Cosmetic color raw material acquisition device

Abstract

The present disclosure relates to a raw material acquisition device that extracts color raw materials for cosmetics having a target color, and when a target color value is input, the raw material acquisition device extracts raw material information using a genetic algorithm.

Description

[Technical field]

[0001] The present disclosure relates to a cosmetic color raw material acquisition device, and more particularly to prediction and formulation of cosmetic color raw materials using a genetic algorithm. [Background technology]

[0002] Among the various cosmetic product groups, color cosmetics are a product group where market needs change very rapidly. If we can introduce new technology and improve the efficiency of research and development of new color products, we can expect to increase our brand value and achieve economic benefits through quicker responses than our competitors.

[0003] However, since developing a new color requires a lot of time and experimentation, there is a problem that the development of a new color requires not only high costs but also a long time. Summary of the Invention [Problem to be solved by the invention]

[0004] The present disclosure aims to provide a raw material acquisition device, a raw material acquisition method, and a raw material acquisition system for extracting color raw materials for cosmetics.

[0005] The present disclosure aims to minimize the time, cost, and effort involved in developing new colors.

[0006] The present disclosure aims to provide a raw material acquisition device that is capable of formulating raw material color tones for cosmetics taking into consideration restrictions such as the maximum number of pigments used in the formulation, the range of content ratios for each pigment, and the overall content ratio of the pigments used.

[0007] The present disclosure aims to provide a raw material acquisition device capable of formulating cosmetic color raw materials capable of expressing a nearly infinite and extremely detailed range of colors. [Means for solving the problem]

[0008] The raw material acquisition device according to the present disclosure aims to replace the series of processes required for developing a new color, such as dye prescription, color matching experiments, and correction experiments, with machine learning.

[0009] The raw material acquisition device according to the present disclosure allows the user to set options according to their needs, such as the maximum number of dyes used in the recipe, the range of content ratios for each dye, and the overall content ratio of the dyes used.

[0010] The raw material acquisition device according to the present disclosure can include an input unit that receives a target color value, a prescription sample group generation unit that generates a prescription sample group according to the target color value, a prescription sample improvement unit that outputs the prescription sample group as an improved sample group using a genetic algorithm, and an output unit that outputs raw material information for manufacturing a color cosmetic having the target color value based on the improved sample group output from the prescription sample improvement unit.

[0011] The input unit may further receive optional items, and the prescription sample improving unit may output the raw material information reflecting the optional items.

[0012] The options may include at least one of the maximum number of ingredients used, the range of content ratios for each ingredient, and the total ingredient content.

[0013] The prescription sample improving unit includes a sample candidate derivation unit that outputs a combination sample based on an input existing sample, a sample evaluation unit that calculates a first color difference between a predicted color of the existing sample and the target color value and a second color difference between a predicted color of the combination sample and the target color value, and a sample having a smaller value of the first color difference and the second color difference. Improved samples within the improved sample group and a sample selector for outputting the sample as

[0014] When an optional item is present, the prescription sample improving unit may compare a first color difference reflecting a penalty due to the optional item with a second color difference reflecting a penalty due to the optional item.

[0015] The raw material acquisition device may further include a color prediction calculation unit that calculates a predicted color of the existing sample and a predicted color of the combination sample through a regression model.

[0016] The sample candidate derivation unit may select one of the existing samples and one of the remaining existing samples as a random sample, mutate the random sample, and then combine the random sample with the one of the existing samples to generate the combined sample.

[0017] The raw material acquisition device may further include an improved sample group convergence confirmation unit that outputs the improved sample group to the prescription sample improvement unit or the output unit depending on whether all samples in the improved sample group converge to a previously set reference range.

[0018] A raw material acquisition device according to an embodiment of the present disclosure includes a color prediction model generation unit that generates a regression model indicating color data according to prescription data, and a color prediction calculation unit that outputs a predicted color value through the regression model when a predicted prescription value is input, wherein the prescription data may be color raw material information of a cosmetic product that has already been manufactured, and the color data may be a color value of a color cosmetic product.

[0019] The raw material acquisition device may further include a predicted prescription generation unit that transmits a predicted prescription value corresponding to the target color value to the color prediction calculation unit when a target color value is input, and receives a predicted color value corresponding to the predicted prescription value from the color prediction calculation unit.

[0020] The predicted prescription generating unit may improve the predicted prescription value until a color difference between the predicted color value and the target color value converges within a preset reference range.

[0021] The predicted prescription generating unit may use a genetic algorithm to improve the predicted prescription value. Effect of the Invention

[0022] According to the present disclosure, since color raw materials for producing a cosmetic product having a specific color are output through a color prediction model, there is an advantage that the time, cost, and effort required to produce a cosmetic product of a new color can be minimized.

[0023] According to the present disclosure, the options applied to the raw material formulation can be set according to the needs of the user, which has the advantage that it is possible to formulate the color raw material formulation of a cosmetic product while taking into account practical limitations.

[0024] According to the present disclosure, there is an advantage in that it is possible to manufacture cosmetics in an almost infinite number of colors as desired by users. [Brief description of the drawings]

[0025] [Figure 1] FIG. 2 is a control block diagram of a cosmetic color raw material acquisition device according to an embodiment of the present disclosure. [Diagram 2] FIG. 2 is a control block diagram of a prescription sample improving unit according to an embodiment of the present disclosure. [Diagram 3] 2 is a flowchart showing an operation method of a cosmetic color raw material acquisition device according to an embodiment of the present disclosure. [Figure 4] 1 is a flowchart showing a method for storing a color prediction model by a raw material acquisition device according to an embodiment of the present disclosure. [Diagram 5] 1 is a flowchart showing a method in which a raw material acquisition device according to an embodiment of the present disclosure outputs a predicted recipe according to a target color value. [Figure 6] FIG. 13 is a diagram showing an example in which the raw material acquisition device according to the embodiment of the present disclosure outputs a result recipe according to a target color value. [Figure 7] FIG. 13 is a diagram showing an example in which the raw material acquisition device according to the embodiment of the present disclosure outputs a result recipe according to a target color value. [Figure 8] FIG. 13 is a diagram showing an example in which the raw material acquisition device according to the embodiment of the present disclosure outputs a result recipe according to a target color value. [Figure 9] FIG. 13 is a diagram showing an example in which the raw material acquisition device according to the embodiment of the present disclosure outputs a result recipe according to a target color value. [Figure 10] FIG. 13 is a diagram showing an example in which the raw material acquisition device according to the embodiment of the present disclosure outputs a result recipe according to a target color value. [Figure 11] FIG. 13 is a diagram showing an example in which the raw material acquisition device according to the embodiment of the present disclosure outputs a result recipe according to a target color value. [Figure 12] FIG. 12 is a diagram showing the inner and outer colors of a color cosmetic product using the resultant formulation derived from the examples of FIGS. 6 to 11. [Figure 13] FIG. 12 is a diagram showing the inner and outer colors of a color cosmetic product using the resultant formulation derived from the examples of FIGS. 6 to 11. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0026] Hereinafter, the embodiments disclosed in the present specification will be described in detail with reference to the accompanying drawings, in which the same or similar components will be given the same reference symbols regardless of the drawing symbols, and duplicate descriptions thereof will be omitted.

[0027] The suffixes "module" and "section" for components used in the following description are given or mixed for the sole consideration of ease of writing the specification, and do not have any distinct meanings or roles by themselves.

[0028] In addition, in describing the embodiments disclosed herein, if it is determined that a detailed description of related known technology may obscure the gist of the embodiments disclosed herein, the detailed description will be omitted. In addition, the attached drawings are merely for the purpose of making the embodiments disclosed herein easier to understand, and the technical ideas disclosed herein are not limited by the attached drawings, and should be understood to include all modifications, equivalents, or alternatives included in the ideas and technical scope of the present invention.

[0029] Terms including ordinal numbers such as "second," "first," etc. may be used to describe various components, but the components are not limited by the terms. The terms are used only to distinguish one component from another.

[0030] A singular expression includes a plural expression unless the context clearly indicates otherwise.

[0031] In this application, the terms "comprise" or "have" and the like are intended to specify the presence of features, numbers, steps, operations, components, parts, or combinations thereof set forth in the specification, but are to be understood as not precluding the presence or additional possibility of one or more other features, numbers, steps, operations, components, parts, or combinations thereof.

[0032] FIG. 1 is a control block diagram of a cosmetic color raw material acquisition device according to an embodiment of the present disclosure.

[0033] The cosmetic color tone raw material acquisition device may include at least a part or all of a predictive prescription generation unit 10, a color prediction model generation unit 20, and a color prediction calculation unit 30.

[0034] The predicted prescription generating unit 10 can input a target color value and output a predicted prescription. The predicted prescription generating unit 10 can be configured to output predicted prescription data for deriving a target color value when it receives a target color value and optional items (prescription restrictions) described below as input values.

[0035] The target color value may refer to a color value that is expressed by a user using a color cosmetic product.

[0036] A predictive formula can refer to information about ingredients used in a cosmetic product having a target color value.

[0037] In this specification, raw materials can refer to materials used to develop color in color cosmetics. Raw materials can include pigments.

[0038] The predicted prescription generating unit 10 may include at least some or all of an input unit 110 , a prescription sample group generating unit 120 , a prescription sample improving unit 130 , an improved sample group convergence checking unit 140 , and an output unit 150 .

[0039] A target color value may be inputted to the input unit 110. In addition, the input unit 110 may further input optional items.

[0040] The target color value may be the L*, a*, and b* values ​​of the inner color / outer color. That is, the target color value may be six values ​​in total, including the inner color L*, a*, and b* and the outer color L*, a*, and b*. However, this is merely an example, and the target color value may be in other forms.

[0041] The optional items may refer to restrictions that are reflected when outputting a predicted prescription, i.e., the optional items may be prescription restrictions.

[0042] For example, the options may include the maximum number of ingredients used, the range of the content ratio of each ingredient, the total ingredient content, etc., but are not limited thereto as they are merely illustrative. Here, the ingredient may refer to a pigment. In this specification, the ingredient and the pigment may be used in combination in some cases.

[0043] The input unit 110 may include a target color input unit 112 to which a target color value is input, and an option input unit 114 to which optional items are input.

[0044] The target color value inputted through the target color input unit 112 may be outputted to the prescription sample group generating unit 120 and the prescription sample improving unit 130.

[0045] The prescription sample group generating unit 120 can generate prescription sample groups according to the target color values.

[0046] Specifically, when the target color value is input from the target color input unit 112, the prescription sample group generating unit 120 can generate a prescription sample group according to the input target color value.

[0047] The prescription sample group generating unit 120 can randomly generate prescription sample groups.

[0048] A prescription sample group may include multiple prescription samples. The number of prescription samples in a prescription sample group may be determined according to Equation 1 below.

[0049]

number

[0050] Nsample means the number of formula samples in the formula sample group, Npigment means the number of raw materials, and m can be 5 or more, but this is merely exemplary.

[0051] The prescription sample group generating unit 120 may generate prescription samples within the range of the content ratio of each ingredient when there is an option for the range of the content ratio of each ingredient.

[0052] The prescription sample improving unit 130 may receive prescription sample groups and derive improved sample groups. According to an embodiment, the prescription sample improving unit 130 may derive improved sample groups using a genetic algorithm. When prescription sample groups are received, the prescription sample improving unit 130 may output improved sample groups for the received prescription sample groups.

[0053] The improved sample group may refer to a sample group in which each of the prescription samples in the prescription sample group is improved to be closer to a predicted prescription for a target color value. The improved sample group may be composed of improved samples in which each of the prescription samples in the prescription sample group is adjusted.

[0054] The prescription sample improving unit 130 may output the prescription sample group as an improved sample group using a genetic algorithm. That is, the prescription sample improving unit 130 is configured to apply a genetic algorithm, and when the prescription sample group is received as an input value, it may transmit the prescription sample group to the color prediction calculation unit 30 to compare the predicted color value, and then perform the function of evolving the prescription sample group. The prescription sample improving unit 130 may perform this process until all the prescription sample groups are similar, and may ultimately converge all the prescription sample groups to a prescription that can derive a predicted color closest to the target color value. In this case, it may be desirable to further include a process of giving a penalty to the optional items according to the present invention to converge the prescription sample group. For example, the prescription sample improving unit 130 may give a penalty according to the optional items when deriving a predicted color closest to the target color value for all the prescription sample groups, thereby converging the prescription sample group.

[0055] When the target color value is input, the predicted prescription generating unit 10 transmits a predicted prescription value corresponding to the target color value to the color prediction calculating unit 30, and may receive a predicted color value corresponding to the predicted prescription value from the color prediction calculating unit 30. The predicted prescription generating unit 10 may improve the predicted prescription value until the color difference between the predicted color value and the target color value converges within a previously set reference range. The predicted prescription generating unit 10 may improve the predicted prescription value using a genetic algorithm.

[0056] The prescription sample improving unit 130 can output raw material information for manufacturing a color cosmetic having a target color value. In addition, the prescription sample improving unit 130 can output raw material information reflecting optional items, if any.

[0057] The improved sample group convergence confirmation unit 140 can obtain whether the improved sample group has converged. The improved sample group convergence confirmation unit 140 can obtain whether the improved sample group has converged by determining whether the improved samples in the improved sample group have converged to a previously set reference range. The improved sample group convergence confirmation unit 140 can output the improved sample group to the prescription sample improvement unit 130 or the output unit 150 depending on whether all samples in the improved sample group have converged to a previously set reference range.

[0058] According to one embodiment, the improved sample group convergence confirmation unit 140 is configured to check the predicted prescription amount {A n.0.1 ,A n.1.1 ,…,A n.m.1} has a variance of m a If we assume that all the raw materials have a variance m a , m b , m c , ... is less than the previously set reference value, it can be determined that the improved samples in the improved sample group have converged to the previously set reference range. Here, the previously set reference value may be less than 0.1, but this is merely an example and is not limited thereto.

[0059] If the improvement sample group convergence confirmation unit 140 determines that the improvement sample group has converged to the already set reference range, it can determine that the improvement sample group has converged, and if it determines that the improvement sample group has not converged to the already set reference range, it can determine that the improvement sample group has not converged.

[0060] If the improved sample group does not converge to a previously set reference range, the improved sample group convergence confirmation unit 140 can output the improved sample group as a prescription sample group to the prescription sample improvement unit 130. That is, the improved sample group determined not to converge by the improved sample group convergence confirmation unit 140 can be fed back to the prescription sample improvement unit 130 as a prescription sample group.

[0061] Thus, the prescription sample improving unit 130 can output an improved sample group for the input improved sample group. The prescription sample improving unit 130 can recognize the input improved sample group as a prescription sample group. In this manner, the prescription sample improving unit 130 can output an Nth order sample group as an N+1th order sample group.

[0062] The improved sample group convergence checking unit 140 may transmit the improved sample group to the output unit 150 if the improved sample group converges to a preset reference range.

[0063] The output unit 150 may output raw material information for manufacturing a color-tone cosmetic having a target color value based on the improved sample group output from the prescription sample improvement unit, where the raw material information may include a content ratio of each raw material. The output unit 150 may output the improved sample group input from the improved sample group convergence confirmation unit 140 as a predicted prescription. The output unit 150 may output raw material information for manufacturing a color-tone cosmetic having a target color value based on the improved sample group input from the improved sample group convergence confirmation unit 140. That is, the predicted prescription (result prescription) may include raw material information for manufacturing a color-tone cosmetic having a target color value. The output unit 150 may include a predicted prescription derivation unit 152 that outputs the predicted prescription.

[0064] Meanwhile, the prescription sample improving unit 130 can derive the prescription sample group as an improved sample group through the color prediction calculation unit 30.

[0065] The color prediction calculation unit 30 can calculate the predicted color of an existing sample and the predicted color of a combination sample through a regression model.

[0066] Specifically, the prescription sample improving unit 130 can transmit the predicted prescription value according to the input prescription sample group to the color prediction calculation unit 30. The color prediction calculation unit 30 can output the predicted color value derived from the predicted prescription value input from the prescription sample improving unit 130 to the prescription sample improving unit 130. The prescription sample improving unit 130 can output the improved sample group by referring to the predicted color value input from the color prediction calculation unit 30.

[0067] The color prediction calculation unit 30 may be configured to calculate an actual color prediction using the color prediction model generated by the color prediction model generation unit 20. The color prediction calculation unit 30 can output predicted color values ​​(external color value, internal color value) when receiving any raw material recipe data as an input value.

[0068] The color prediction calculation unit 30 can include at least a part or all of a prescription input unit 310 , a color prediction model management unit 320 , and a predicted color deriving unit 330 .

[0069] The prescription input unit 310 can receive predicted prescription values ​​from the predicted prescription generating unit 10, particularly the prescription sample improving unit 130.

[0070] The prescription input unit 310 can output the input predicted prescription value to the color development prediction model management unit 320.

[0071] The color prediction model management unit 320 may store a color prediction model. The color prediction model may include a color regression model based on a prescription.

[0072] The color prediction model management unit 320 can input the predicted prescription value to the color prediction model and output the output of the color prediction model to the predicted color deriving unit 330. The predicted color deriving unit 330 can transmit the output input from the color prediction model management unit 320 as a predicted color value to the predicted prescription generating unit 10, particularly the prescription sample improving unit 130.

[0073] In this manner, the prescription sample improvement unit 130 obtains the predicted color value for the predicted prescription value through the color prediction model management unit 320, and by referring to this, can improve the prescription samples of the prescription sample group and output the improved sample group.

[0074] Meanwhile, the regression model stored in the color prediction model management unit 320 may be input from the color prediction model generation unit 20 .

[0075] The color prediction model generating unit 20 can generate a regression model that indicates color data according to the prescription data. Here, the prescription data is color raw material information of a cosmetic product that has already been manufactured, and the color data includes color values ​​of the cosmetic product (e.g., L*, a*, b* of each of the inner and outer colors).

[0076] The color development prediction model generating unit 20 may include at least a part or the whole of a database 210 and a regression model generating unit 220 .

[0077] The database 210 can store color development data corresponding to the prescription data. The prescription data includes the content ratio of each ingredient, and the color development data can include the L*, a*, and b* of each of the developed inner and outer colors. That is, the database 210 can include the L*, a*, and b* values ​​of each of the inner and outer colors according to the content ratio of each ingredient.

[0078] The formula data stored in the database 210 may include the type of base formula and the composition of color-developing raw materials. Here, the base formula refers to the composition of raw materials excluding color-developing raw materials involved in color development among all formulas for producing the final color cosmetic, and a regression model may be generated for each type of base formula through the regression model generating unit 220. This may be necessary to minimize the influence of the base formula on the predicted formula obtained through the cosmetic color-developing raw material acquisition device according to the present invention. In this case, the cosmetic color-developing raw material acquisition device according to the present invention may select the type of base formula along with the target color development value as an input value, and the predicted formula may include the type of base formula or the composition of the base formula and the composition of the predicted color development raw materials as a result value.

[0079] The regression model generating unit 220 can generate a regression model that indicates color development according to a prescription. That is, the regression model generating unit 220 can generate a regression model using color development data according to prescription data stored in the database 210. The regression model can be a multiple regression model. The regression model can be a linear or nonlinear method for predicting a result vector (Y) having a trend from multiple vectors (X).

[0080] The regression model generating unit 220 can transmit the generated multiple regression model to the color prediction calculating unit 30, particularly to the color prediction model managing unit 320.

[0081] In summary, the color prediction model generating unit 20 may be configured to generate a color prediction model based on color value experimental data from raw material recipes. The color prediction model may be a model that takes any raw material recipe data as input values ​​and outputs predicted color values ​​(external color value, internal color value). The greater the number of experimental data and the wider and more detailed the color gamut it encompasses, the more accurate the color prediction model the color prediction model generating unit 20 can generate.

[0082] Next, the operation method of the prescription sample improving unit 130 will be described in more detail with reference to Fig. 2. The operation method of the prescription sample improving unit 130 described below may be an operation method utilizing a genetic algorithm.

[0083] FIG. 2 is a control block diagram of a prescription sample improving unit according to an embodiment of the present disclosure.

[0084] The prescription sample improvement unit 130 may include at least a part or all of a sample candidate derivation unit 1310, a sample evaluation unit 1320, and a sample selection unit 1330.

[0085] The sample candidate derivation unit 1310 may output a sample candidate that may be an improved sample. The sample candidate derivation unit 1310 may output a combination sample based on the input existing samples. The sample candidate derivation unit 1310 may generate a combination sample by selecting one of the existing samples and one of the remaining existing samples as a random sample, mutating the random sample, and combining it with one of the existing samples. A method of generating a combination sample will be described in detail below.

[0086] The sample candidate derivation unit 1310 may generate a random sample, generate a mutant sample by performing an interpolation operation between an existing sample and the random sample, and derive a combination sample by probabilistically selecting and combining the existing sample and the mutant sample. To this end, the sample candidate derivation unit 1310 may include a random sample selection unit 1311 that randomly selects one of the existing samples, a mutant sample derivation unit 1313 that generates a mutant sample, and a combination sample derivation unit 1315 that generates a combination sample. Here, the existing sample may be a prescription sample input to the prescription sample improvement unit 130.

[0087] Meanwhile, the generated combination sample may be output together with the existing sample to the color prediction calculation unit 30. The color prediction calculation unit 30 may calculate predicted colors based on the existing samples and predicted colors based on the combination samples, and output the calculated colors to the sample evaluation unit 1320.

[0088] The sample evaluation unit 1320 may calculate a first color difference between a predicted color according to an existing sample and a target color value and a second color difference between a predicted color according to a combination sample and a target color value. If there is an optional item, the sample evaluation unit 1320 may compare the first color difference reflecting a penalty due to the optional item with the second color difference reflecting a penalty due to the optional item.

[0089] The sample evaluation unit 1320 can compare the predicted color development of the existing sample with the predicted color development of the combined sample.

[0090] The sample evaluation unit 1320 may include a color difference calculation unit 1321 and a color difference comparison unit 1325. The color difference calculation unit 1321 receives a target color value, a predicted color based on an existing sample, and a predicted color based on a combination sample, and calculates a first color difference between the target color value and the predicted color based on the existing sample and a second color difference between the target color value and the predicted color based on the combination sample.

[0091] The color difference comparator 1325 can compare the first color difference with the second color difference. 既存 ) is the color difference between the target color value and the predicted color by the existing sample, and the second color difference ( 組合せ ) can refer to the color difference between the target color value and the predicted color from the combination sample.

[0092] The color difference comparator 1325 can output the comparison result between the first color difference and the second color difference to the sample selector 1330 .

[0093] The sample selection unit 1330 selects a sample having a smaller value among the first chrominance and the second chrominance. Improved samples within the improved sample group It can be output as:

[0094] The sample selection unit 1330 may output the existing sample as an improved sample if the first color difference is smaller than the second color difference, and may output the combined sample as an improved sample if the first color difference is larger than the second color difference.

[0095] That is, the sample selection unit 1330 can output a sample having a small color difference from the target color value as an improved sample.

[0096] Meanwhile, the sample evaluator 1320 may further include a penalty calculator 1323 .

[0097] The penalty calculation unit 1323 may receive the existing sample, the combined sample, and an option item as input.

[0098] The penalty calculation unit 1323 may calculate a first penalty-reflecting color difference and a second penalty-reflecting color difference if there is an option. 既存 + Penalty 既存 ) is the sum of the first color difference and the first penalty due to the existing sample, and the second penalty reflected color difference ( 組合せ + Penalty 組合せ ) may be the sum of the second color difference and the second penalty due to the combined sample.

[0099] In this case, the color difference comparator 1325 can compare the first penalty-reflecting color difference with the second penalty-reflecting color difference, and output the comparison result to the sample selector 1330 .

[0100] The sample selection unit 1330 can output the existing sample as an improved sample if the first penalty-reflecting color difference is smaller than the second penalty-reflecting color difference, and can output the combined sample as an improved sample if the first penalty-reflecting color difference is larger than the second penalty-reflecting color difference.

[0101] In this manner, the improved sample derived by the prescription sample improving unit 130 may be input again to the prescription sample improving unit 130 through the improved sample group convergence checking unit 140 or may be input to the output unit 150.

[0102] Through the above-mentioned method, the prediction prescription generating unit 10 creates a function for calculating color difference and a function for calculating a penalty for an optional item, and applies a genetic algorithm to induce the result values ​​of the two functions to converge to 0, thereby deriving an optimal prediction prescription.

[0103] FIG. 3 is a flowchart showing an operation method of the cosmetic color raw material acquisition device according to an embodiment of the present disclosure.

[0104] Referring to FIG. 3, the raw material acquisition device can store a color prediction model (S10).

[0105] Next, a method for the raw material recovery device to store the color prediction model will be described in more detail with reference to Fig. 4. That is, step S10 in Fig. 3 will be described in detail with reference to Fig. 4.

[0106] FIG. 4 is a flowchart showing a method for storing a color prediction model by a raw material acquisition device according to an embodiment of the present disclosure.

[0107] The color prediction model generating unit 20 can obtain color data according to a prescription (S10).

[0108] The color data according to the prescription may include color values ​​according to the ingredient content ratio. Specifically, the color data according to the prescription may include a first color value according to a first ingredient content ratio, a second color value according to a second ingredient content ratio, ..., an nth color value according to an nth ingredient content ratio. The color values ​​may refer to the L*, a*, and b* values ​​of the inner and outer colors of the cosmetic product.

[0109] The internal color may refer to the color that is developed by using the cosmetic product, and the external color may refer to the color of the cosmetic product itself before use.

[0110] The color development prediction model generating unit 20 can collect and additionally update a large amount of color development data according to prescriptions.

[0111] The color prediction model generating unit 20 can generate a color prediction model according to the prescription based on the acquired color data according to the prescription (S20).

[0112] The color prediction model generating unit 20 can generate a color prediction model by prescription using the collected color data by prescription. The color prediction model by prescription can be a DNN (Deep Neural Network) model. The DNN model can be a model that outputs a predicted color value when the content ratio of each raw material is input.

[0113] The color prediction model generating unit 20 can store the color prediction model based on the generated prescription (S30).

[0114] Referring again to FIG.

[0115] The raw material acquisition device may receive a target color value as input (S20).

[0116] For example, the raw material acquisition device may receive input of inner colors L*, a*, b* and outer colors L*, a*, b* as target color values.

[0117] Meanwhile, the raw material acquisition device may further input optional items together with the target color value.

[0118] For example, the options may include at least one of a maximum number of ingredients used in the recipe, a content ratio range for each ingredient, and a total ingredient content ratio.

[0119] The raw material acquisition device can output a predicted recipe according to the input target color value (S30).

[0120] Next, a method in which the raw material acquisition device outputs a predicted recipe according to a target color value will be described with reference to Fig. 5. Fig. 5 is a flow chart embodying step S30 in Fig. 3.

[0121] FIG. 5 is a flowchart showing a method in which a raw material acquisition device according to an embodiment of the present disclosure outputs a predicted recipe according to a target color value.

[0122] The prescription sample group generating unit 120 can generate a prescription sample group (S301).

[0123] The prescription sample group generating unit 120 can generate a prescription sample group when the target color developing value is input.

[0124] The formulation sample groups are G0, G1, G2, G3, …, G n , G0 is the initial prescription sample group, G1 is the first-improvement sample group (first-generation improvement sample group), G2 is the second-improvement sample group (second-generation improvement sample group), G3 is the third-improvement sample group (third-generation improvement sample group), ..., G n may be an nth-order improved sample group (nth generation improved sample group).

[0125] Initial prescription sample group G0 and improved sample group G n can be as follows:

[0126] G0={S 0.0 ,S 0.1 ,S 0.2,…,S 0.m}, G n ={S n.0 ,S n.1 ,S n.2 ,…,S n.m}

[0127] Here, S 0.0 , S 0.1 , S 0.2 , … , S 0.m Each of S can represent one prescription sample. n.m indicates the mth sample of the nth generation.

[0128] And S 0.0 ={A 0.0.1 ,A 0.0.2 ,A 0.0.3 , …}, S 0.1 ={A 0.1.1 ,A 0.1.2 ,A 0.1.3 ,…},…,S n.m {A n.m.1 ,A n.m.2 ,A n.m.3 ,…,A n.m.l}, where each element constituting each recipe sample indicates the content of the raw material, and the number of elements may be the same as the number of raw materials to be formulated. That is, A n.m.l can mean the content of the lth dye in the mth sample of the nth generation.

[0129] The prescription sample improving unit 130 can generate a variant sample based on each prescription sample (S303), and generate a combination sample by combining the existing sample and the variant sample (S305).

[0130] Here, the combination sample may indicate a combination sample generated by the prescription sample improving unit 130 based on the input prescription sample.

[0131] If the prescription sample improving unit 130 receives the initial prescription sample group G0, it first selects the existing sample S 0.0 In contrast, other existing samples S 0.1 , S0.2 , … , S 0.m For example, the prescription sample improving unit 130 may select one of the other existing samples S 0.2 Assume that you have selected

[0132] The prescription sample improvement unit 130 is a two prescription sample S 0.0 and S 0.2 For example, we can perform an interpolation on S 0.0 = {0,1,3,5}, and S 0.2 In the case of {1, 4, 6, 8}, the prescription sample improving unit 130 selects one of the values ​​between 0 and 1 as the value of the first element, selects one of the values ​​between 1 and 4 as the value of the second element, selects one of the values ​​between 3 and 6 as the value of the third element, and selects one of the values ​​between 5 and 8 as the value of the fourth element. As a result, S 0.0変異 That is, the prescription sample improving unit 130 can generate a mutant sample such as the existing sample S 0.0 The mutant sample for S 0.0変異 ={A 0.0.1 ~A 0.2.1 Any value between ,A 0.0.2 ~A 0.2.2 Any value between ,A 0.0.3 ~A 0.2.3 It can be generated with any value between {,…}.

[0133] Then, the prescription sample improving unit 130 selects the existing sample S 0.0 ={A 0.0.1 ,A 0.0.2 ,A 0.0.3 ,…} and the mutated sample S 0.0変異 ={A 0.0.1変異 ,A 0.0.2変異 ,A 0.0.3変異 ,…}, where S is a randomly selected value. 0.0組合せ ={A 0.0.1 and A 0.0.1変異 Any one of the following: A 0.0.2 and A 0.0.2変異 Any one of the following: A 0.0.3and A 0.0.3変異 Any one of ... through S 0.0組合せ ={A 0.0.1組合せ ,A 0.0.2組合せ ,A 0.0.3組合せ , …} can be generated. 0.0 ={0,1,3,5} and S 0.0変異 = {0.5,2,5,7}, S 0.0組合せ = {0.5, 2, 3, 5}, but this is merely an example and is not limited to this.

[0134] On the other hand, S 0.0 The method for generating a combination sample for S has been described as an example. 0.1 , S 0.2 , … , S 0.m For each one, the methods described above can be applied to generate a combination sample.

[0135] In summary, the prescription sample improving unit 130 is n.m.r If is a random sample of the mth sample of n generations, then S n.m.r is G n Among the samples constituting the sample group, S n.m Select one of the other samples that is not S n.m.r ={A n.m.1.r ,A n.m.2.r ,A n.m.3.r ,…,A n.m.l.r}, where A n.m.1.r can mean the lth ingredient content of the mth random sample of the nth generation. n.m.変異 is the mutant sample of the mth sample in the nth generation, and S n.m.変異 ={A n.m.1.変異 ,A n.m.2.変異 ,A n.m.3.変異 ,…,A n.m.l.変異} and A n.m.l変異 A n.m.l and A n.m.l.r It can be any value between

[0136] S n.m.組合せ If is a combination sample of the mth sample in the nth generation, then Sn.m.組合せ ={A n.m.1.組合せ ,A n.m.2.組合せ ,A n.m.3.組合せ ,…,A n.m.l.組合せ} and A n.m.1.組合せ A n.m.l and A n.m.l.変異 Each ingredient is one of the following: n.m.l and A n.m.1.変異 can be arbitrarily selected from among

[0137] Then, the prescription sample improving unit 130 can input the existing sample and the combination sample into the color prediction model and calculate each color difference (S307). If there is an option, the prescription sample improving unit 130 can calculate a penalty-reflected color difference for each of the existing sample and the combination sample (S309).

[0138] The color difference can be calculated using Equation 2 below, and the penalty can be calculated using Equation 3 below.

[0139]

number

[0140]

number

[0141] In formula 2, L 外色目標 is the external color L* received at the target input value, L 外色計算 is the outer color L* of the predicted color value calculated by the color prediction calculation unit 30, 外色目標 is the external color a* received as the target input value, a 外色計算 are the outer colors a*, b of the predicted color values ​​calculated by the color prediction calculation unit 30. 外色目標 is the external color b* received as the target input value, b 外色計算 is the outer color b* of the predicted color value calculated by the color prediction calculation unit 30, L 内色目標 is the inner color L* received as the target input value, L inner color calculation is the inner color L* of the predicted color value calculated through the color prediction calculation unit 30, a 内色目標is the inner color a* received as the target input value, a 内色計算 are the colors a*, b of the predicted color values ​​calculated by the color prediction calculation unit 30. 内色目標 is the inner color b* received as the target input value, b 内色算出 may represent the inner color b* of the predicted color value calculated by the color prediction calculation unit 30.

[0142] The prescription sample improving unit 130 calculates the color difference (color difference 組合せ ) and color difference against existing samples (color difference 既存 ) can be compared. If there is an option, the prescription sample improving unit 130 compares the penalty reflection color difference (color difference 組合せ + Penalty 組合せ ) and the penalty reflection color difference for existing samples (color difference 既存 + Penalty 既存 ) can be compared.

[0143] The following is the color difference for the combination sample (color difference 組合せ ) is the penalty reflection color difference (color difference 組合せ + Penalty 組合せ ) and the color difference (color difference 既存 ) is the penalty reflection color difference (color difference 既存 + Penalty 既存 ).

[0144] The prescription sample improving unit 130 may obtain whether the color difference for the combination sample is smaller than the color difference for the existing sample (S311).

[0145] If the color difference for the combination sample is smaller than the color difference for the existing sample, the prescription sample improving unit 130 can replace the existing sample with the combination sample (S313).

[0146] As a result, the prescription sample improving unit 130 can output an improved sample group G1 composed of combination samples, which is an output for the input initial prescription sample group G0.

[0147] If the color difference for the combination sample is greater than the color difference for the existing sample, the prescription sample improving unit 130 may maintain the existing sample (S315).

[0148] As a result, the prescription sample improving unit 130 can output an improved sample group G1 composed of existing samples, which is an output for the input initial prescription sample group G0.

[0149] The improved sample group convergence confirmation unit 140 may obtain whether all samples of the improved sample group output from the prescription sample improvement unit 130 converge within a reference range (S317).

[0150] Hereinafter, the improvement sample group convergence confirmation unit 140 receives the improvement sample group G n The following explanation will be given assuming that the following is input. n ={S n.0 ,S n.1 ,S n.2 ,…,S n.m}Improvement sample S n.0 , S n.1 , S n.2 , … , S n.m The content of the first ingredient {A n.0.1 ,A n.1.1 ,…,A n.m.1} variance m a , the content of the second ingredient {A n.0.2 ,A n.1.2 ,…,A n.m.2} variance m b , the content of the third ingredient {A n.0.3 ,A n.1.3 ,…,A n.m.3} variance m c , ..., if the variances ma, mb, mc, ... for all the raw materials are less than the preset reference value, it can be determined that the improved samples in the improved sample group have converged to the preset reference range. Here, the preset reference value may be less than 0.1, but this is merely an example and is not limited thereto.

[0151] If the improved sample group does not converge to the already set reference range, the improved sample group convergence confirmation unit 140 can input the improved sample group to the prescription sample improvement unit 130. As a result, the prescription sample improvement unit 130 can recognize the improved samples of the input improved sample group as prescription samples and generate mutant samples based on each prescription sample.

[0152] In this way, the prescription sample improving unit 130 can recognize the input sample group as a prescription sample group, improve the recognized prescription sample group, and output it as an improved sample group. Thus, when the first prescription sample group G0 is input, the prescription sample improving unit 130 outputs a first improved sample group G1, when the first improved sample group G1 is input, it outputs a second improved sample group G2, ..., an nth improved sample group G n can be output.

[0153] First-order improvement sample group G1, ..., nth-order improvement sample group G n can be as follows:

[0154] G1={S 1.0 ,S 1.1 ,S 1.2 ,…,S 1.m}, G n ={S n.0 ,S n.1 ,S n.2 ,…,S n.m}, S 1.0 ,S 1.1 ,S 1.2 ,…,S 1.m ,…,S n.0 ,S n.1 ,S n.2 ,…,S n.m Each of the formulas can represent one sample. 1.0 ={A 1.0.1 ,A 1.0.2 ,A 1.0.3 , …}, S 1.1 ={A 1.1.1 ,A 1.1.2 ,A 1.1.3 ,…},…,Sn.m ={A n.m.1 ,A n.m.2 ,A n.m.3 , ...}, where each element constituting each formulation sample indicates the content of a raw material, and the number of elements may be the same as the number of raw materials to be formulated.

[0155] When all the samples in the improved sample group converge within the reference range, the improved sample group convergence confirmation unit 140 can derive a result prescription (S319).

[0156] Specifically, when all samples in the improved sample group converge within the reference range, the improved sample group convergence confirmation unit 140 outputs the improved sample group to the output unit 150, and the output unit 150 can output the improved samples of the improved sample group as a result recipe.

[0157] 6 to 11 are diagrams showing examples of result recipes output by the raw material acquisition device according to the embodiment of the present disclosure in accordance with target color values.

[0158] More specifically, Figures 6 to 11 are examples showing the experimental results of ensuring the performance of the raw material acquisition device using actual commercially available cosmetics whose color tones and pigment types are unknown. Figures 6 to 11 are diagrams showing the result formula and the color according to the result formula that are output when the outer color and inner color of each cosmetic product actually available on the market are input as target color values ​​to the raw material acquisition device. In the experiment of Figures 6 to 11, the options were set to a total number of raw materials of 6 or less, the range of the content ratio of each pigment to a content ratio of black pigment of 0.5% or less, and the total content ratio of the pigments used to 20%.

[0159] Figure 6 shows an example of a case where the raw material acquisition device inputs 38.13, 44.30, 26.05, 42.39, 53.77, and 32.59 as the target color values ​​for the outer color L*, a*, and b* and the inner color L*, a*, and b*, and predictive prescriptions are derived with the contents of raw materials P1, P3, P4, P8, and P9 being 5.66, 3.58, 3.67, 0.69, and 6.41, respectively. In this case, it can be confirmed that the target colors according to the outer color values ​​and inner color values ​​are very similar to the predicted colors according to the actual prescription.

[0160] Figure 7 shows an example of a case where the raw material acquisition device inputs 32.95, 37.85, 21.00, 34.41, 38.02, and 19.89 as the target color values ​​for the outer color L*, a*, and b* and the inner color L*, a*, and b*, and a predicted recipe is derived with the contents of raw materials P1, P3, P4, P11, P12, and P13 being 6.08, 1.77, 11.0, 0.54, 0.54, and 0.03, respectively. In this case, it can be confirmed that the target color according to the outer color value and the inner color value is very similar to the predicted color according to the actual recipe.

[0161] Figure 8 shows an example of a case where the raw material acquisition device inputs 28.59, 20.12, 9.57, 28.77, 19.27, and 8.74 as the target color values ​​for the outer color L*, a*, and b* and the inner color L*, a*, and b*, and a predicted recipe is derived with the contents of raw materials P1, P4, P5, P11, and P12 being 0.66, 9.65, 5.99, 1.54, and 2.14, respectively. In this case, it can be confirmed that the target color according to the outer color value and the inner color value is very similar to the predicted color according to the actual recipe.

[0162] Figure 9 shows an example of a case where the raw material acquisition device inputs 46.93, 53.07, 39.18, 48.43, 56.60, and 40.95 as the target color values ​​for the outer color L*, a*, and b* and the inner color L*, a*, and b*, and a predicted recipe with the contents of raw materials P1, P5, and P12 of 10, 1.45, and 3.53, respectively, is derived. In this case, it can be confirmed that the target color according to the outer color value and the inner color value is very similar to the predicted color according to the actual recipe.

[0163] Figure 10 shows an example of a case where the raw material acquisition device inputs 41.85, 51.91, 21.86, 44.43, 53.77, and 22.59 as the target color values ​​for the outer color L*, a*, b* and the inner color L*, a*, b*, and derives a predicted recipe with the contents of raw materials P1, P3, P4, P9, and P12 being 1.72, 3.38, 3.0, 5.7, and 6.19, respectively. In this case, it can be confirmed that the target color according to the outer color value and the inner color value is very similar to the predicted color according to the actual recipe.

[0164] Figure 11 shows an example of a case where the raw material acquisition device inputs 49.32, 16.54, 17.32, 49.55, 16.71, and 17.11 as the target color values ​​for the outer color L*, a*, b* and the inner color L*, a*, b*, and derives a predicted recipe with the contents of raw materials P1, P7, P11, P12, and P13 being 2.7, 5.24, 1.24, 10.55, and 0.42, respectively. In this case, it can be confirmed that the target color according to the outer color value and the inner color value is very similar to the predicted color according to the actual recipe.

[0165] 12 and 13 are diagrams showing the inner and outer colors of a color cosmetic product using the resultant formulation derived from the examples of FIGS. 6 to 11. In FIG.

[0166] Specifically, the upper side 1210 of FIG. 12 shows the outer color of the cosmetic produced according to the resultant formula, and the lower side 1220 of FIG. 12 shows the outer color of the target color cosmetic to be realized through the raw material acquisition device.

[0167] The upper side 1310 of FIG. 13 shows the inner color of the cosmetic produced according to the resultant formula, and the lower side 1320 of FIG. 13 shows the inner color of the target color cosmetic to be realized through the raw material acquisition device.

[0168] More specifically, an average error of 4.6% was measured for the inner color and an average error of 6.4% for the outer color, confirming that the raw material acquisition device can more accurately derive raw materials for manufacturing cosmetics capable of developing the desired color.

[0169] The present invention described above can be embodied as computer readable codes on a medium having a program recorded thereon. The computer readable medium includes all kinds of recording devices in which data that can be read by a computer system is stored. Examples of the computer readable medium include a hard disk drive (HDD), a solid state disk (SSD), a silicon disk drive (SDD), a ROM, a RAM, a CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device. The computer may also include a control unit 11 of the terminal 1 or a control unit 31 of the cosmetics manufacturing device 3. Therefore, the above detailed description should not be construed as limiting in all respects, but should be considered as illustrative. The scope of the present invention should be determined by a reasonable interpretation of the appended claims, and all modifications within the scope of the present invention that are equivalent to the scope of the present invention are included in the scope of the present invention.

[0170] The above description is merely an illustrative example of the technical concept of the present invention, and various modifications and variations may be made by a person having ordinary knowledge in the technical field to which the present invention pertains without departing from the essential characteristics of the present invention.

[0171] Therefore, the embodiments disclosed herein are intended to illustrate, rather than limit, the technical concept of the present invention, and the scope of the technical concept of the present invention is not limited by such embodiments.

[0172] The scope of protection of the present invention should be interpreted by the following claims, and all technical ideas within the scope equivalent thereto should be interpreted as being included in the scope of the present invention. [Explanation of symbols]

[0173] 10. Predictive prescription generation unit 110 Input section 112 Target color input section 114 Optional input section 120 Prescription sample group generation unit 130 Prescription Sample Improvement Department 1310 Sample Candidate Derivation Unit 1311 Random Sample Selection Section 1313 Mutation sample extraction section 1315 Combination sample extraction section 1320 Sample Evaluation Department 1321 Color difference calculation section 1323 Penalty Calculation Unit 1325 Color difference comparison section 1330 Sample Selection Section 140 Improved sample group convergence confirmation section 150 Output section 152 Predictive prescription derivation part 20 Color prediction model generation unit 210 Color Database by Prescription 220 Regression model generation unit 30 Color prediction calculation section 310 Prescription input unit 320 Color Prediction Model Management Department 330 Predicted color generation derivation part

Claims

1. an input section for inputting a target color value; A prescription sample group generating unit that generates a prescription sample group according to the target color value; A prescription sample improving unit that outputs the prescription sample group as an improved sample group using a genetic algorithm; and an improved sample group convergence confirmation unit that outputs the improved sample group to the prescription sample improvement unit or the output unit depending on whether all samples in the improved sample group converge to a previously set reference range, the output unit outputs raw material information for manufacturing a color cosmetic having the target color value based on the improved sample group output from the prescription sample improving unit; The ingredient information includes a content of at least one ingredient, The prescription sample improvement unit includes: a sample candidate derivation unit that selects one of the existing samples and one of the remaining existing samples as a random sample, performs an interpolation operation to select a value between a content amount included in the one existing sample and a content amount included in the random sample, mutates the random sample into a mutant sample, and then generates a combination sample by combining the one existing sample and the mutant sample; a sample evaluation unit that calculates a first color difference between a predicted color of the one existing sample and the target color value and a second color difference between a predicted color of the combination sample and the target color value; a sample selection unit that outputs a sample having a smaller value of the first color difference and the second color difference as an improved sample in the improved sample group.

2. The input unit further receives an option item, The raw material acquisition device according to claim 1 , wherein the prescription sample improvement unit outputs the raw material information by reflecting the optional items.

3. The raw material acquisition device according to claim 2 , wherein the options include at least one of a maximum number of raw materials to be used, a range of content ratios for each raw material, and a total raw material content.

4. The prescription sample improvement unit includes:

2. The raw material obtaining device according to claim 1, further comprising: a step of comparing a first color difference reflecting a penalty due to the optional item with a second color difference reflecting a penalty due to the optional item when the optional item is present.

5. The improvement sample group convergence confirmation unit The raw material acquisition device of claim 1, wherein when the variance of the content of each raw material in the improved sample group is less than a preset reference value, it is acquired that all samples in the improved sample group converge to a preset reference range.

6. The improvement sample group convergence confirmation unit The raw material acquisition device according to claim 5 , wherein when it is confirmed that all samples in an improved sample group converge to a preset reference range, the improved sample group is output to the output unit.

7. The improvement sample group convergence confirmation unit The raw material acquisition device according to claim 5, wherein when it is determined that all samples in the improved sample group do not converge to a pre-set reference range, the improved sample group is output to the prescription sample improvement unit.

8. The prescription sample improvement unit includes:

8. The raw material obtaining device according to claim 7, further comprising: a second improved sample set output section for the improved sample set input from the improved sample set convergence confirmation section.

9. The prescription sample improvement unit includes: The raw material acquisition device according to claim 1 , wherein the combined sample is generated by randomly selecting one value of the existing sample or the mutant sample.

10. The raw material obtaining device according to claim 1 , further comprising a color prediction calculation unit for calculating a predicted color of the existing sample and a predicted color of the combined sample through a regression model.

11. A color prediction model generating unit that generates a regression model showing color data according to prescription data; and a color prediction calculation unit that outputs a predicted color value indicating a predicted color calculated through the regression model when a predicted prescription value is input, The prescription data is information on color raw materials of a cosmetic product that has already been manufactured, The raw material acquisition device according to claim 10 , wherein the color development data is a color value of a color cosmetic.

12. The raw material acquisition device according to claim 11, further comprising a predicted prescription generating unit that, when a target color value is input, transmits a predicted prescription value corresponding to the target color value to the color prediction calculation unit, and receives a predicted color value corresponding to the predicted prescription value from the color prediction calculation unit.

13. The predictive prescription generating unit includes: The raw material acquisition device according to claim 12, further comprising: improving the predicted prescription value until a color difference between the predicted color value and the target color value converges within a preset reference range.

14. Any one of the existing samples and the random sample includes information related to at least one ingredient and an ingredient content, The sample candidate derivation unit The raw material acquisition device according to claim 13, wherein for at least one of the contents of the raw materials contained in the one existing sample and the random sample, one of the values ​​between the content of the raw material contained in the one existing sample corresponding to the same raw material and the content of the raw material contained in the random sample is selected to perform the interpolation calculation.

15. A step in which an input unit inputs a target color value; A prescription sample group generating unit generates a prescription sample group according to a target color value; A step in which a prescription sample improvement unit outputs the prescription sample group as an improved sample group using a genetic algorithm; An improved sample group convergence confirmation unit acquires whether all samples in the improved sample group converge to a previously set reference range; The formula sample improvement unit outputs ingredient information according to whether all samples in the improved sample group converge to a previously set reference range, or outputs a secondary improved sample group for the improved sample group, The step of outputting the prescription sample group as an improved sample group includes: a sample candidate derivation unit selecting one of the existing samples and one of the remaining existing samples as a random sample, performing an interpolation operation to select a value between a content amount contained in the one existing sample and a content amount contained in the random sample, mutating the random sample into a mutant sample, and then combining one of the existing samples with the mutant sample to generate a combined sample; A sample evaluation unit calculates a first color difference between a predicted color of the existing sample and a target color value and a second color difference between a predicted color of the combined sample and a target color value; a sample selection unit outputting a sample having a smaller value of the first color difference and the second color difference as an improvement sample in the improvement sample group; The step of outputting the raw material information includes: The raw material acquisition method includes a step in which the prescription sample improvement unit outputs raw material information for manufacturing a color cosmetic having the target color value based on the improved sample group output in the step of outputting the prescription sample group as an improved sample group.

16. The input unit further includes a step of inputting an option item, The step of outputting the raw material information includes:

16. The raw material acquisition method according to claim 15, further comprising a step of: the recipe sample improvement unit outputting the raw material information by reflecting the optional items.

17. The step of inputting the optional items includes:

17. The method for obtaining raw materials according to claim 16, wherein the input unit is a step for inputting at least one of a maximum number of raw materials to be used, a range of content ratios for each raw material, and a total raw material content.

18. The step of outputting the prescription sample group as an improved sample group includes: The method for obtaining raw materials according to claim 15, further comprising a step of comparing a first color difference reflecting a penalty due to the optional item with a second color difference reflecting a penalty due to the optional item when the optional item is inputted.

19. The step of obtaining whether all samples in the improvement sample group converge to a previously set reference range includes: The raw material acquisition method of claim 15, wherein the improved sample group convergence confirmation unit determines that all samples in the improved sample group converge to a pre-set standard range when the variance of the content of each raw material in the improved sample group is less than a pre-set standard value.

20. The step of generating the combined sample comprises: The method for obtaining raw materials according to claim 15, further comprising the step of: the formula sample improving unit randomly selecting one value of the existing sample and the mutant sample to generate the combination sample.

Images