing a plurality of monomers including an essential monomer within a blending ratio range is generated.

5. The material design device according to claim 1, When a specific monomer is selected as the first-stage monomer, the learned model learns the monomer and the monomer constituting the main-chain polymer as different substances.

6. A material design method for designing a graft polymer obtained by two-stage polymerization of multiple monomers, comprising: Creating a learning model that learns the correspondence between input information of monomer mix ratios and output information of polymer physical property values; Enter the mix ratio range of at least one monomer; Enter the required range of at least one physical property value of the polymer; Selecting at least one first-stage monomer to be used in the first-stage polymerization from the monomers having a mixing ratio range input thereto, and generating a comprehensive analysis point of a main chain polymer polymerized using a plurality of monomers, the plurality of monomers including the at least one first-stage monomer within the mixing ratio range; Based on the comprehensive analysis points of the first stage, propose at least one monomer used for the second stage polymerization with the main chain polymer; A graft polymer is polymerized from the second-stage monomer and the main chain polymer to generate a comprehensive analytical point of the graft polymer; inputting the comprehensive analysis points into the learned model to calculate the physical property values ​​of the graft polymer, creating a data set in which the comprehensive analysis points are associated with the calculated physical property values ​​of the graft polymer, and storing the created data set; and A graft polymer within a required range of the physical property value inputted in the required physical property input section is selected from the data set.

7. A non-transitory computer-readable storage medium storing a material design program for designing a graft polymer polymerized in two stages from a plurality of monomers, the program being configured to cause a computer to implement the following functions: Creating a learning model that learns the correspondence between input information of monomer mix ratios and output information of polymer physical property values; Enter the mix ratio range of at least one monomer; Enter the required range of at least one physical property value of the polymer; Selecting at least one first-stage monomer to be used in the first-stage polymerization from the monomers having a mixing ratio range input thereto, and generating a comprehensive analysis point of a main chain polymer polymerized using a plurality of monomers, the plurality of monomers including the at least one first-stage monomer within the mixing ratio range; Based on the comprehensive analysis points of the first stage, propose at least one monomer used for the second stage polymerization with the main chain polymer; A graft polymer is polymerized from the second-stage monomer and the main chain polymer to generate a comprehensive analytical point of the graft polymer; inputting the comprehensive analysis points into the learned model to calculate the physical property values ​​of the graft polymer, creating a data set in which the comprehensive analysis points are associated with the calculated physical property values ​​of the graft polymer, and storing the created data set; and A graft polymer within a required range of the physical property value inputted in the required physical property input section is selected from the data set.

Images