Program, information processing apparatus, system, server, terminal, and method

Abstract

To generate recipe information more appropriately.SOLUTION: A program includes: an acquisition unit for acquiring recipe initial information, environmental condition information, and target physical property information; a storage unit which stores a regression equation that specifies a relation between recipe information, the environmental condition information, and the physical property information; and a deriving unit for deriving recipe information which is generated by applying an error function so as to reduce a difference between the target physical property information and predicted physical property information to be obtained by applying the recipe initial information and the acquired environmental condition information to the regression equation. The physical property information and the target physical property information each include density and compression strength. The regression equation includes a regression equation on the density and a regression equation on the compression strength.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to a program, an information processing apparatus, a system, a server, a terminal, and / or a method for information processing related to raw materials of chemical products.

Background Art

[0002] In the preparation of raw materials for chemical products, in order to obtain target properties, it is necessary to generate a recipe (information related to raw materials of chemical products).

[0003] Patent Document 1 describes a method of creating a recipe for obtaining target properties by a computer. However, that method is premised on the relationship between target properties and the mixing weight when using various raw materials being clear, and it does not describe how to create a recipe when that premise is different.

[0004] Also, Patent Document 2 describes a system for deriving recipe information for obtaining target physical properties of a polymer composition. However, in the reference method, after designing an equation for predicting physical properties from recipe information, the user needs to search by trial and error for a combination of recipe values such that the calculation result by the prediction equation satisfies the target value of the physical properties.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0006] Even in various cases where the relationship between the target properties and the mixing weight when using various raw materials is not clear, it is preferable to obtain a recipe (NCO / OH ratio, water addition amount) for obtaining the target foam physical properties (density, compressive strength). In addition, since it is a burden for the user to search for recipe values such that the calculation results according to the formula satisfy the target values of the physical properties through trial and error, it is preferable that there are few such trial-and-error steps.

[0007] Therefore, the technology according to the present application relates to a program, an information processing apparatus, a system, a server, a terminal, and / or a method for information processing for obtaining a recipe for obtaining target foam physical properties by a computational approach.

Means for Solving the Problem

[0008] A computer program according to an embodiment of the present application causes a computer to acquisition means for acquiring initial recipe information, environmental condition information, and target property information, storage means for storing a regression formula that defines the relationship between the recipe information and the environmental condition information and the property information, means for deriving recipe information generated by applying an error function that reduces the difference between the predicted property information obtained by applying the initial recipe information and the environmental condition information to the regression formula and the target property information, and may be a program for functioning as such.

[0009] A method according to an embodiment of the present application includes steps in which a computer acquires initial recipe information, environmental condition information, and target property information, stores a regression formula that defines the relationship between the recipe information and the environmental condition information and the property information, derives recipe information generated by applying an error function that reduces the difference between the predicted property information obtained by applying the initial recipe information and the environmental condition information to the regression formula and the target property information, It may be a method of executing.

[0010] The device according to one embodiment of the present application is An acquisition unit that acquires initial recipe information, environmental condition information, and target property information A storage unit that stores a regression equation that defines the relationship between recipe information and environmental condition information, and property information A derivation unit that derives recipe information generated by applying an error function that reduces the difference between predicted property information obtained by applying the initial recipe information and the environmental condition information to the regression equation, and the target property information It may be a device including.

Effect of the Invention

[0011] According to one embodiment of the present invention, a recipe value can be calculated more appropriately.

Brief Description of the Drawings

[0012]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

[0013] 1. Functions of the information processing apparatus The information processing apparatus according to the present application may be of various types. An example of the information processing apparatus will be described with reference to FIG. 1. The information processing apparatus 10 may include an arithmetic unit 11, a storage device 12, an input device 13, an output device 14, a communication IF 15, and a line 16 connecting these.

[0014] The arithmetic unit is a device that can execute program instructions. The arithmetic unit may be, for example, a processor, a CPU, an MPU, etc. The arithmetic unit may be capable of sequential calculation or parallel calculation. The arithmetic unit may have a graphics processing unit, a digital signal processor, etc.

[0015] The storage device is a device that records information. The storage may be temporary storage or long-term storage. The storage device may be either an internal memory or an external memory or both. Also, the storage device 13 may be a magnetic disk (hard disk), an optical disk, a magnetic tape, a semiconductor memory, etc. Further, although not shown, the storage device may be a storage device via a network or a storage device on a cloud via a network. The storage device can store a program for executing the technology according to the present application, and can also appropriately record data necessary for executing the processing of the technology according to the present application. Also, the storage device may include a database.

[0016] The line may be any that can transmit information between devices. The line may be, for example, a bus. The bus has a function of transmitting information between the arithmetic unit, the storage device, the input device, the display device, and the communication IF.

[0017] An input device is for inputting information. Examples of input devices include keyboards, mice, touch panels, or pen-type pointing devices. Note that the input device may have other functions so that the touch panel has a display function.

[0018] A display device has a function of displaying information. For example, a liquid crystal display, a plasma display, an organic EL display, etc. may be mentioned, but in short, any device that can display information is acceptable. Also, it may be partially equipped with an input device such as a touch panel.

[0019] A communication IF has a function of being able to communicate information with an external network. The communication IF may have any connection form. For example, USB, IEEE1394, Ethernet (registered trademark), PCI, SCSI, etc. are acceptable. Note that the network may be either wired or wireless, and in the case of wired, an optical fiber or a coaxial cable, etc. may be used.

[0020] Also, the above was described by taking as an example the case where the arithmetic unit is executed based on a program provided in the storage device. However, as one form in which the above bus, arithmetic unit, and storage device are combined, the information processing according to the system of the present application may be realized by a programmable logic device capable of changing the hardware circuit itself or a dedicated circuit in which the information processing to be executed is determined. When realized by a programmable logic device or a dedicated circuit, there is an advantage that the technology according to the present application can be processed at a higher speed.

[0021] The information processing device may be a dedicated information processing device for part or all of the technology according to the present application, or may be an information processing device capable of executing technologies other than the technology according to the present application. The information processing device may be a server, a workstation, a desktop personal computer, a laptop personal computer, a notebook personal computer, a PDA, a mobile phone, a smartphone, etc.

[0022] In FIG. 1, the description was given for one information processing apparatus, but the information processing apparatus capable of processing the information according to the present application may be composed of a plurality of information processing apparatuses. The plurality of information processing apparatuses may be internally connected or externally connected via a network. Further, when the information processing apparatus is composed of a plurality of information processing apparatuses, even if their owners or administrators are different, when executing the technology described in this process, it is sufficient that the information processing apparatus can be used due to having access rights or the like. Further, the information processing apparatus may be a physical existence or a virtual one. For example, using cloud computing, the information processing apparatus may be virtually realized. Further, although FIG. 1 explains the basic configuration of the information processing apparatus according to the present application, when cooperating with other information processing apparatuses, etc., it may be in a form without an input device or a display device. In the present application, the term "system" may be used for something composed of one or a plurality of information processing apparatuses. Further, the term "system" may include a part or all of the technologies according to the technology of the present application. Further, the term "system" may indicate only a server, only a terminal, a server and a terminal, only a cloud, or a cloud and a terminal.

[0023] 2. Outline of the technology according to the present application FIG. 2 schematically explains the outline of the technology according to the present application. There may be a case where past manufacturing data is collected and a regression equation for predicting the physical properties of the form can be constructed from the recipe and manufacturing environment conditions. In this case, by substituting the initial value of the recipe and the environmental conditions into the regression equation, the physical properties of the form when such a recipe is used can be predicted. Here, when a target value of the physical properties of the form is given, by applying an error function L so as to minimize the error between the predicted value of the form physical properties and the target value of the form physical properties, there may be a case where the optimum value of the recipe corresponding to the target value of the form physical properties can be derived. Regarding an example in which such an outline is embodied, it will be described below as the first embodiment.

[0024] Further, FIG. 3 schematically illustrates an overview of another technique according to the present application. It may be possible to collect past manufacturing data and calculate the average value of the resin strength for each form type therefrom. And it may be possible to construct a regression equation for predicting the form physical properties from the resin strength, the recipe, and the manufacturing environment. As will be described later, by calculating the average value of the resin strength, there is an advantage that a regression equation applicable to various form types can be constructed. The derivation of the optimal recipe value using such a regression equation is the same as the overview regarding the above-described first embodiment, and the details thereof will be described as the second embodiment below.

[0025] 3. Embodiments 3.1. First Embodiment Next, as an embodiment of the system according to the present invention, the system according to the first embodiment will be described.

[0026] The system of this embodiment may be a program composed of a plurality of functions including a macro executed on Excel. In this case, the program composed of functions has an advantage that it can utilize the functions implemented in Excel.

[0027] First, the system of this embodiment acquires a form type, target property information, environmental condition information, and recipe initial information. In the present documents, information to be substituted into the regression equation, such as environmental condition information and recipe initial information, may be referred to as "regression equation input information". The system of this embodiment implemented by a macro on Excel may acquire this information using the input function of Excel.

[0028] FIG. 4 is a diagram showing an example of receiving input information in the system of this embodiment. In this example, the system of this embodiment is configured to be able to receive the input of target property information, environmental condition information, and recipe initial information.

[0029] The form type is information for identifying the form that the user of this system wishes to manufacture.

[0030] The target object property information is the information on the characteristics of the form that the user of this system hopes to manufacture. The target object property information may include the density of the form and the compressive strength of the form.

[0031] Here, the form may be a polyurethane foam. Further, the polyurethane foam may be a soft or hard foam. Also, in this specification, the "soft polyurethane foam" refers to a urethane foam molded using a blended polyol containing 50% by mass or more of a polyol having a hydroxyl value of 56 mgKOH / g or less and having a product density of less than 100 kg / m3. The "rigid polyurethane foam" refers to a urethane foam molded using a blended polyol containing 50% by mass or more of a polyol having a hydroxyl value of 300 to 500 mgKOH / g.

[0032] Also, the application uses of the form according to this application may be various and its use is not limited. For example, the technology according to this application may be a foam for furniture, vehicle materials, consumer products (such as sponges and cushions) in the case of soft foams, and may be a foam for building materials, vehicle materials, etc. in the case of rigid foams.

[0033] Recipe initial information may be information that is initially required in a manufacturing apparatus that manufactures a foam. For example, the recipe initial information may include the NCO / OH ratio and / or the amount of water added. The NCO / OH ratio may be a ratio in a predetermined unit. The amount of water added may be the amount of moisture added per predetermined unit, and may be information expressed in weight or volume. The recipe initial information may be set based on the past knowledge and experience of the user of this system, but may also be a value tentatively input even when the user of this system has no past knowledge or experience. Also, the recipe initial information may be the recipe initial information or target property information used when this system was last used, or may be information obtained from another system by cooperating with another system. Furthermore, the recipe initial information may be a combination of the NCO / OH ratio and the amount of water added corresponding to a location selected by the user within the area that displays the physical property error described later.

[0034] The environmental condition information may include the temperature, pressure, and / or humidity.

[0035] The target property information may include the density of the foam and / or the compressive strength of the foam.

[0036] Next, the system of this embodiment substitutes the acquired regression equation input information into the regression equation. The regression equation may be, for example, the following equation. Here, as the regression equations, Equations 1 to 3 are exemplified, but the regression equations used are not limited to these and may be in various forms. In these equations, X1 represents the index (NCO / OH ratio), X2 represents the amount of moisture, X3 represents the absolute humidity, X4 represents the pressure, X5 represents the temperature, Y1 represents the foam density, Y2 represents the compressive strength, and w and v represent coefficients. Also, these coefficients of w and v may be determined manually in advance or automatically calculated using information related to the foam, and these coefficients of w and v may use, for example, sparse regression in addition to simple linear regression, and may be determined using Lasso or Elastic Net.

[0037]

Number

[0038]

Number

[0039]

Number

[0040] The above-described regression equations shown may be capable of expressing the density and compressive strength of the foam with respect to the form. These regression equations may enable a person to accurately express such a specific form based on information regarding a specific form type (environmental conditions such as humidity, atmospheric pressure, and temperature, and recipe initial values such as the amount of water added and the NCO / OH ratio, physical property information such as the density and compressive strength of the foam), and the coefficients may be set in advance so that a person can express such a specific form. The system of the present embodiment may store, in association with each other, the regression equation with coefficients set in this way and the corresponding form type. The system of the present embodiment may use the regression equation corresponding to the input form type described above.

[0041] Next, the system of the present embodiment uses an error function to calculate recipe information such that the difference between the density and compressive strength of the foam based on the recipe initial information obtained by the above substitution and the density and compressive strength within the target physical property information (which may also be referred to as "physical property error" in this application documents) decreases.

[0042] Here, various methods for calculating the error between two arguments may be used as the error function for calculating the physical property error. Here, since it is the calculation of the error of a plurality of values like two arguments, if one tries to eliminate the error simply by subtraction, the value with a larger absolute value will be emphasized. Therefore, after normalizing a plurality of values like two arguments, the calculation is performed. As a result, there is an advantage that the error between a plurality of values can be appropriately evaluated, rather than just an error function that emphasizes a value with a large absolute value.

[0043] The system of this embodiment may utilize the Generalized Reduced Gradient method for the regression equation implemented in Excel.

[0044] The system of this embodiment derives recipe information (NCO / OH ratio and moisture content) corresponding to target physical property information (density and compressive strength) such that the physical property error decreases, which is generated by using the generalized reduced gradient method, the regression equation, and the error function. Here, "deriving" means that the recipe information (NCO / OH ratio and moisture content) includes both cases where it is obtained based on the generalized reduced gradient method implemented inside the system of this embodiment and cases where it is obtained based on the generalized reduced gradient method implemented outside the system of this embodiment. Also, the recipe information (NCO / OH ratio and moisture content) corresponding to the target physical property information (density and compressive strength) with a reduced physical property error may be the recipe information corresponding to the target physical property information with the physical property error reduced to a predetermined value or less, or may be the recipe information corresponding to the target physical property information with the physical property error reduced to 0.

[0045] As described above, in the production line of the foam, there was a recipe (NCO / OH ratio, water addition amount) to obtain the target foam physical properties (density, compressive strength). However, since the foam physical properties were also affected by the surrounding environment (temperature, humidity, etc.) of the production line, the foam physical properties obtained even when produced according to the recipe often deviated from the target values. Although some elimination of the deviation could be expected based on the knowledge and experience of the manufacturer, in order to further stabilize the quality, an impersonal recipe optimization method based on a computational approach was desired. Since the technology according to the present application is a computational approach that uses the above regression formula and generalized reduced gradient method, there is an advantage that the recipe information (NCO / OH ratio and water content) corresponding to the target physical property information (density and compressive strength) that reduces the physical property error as described above can be derived impersonally. Further, when the physical property information can be expressed with a certain degree of accuracy by the regression formula, it is applicable even when the relationship between the target characteristics and the mixing weights when using various raw materials is not clear, and since the user does not need to search by trial and error, there is an advantage that the recipe can be created without burdening the user.

[0046] Note that the system of this embodiment may be configured to display the physical property error corresponding to the initial recipe value.

[0047] For example, the system of this embodiment may be configured to display a Cartesian coordinate with the NCO / OH ratio on the vertical axis and the water addition amount on the horizontal axis. Here, in such a Cartesian coordinate, the values corresponding to each NCO / OH ratio and each water addition amount may be the physical property errors between the physical property information (density, compressive strength) when these NCO / OH ratios and water addition amounts are applied to the regression formula and the input target physical property information.

[0048] Such physical property errors may be displayed separately in various manners. For example, the information for separate display may include at least one of color, pattern, etc. as values corresponding to the NCO / OH ratio and each water addition amount in Cartesian coordinates. In this case, information for separate display may be displayed corresponding to a physical property error within a predetermined range. For example, in the case of color, color 1 when the physical property error is less than 0.5, color 2 when the physical property error is 0.5 or more and less than 1.0, etc. Also, a display explaining that these pieces of information indicate a predetermined physical property error may be provided. For example, information such as color 1 when the physical property error is less than 0.5, color 2 when the physical property error is 0.5 or more and less than 1.0, etc. in the above example may be displayed.

[0049] FIG. 5 illustrates one aspect of displaying such physical property errors. In this figure, within the Cartesian coordinates, regions 501 to 503 are displayed in a different manner from others, indicating the difference in physical property errors. In this example, particularly, region 503 is shown in a different manner by a darker display than the other regions 501 and 502, and is depicted as an example indicating a smaller physical property error.

[0050] When the system of this embodiment includes a configuration for displaying a physical property error corresponding to such a recipe initial value, the following processing may be performed.

[0051] First, the system of this embodiment acquires a form type, target physical property information, environmental condition information, and recipe initial information.

[0052] Next, the system of this embodiment identifies a regression equation corresponding to the form type.

[0053] Next, the system of this embodiment sets a range in which the recipe information can be applied based on the recipe initial information. For example, the system of this embodiment sets values obtained by increasing and decreasing a predetermined amount from the NCO / OH ratio and each water addition amount related to the recipe initial value as the lower limit value and upper limit value of the NCO / OH ratio and the water addition amount. Note that these values may correspond to the lower limit value and upper limit value of the Cartesian coordinates in the display indicating the physical property error.

[0054] The system of this embodiment generates corresponding physical property information for each combination of values at predetermined intervals within the ranges of the lower and upper limits of the above-mentioned NCO / OH ratio and water addition amount, using the environmental condition information and the regression formula corresponding to the above-mentioned form type. For example, consider a case where the lower and upper limits of the NCO / OH ratio are set to 80 to 150, the lower and upper limits of the water addition amount are set to 300 to 500, the predetermined interval of the NCO / OH ratio is 10, and the predetermined interval of the water addition amount is 30. In this case, the system of this embodiment generates physical property information using the environmental condition information, the regression formula corresponding to the above-mentioned form type, an NCO / OH ratio of 80, and a water addition amount of 300, and sets it as the physical property information at the location associated with an NCO / OH ratio of 80 and a water addition amount of 300. Next, the system of this embodiment generates physical property information using the environmental condition information, the regression formula corresponding to the above-mentioned form type, an NCO / OH ratio of 90, and a water addition amount of 330, and sets it as the physical property information at the location associated with an NCO / OH ratio of 90 and a water addition amount of 330. In this way, physical property information is generated at predetermined intervals within the ranges of the NCO / OH ratio and the water addition amount. Here, the predetermined amounts for setting the lower and upper limits of the NCO / OH ratio and the water addition amount and / or the predetermined intervals for calculating the physical property information may each be stored in advance in the system of this embodiment, or may be values input by the user.

[0055] Next, the system of this embodiment may display the physical property error, which is the difference between each physical property information and the target physical property information, as the values at the locations of the corresponding NCO / OH ratio and water addition amount.

[0056] In such a case, the user has the advantage of being able to understand at a glance the physical property errors when these values are input without inputting other recipe initial values other than the recipe initial values input by the user himself / herself. In particular, depending on the applied regression formula and the applied generalized reduced gradient method, the user can notice that recipe initial values that cannot reduce the physical property error have been selected.

[0057] In addition, as a configuration for displaying physical property errors, although Cartesian coordinates were described above, any graph that can display three components capable of displaying values corresponding to each NCO / OH ratio and each water addition amount and that includes a coordinate system capable of such display is acceptable, and it is not limited to Cartesian coordinates. For example, cylindrical coordinates, spherical coordinates, elliptical coordinates, parabolic coordinates, etc. may be used, or even modified versions of these may be used.

[0058] The flowchart of FIG. 6 briefly shows the flow of the system of this embodiment.

[0059] As described above, the system of this embodiment has been described by way of an example implemented as a function executed on Excel, but it is not limited to such an example and may be executed on various systems. In particular, as a sparse regression method, Lasso and Elastic Net, and as a method for obtaining recipe optimum values, the Generalized Reduced Gradient method can be used respectively, and it may be a system executed on R, which is a program execution environment for statistical analysis, or Python, which is an execution environment.

[0060] Moreover, not limited to these, the system of this embodiment may be an aspect that incorporates the Generalized Reduced Gradient method as an implementation form. In this case, there is an advantage that it is not necessary to use the Generalized Reduced Gradient method in the execution environment.

[0061] 3.2. Second Embodiment Next, as an embodiment of the system according to the present invention, the system according to the second embodiment will be described. The system of this embodiment uses one type of regression formula that can be used for various form types for each piece of physical property information. Note that only the parts different from the first embodiment will be described, and the parts the same as the first embodiment will be omitted from the description.

[0062] That is, in the system of the first embodiment, the regression formula used is the one corresponding to the form type. In this case, when the number of form types to be used particularly increases, the following problems may occur. That is, first, when the information dependent on the form type is not included in the regression formula, since the applicable regression formula is different for each form type, as the number of form types increases, the number of regression formulas increases, making implementation difficult and at the same time increasing the burden on humans to determine in advance. Also, it is difficult to construct a regression formula with high predictability for form types with few samples. In particular, when creating a prediction formula that incorporates the form type as a fixed effect, as the number of form types to be handled increases, the number of explanatory variables becomes enormous, making it difficult to perform optimal modeling, and there is also a problem that it cannot cope with unknown form types.

[0063] Therefore, the system of the second embodiment is one that deforms and uses a predetermined number of regression formulas in advance so that they can be used for various form types equal to or more than such a predetermined number.

[0064] In this regard, the inventor conducted the following examination. That is, the compressive strength of the foam is directly proportional to the foam density, but even if the foam density is the same, if the strength of the resin itself is high, the compressive strength of the foam will be high. For example, FIG. 7 schematically shows the relationship between the density of the foam and the compressive strength of the foam. Therefore, it can be considered that the compressive strength per density (compressive strength / density) reflects the strength derived from the resin.

[0065] From the above viewpoints, the inventors noticed that, for example, when a regression formula includes the compressive strength per density (the compressive strength divided by the density), it can be used by changing such compressive strength per density according to the form type.

[0066] Based on the above technical idea, when the compressive strength per density described above is defined as σ_i, the following regression formula can be generated.

[0067]

Equation

[0068] In Equation 4, depending on the form type, regression equations can be made to correspond by ws*σ_i and vs*σ_i.

[0069]

Number

[0070] Here, in Y1, by ws*σ_i + w1s*X1*σ_i + w2s*X2*σ_i + w3s*X3*σ_i + w4s*X4*σ_i + w5s*X5*σ_i, and in Y2, by vs*σ_i + v1s*X1*σ_i + v2s*X2*σ_i + v3s*X3*σ_i + v4s*X4*σ_i + v5s*X5*σ_i, regression equations corresponding to the form type can be used. Equation 5 can also represent the case where there is an interaction between the compressive strength per density, the recipe information, and / or the environmental condition information.

[0071] Note that the selection of Equation 4 and Equation 5 and the coefficients w and v of these equations may be selected based on values determined in advance by a person, or may be selected by automatic processing. For example, for the model without interaction of Equation 4 and the model with interaction of Equation 5, in each model, the learning data is regressed, and the one with a higher coefficient of determination or a smaller mean prediction error may be adopted, and these may be selected by automatic processing.

[0072] Next, the processing flow of the system of the present embodiment using the compressive strength per density as described above will be described with reference to FIG. 8.

[0073] First, the system of this embodiment acquires and stores information on the compressive strength and density of the foam. Here, the information on the compressive strength and density of the foam may be acquired by a file or the like that includes information on the foam. For example, FIG. 9 shows an example in which a plurality of sets of compressive strength and density (compressive strength 1 and density 1, compressive strength 2 and density 2, etc.) are associated with each foam type (foam type IDs 1 to 3). The system of this embodiment may calculate the compressive strength per density corresponding to one foam type based on such a file. In this case, when there are a plurality of sets of compressive strength and density, the compressive strength per density may be calculated by calculating the average value of these. Further, these compressive strength and density information may be used to automatically generate regression equations for the above-described equations 4 and 5. Here, as described above, for the regression equation, for example, the coefficients of these w and v may be dynamically determined by a method automatically calculated using the compressive strength and density information, and in addition to simple linear regression, for example, sparse regression may be used for the coefficients of these w and v, and Lasso or Elastic Net, which are prior arts, may be used for determination. Note that, as described above, instead of automatic and dynamic calculation, a regression equation constructed in advance manually or automatically may be used for these regression equations. When a regression equation is prepared in advance, the regression equation corresponding to the target foam may be specified using the above-described compressive strength per density. For example, in advance, regression equations corresponding to a plurality of foams are constructed manually or automatically, and among the compressive strengths per density in these regression equations, the regression equation related to the compressive strength per density corresponding to the compressive strength per density based on the acquired compressive strength and density information may be specified. Here, for the correspondence of the compressive strength per density, among the compressive strengths per density in each regression equation, the compressive strength per density that is closer to the acquired compressive strength per density than other compressive strengths per density may be selected. Here, the closeness may be such that the difference between the compressive strengths per density is small.

[0074] Further, when the system of the present embodiment does not acquire information on the compressive strength and density of such a form, it may calculate the compressive strength per unit density based on the target property information (i.e., the compressive strength and density of the target form) that is input next. Also in this case, similar to the information on the compressive strength and density based on the above-described file, the calculation and identification of the regression equation may be performed.

[0075] Next, the system of the present embodiment acquires and stores the form type, target property information, environmental condition information, and recipe initial value. Here, the form type may correspond to the form type in the information on the compressive strength and density of the above-described form. Note that when the information on the compressive strength and density of the form is information for only one form type, or when calculating the compressive strength per unit density from the target property information, it is not necessary to acquire the form type information.

[0076] Next, the system of the present embodiment substitutes the compressive strength per unit density corresponding to the form type, environmental condition information, and recipe initial value into the regression equation to generate property information corresponding to the recipe initial value. Here, as the regression equation, Equation 4 or Equation 5 may be selected by automatic processing.

[0077] Thereafter, it is the same as in the first embodiment. The system of the present embodiment uses the error function based on the above-described property information, the regression equation, and the generalized reduced gradient method to derive recipe information corresponding to the target property information and displays the recipe information. Also, similar to the first embodiment, physical property errors corresponding to different plural NCO / OH ratios and water addition amounts may be calculated and those physical property errors may be displayed.

[0078] 3.3. Third Embodiment The system of the third embodiment is a case where the system of the first or second embodiment is performed in a server-client system.

[0079] Referring to FIG. 10, the system according to the third embodiment may be configured such that the system 101, the terminal devices 102A and 102B can be connected to each other via the network 103. Although this figure illustrates the case where there are two terminals, the number of terminals is not limited and may be one or three or more.

[0080] The terminal devices 102A and 102B may mainly implement the configurations related to input and output among the functions in the system according to the first or second embodiment, and the server 103 may implement the configurations not implemented by the terminals.

[0081] For example, the terminal devices 102A and 102B may implement the configurations related to input (for example, the configurations related to the acquisition of information such as form-related information, target property information, environmental condition information, and initial recipe information) and output (for example, the display of information to assist the user in inputting information when acquiring the above-mentioned information, the display of recipe information (NCO / OH ratio and moisture content) corresponding to the target property information (density and compressive strength) with reduced physical property error, the display of physical property error, etc.) in the system according to the first or second embodiment, as well as the processes related to the processing of these information.

[0082] For example, the server may perform processes of substituting the regression formula input information into the regression formula to calculate the regression formula and calculating the physical property information, and processes of calculating the recipe information (NCO / OH ratio and moisture content) corresponding to the target property information (density and compressive strength) such that the physical property error is reduced, which is generated by using the generalized reduced gradient method, the regression formula, and the error function. In particular, since the above-mentioned process of reducing the physical property error increases the calculation processing load, advanced calculation processing capabilities such as parallel calculation processing may be used.

[0083] In addition, the process of calculating the physical property error at predetermined intervals within the range of the lower limit value and the upper limit value of the NCO / OH ratio and the water addition amount may also increase the calculation processing load depending on the range and the interval, and the calculation of the compressive strength per density of the regression formula may also increase the calculation processing load depending on the amount of information related to the form. Therefore, advanced calculation processing capabilities such as parallel calculation processing may be used.

[0084] Although the server - type embodiments have been described above, in addition to or instead of the server, cloud - type embodiments may also be applicable. In the cloud - type embodiment, the owner of the information processing device that constitutes the cloud, the administrator who runs the cloud, and the administrator and / or user of the system according to the present application may be the same or different from each other.

[0085] 3.4. Other aspects The first aspect of the invention according to the present application is The computer program according to the first aspect causes "a computer to an acquisition means for acquiring initial recipe information, environmental condition information, and target property information, a storage means for storing a regression formula that defines the relationship between recipe information, environmental condition information, and property information, a means for deriving recipe information generated by applying an error function that reduces the difference between the predicted property information obtained by applying the initial recipe information and the environmental condition information to the regression formula and the target property information. Here, the term "acquisition" may be a concept including a process of acquiring information input by a user using an input device, a process of acquiring information from a storage device in the information processing device in which the acquisition means is implemented, or a process of acquiring information from an information processing device different from the information processing device in which the acquisition means is implemented. The process of acquiring information from the storage device in the information processing device in which the acquisition means is implemented may include a process of acquiring information calculated using the technology according to the present application. For example, it may include a process of acquiring the NCO / OH ratio and the amount of water added set or calculated in a configuration for displaying property errors.

[0086] The computer program according to the second aspect is, in the first aspect, "wherein the property information and the target property information each include density and compressive strength, and the regression formula includes a regression formula for density and a regression formula for compressive strength."

[0087] The computer program according to the third aspect is the one for "causing the computer to generate information indicating the difference based on the second recipe information, which is associated with the second recipe information including the NCO / OH ratio and the amount of water added, on a graph showing the relationship between the NCO / OH ratio and the amount of water added, for functioning as" in the above first aspect or the above second aspect.

[0088] The computer program according to the fourth aspect is the one for "the regression equation includes a constant indicating the relationship between density and compressive strength corresponding to the form type" in any one of the above first to third aspects.

[0089] The computer program according to the fifth aspect is the one for "causing the computer to generate a regression equation using information indicating the relationship between density and compressive strength, for functioning as" in any one of the above first to fourth aspects.

[0090] The computer program according to the sixth aspect is the one for "causing the computer to acquire first density information for the first form type and first compressive strength information corresponding to the first density information, acquire second density information for the second form type and second compressive strength information corresponding to the second density information, a second acquisition means, acquire a third form type, for functioning as a program, wherein the regression equation generation means generates information indicating the relationship between the density and the compressive strength using the density information and the compressive strength information corresponding to the third form type, and generates a regression equation."

[0091] The computer program according to the seventh aspect is the one where "the information indicating the relationship between the density and the compressive strength is the compressive strength per density" in any one of the first to sixth aspects described above.

[0092] The computer program according to the eighth aspect is the one where "the program does not include the error function" in any one of the first to seventh aspects described above.

[0093] The computer program according to the ninth aspect is the one where "the program includes the error function" in any one of the first to eighth aspects described above.

[0094] The computer program according to the tenth aspect is the one where "the recipe information includes the NCO / OH ratio and the amount of water added, and the environmental condition information includes the temperature, pressure, and humidity" in any one of the first to ninth aspects described above.

[0095] The computer program according to the eleventh aspect is the one where "the physical property information is the physical property information of the polyurethane foam" in any one of the first to tenth aspects described above.

[0096] The computer program according to the twelfth aspect is the one where "the polyurethane foam is a soft or hard foam" in any one of the first to eleventh aspects described above.

[0097] The method according to the thirteenth aspect is "a computer acquisition step of acquiring initial recipe information, environmental condition information, and target physical property information, a storage step of storing a regression equation that defines the relationship between the recipe information and the environmental condition information and the physical property information, a derivation step of deriving recipe information generated by applying an error function that reduces the difference between the predicted physical property information obtained by applying the initial recipe information and the environmental condition information to the regression equation and the target physical property information. which executes

[0098] The apparatus according to the 14th aspect includes: an acquisition unit that acquires initial recipe information, environmental condition information, and target property information; a storage unit that stores a regression formula defining the relationship between the recipe information, the environmental condition information, and the property information; a derivation unit that derives recipe information generated by applying an error function that reduces the difference between predicted property information obtained by applying the initial recipe information and the environmental condition information to the regression formula and the target property information; and is provided with

[0099] The processes and procedures described in this specification can be realized not only by those explicitly described in the embodiments, but also by software, hardware, or a combination thereof. Further, the processes and procedures described in this specification can be implemented as a computer program and executed on various computers. Also, the computer program can be recorded on a non-transitory recording medium. The recording medium may be a computer-readable recording medium.

Description of Reference Numerals

[0100] 10 Information processing apparatus 11 Arithmetic unit 12 Storage device 13 Input device 14 Display device 15 Communication IF 16 Bus 19 Network 501, 502, 503, Displays related to physical property errors 101 Server 102A, 102B Terminal devices 103 Network

Claims

1. A computer, an acquisition means for acquiring initial recipe information, environmental condition information, and target property information including a target density and a target compression strength; a storage means for storing a regression equation that defines the relationship between the recipe information and the environmental condition information and the property information including the density and the compression strength; a means for calculating a property error between predicted property information including a predicted density and a predicted compression strength obtained by applying the initial recipe information and the environmental condition information to the regression equation and the target property information by an error function, and deriving recipe information including an NCO / OH ratio and a water addition amount by a generalized reduced gradient method so as to reduce the property error; a program for causing the computer to function as such.

2. The regression equation includes a regression equation for density and a regression equation for compression strength, The program according to claim 1.

3. The computer, a means for generating information indicating the difference based on the second recipe information, which is associated with the second recipe information including the NCO / OH ratio and the water addition amount, on a graph showing the relationship between the NCO / OH ratio and the water addition amount; The program according to claim 1 or 2 for causing the computer to function as such.

4. The regression equation includes a constant indicating the relationship between density and compression strength corresponding to the form type, The program according to any one of claims 1 to 3.

5. The computer, a regression equation generation means for generating the regression equation using information indicating the relationship between density and compression strength; The program according to any one of claims 1 to 4 for causing the computer to function as such.

6. The computer, first density information for the first form type and first compression strength information corresponding to the first density information, second density information for the second form type and second compression strength information corresponding to the second density information, a second acquisition means for acquiring; a third acquisition means for acquiring a third form type, a program for causing the computer to function as such, wherein the regression equation generation means generates information indicating the relationship between the density and the compression strength using the density information and the compression strength information corresponding to the third form type, and generates a regression equation. The program according to claim 5.

7. The information indicating the relationship between the density and the compression strength is the compression strength per density, The program according to claim 6.

8. The environmental condition information includes temperature, pressure, and humidity, The program according to any one of claims 1 to 7.

9. The physical property information is the physical property information of the polyurethane foam. The program according to any one of claims 1 to 8.

10. The polyurethane foam is a soft or hard foam. The program according to claim 9.

11. A computer An acquisition step of acquiring initial recipe information, environmental condition information, and target physical property information including a target density and a target compression strength; A storage step of storing a regression equation that defines the relationship between the recipe information and the environmental condition information and the physical property information including the density and the compression strength; A derivation step of calculating a physical property error between the predicted physical property information including the predicted density and the predicted compression strength obtained by applying the initial recipe information and the environmental condition information to the regression equation and the target physical property information by an error function, and deriving recipe information including the NCO / OH ratio and the water addition amount by the generalized reduced gradient method so as to reduce the physical property error; A method of executing.

12. An acquisition unit that acquires initial recipe information, environmental condition information, and target physical property information including a target density and a target compression strength; A storage unit that stores a regression equation that defines the relationship between the recipe information and the environmental condition information and the physical property information including the density and the compression strength; A derivation unit that calculates a physical property error between the predicted physical property information including the predicted density and the predicted compression strength obtained by applying the initial recipe information and the environmental condition information to the regression equation and the target physical property information by an error function, and derives recipe information including the NCO / OH ratio and the water addition amount by the generalized reduced gradient method so as to reduce the physical property error; An apparatus comprising.

13. The computer has a memory. The program according to claim 1.

14. The apparatus has a memory. The apparatus according to claim 12.

Images