Formulation determination method and formulation determination system

Abstract

To provide a formulation determination method capable of easily determining the composition of a food product.SOLUTION: The formulation determination method determines the formulation of a food material based on a user's request. The formulation determination method includes a selection step for selecting a desired base formulation from a plurality of base formulations preset for a predetermined food, a change input step for inputting a change to a value related to a functional item of the selected base formulation, and a formulation determination step for determining the formulation of the material based on the changed value related to a functional item.SELECTED DRAWING: Figure 11

Description

[Technical Field]

[0001] The present invention relates to a formulation determination method and a formulation determination system. [Background technology]

[0002] In recent years, food products have been created in response to user requests (see, for example, Non-Patent Document 1). When creating food products in response to user requests, multiple base recipes containing pre-mixed ingredients are prepared, and the user selects one base recipe and requests changes to the flavor, etc. of that base recipe.

[0003] To meet user requests, developers manually changed the ingredient mix from the base mix to create food with the ingredient mix that met the user's needs. [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] https: / / shouse.jp / oem / index.html Summary of the Invention [Problem to be solved by the invention]

[0005] However, it took a long time to create because the developers had to manually change the ingredient mix from the base mix.

[0006] The present invention aims to provide a formulation determination method and formulation determination system that can easily determine the formulation of foods. [Means for solving the problem]

[0007] [1] A method for determining a food ingredient composition based on a user's request, a selection step in which a desired base formulation is selected from a plurality of pre-defined base formulations for a given food product; A change input step in which changes to values ​​related to the selected function items of the base formulation are input; and a blending step of determining a blending of materials based on values ​​relating to the changed function items.

[0008] [2] For each of the functional items of the base formulation, a sensory evaluation value is determined in advance by sensory evaluation, the change input step includes inputting a change to the sensory evaluation value. The method for determining the composition described in [1].

[0009] [3] The blending determination step is A conversion process for converting a change in the sensory evaluation value into a change in score based on the relationship between a score set according to the ingredients included in the base formulation and the sensory evaluation value in a plurality of the base formulations for each function item of the base formulation; A blending determination step of determining a blending of the materials in accordance with the change in the score. [2] The method for determining the composition described above.

[0010] [4] For each of the functional items of the material, a sensory evaluation value is determined in advance by sensory evaluation; The score of the base formulation is set based on the sensory evaluation value, formulation amount, and potency of the ingredients. [3] The method for determining the composition described above.

[0011] [5] The change in the score is represented as a difference vector from a current position based on the sensory evaluation value of the base formulation to a target position based on a target score corresponding to the changed sensory evaluation value; The blending determination step includes: the probability of selecting the ingredient is set higher as the similarity between the direction of movement from the current position based on the change in the score when the ingredient is added to or subtracted from the base recipe and the direction of the difference vector becomes higher; [3] The method for determining the composition described above.

[0012] [6] The blending determination step determining an amount of the ingredient selected according to the set probability to be added or subtracted from the ingredient by a probability distribution centered on the amount that matches the target position; [5] The method for determining the composition of matter described above.

[0013] [7] The blending determination step a position based on the score of a formulation changed by adding or subtracting the determined amount of the selected ingredient to or from the base formulation is set as the current position, and the higher the similarity between the direction of a difference vector from the current position to the target position and the direction of movement from the current position when the ingredient is added to or subtracted from the changed formulation, the higher the probability of selecting the ingredient is set; [6] The method for determining the composition of matter described above.

[0014] [8] The blending determination step determining the amount of the material to be added or dropped selected according to the set probability based on a probability distribution centered on the amount that coincides with the target position; the current position is determined to be a position based on the score of the composition changed by adding or subtracting the determined amount of the selected ingredient to or from the changed composition, and the probability of selecting the ingredient is set higher the higher the similarity between the direction of the difference vector from the current position to the target position and the direction of movement from the current position when the ingredient is added to or subtracted from the changed composition; and the composition of the ingredients is determined by repeating this process a predetermined number of times. [7] The method for determining the composition of matter described above.

[0015] [9] The method further includes a display step of displaying the determined combinations of the plurality of ingredients in order of similarity to the target score corresponding to the changed sensory evaluation value. [3] The method for determining the composition described above.

[0016]

[11] The relationship is updated using a sensory evaluation value determined by a sensory evaluation for each function of the ingredient combination determined in the ingredient combination determination step, and a score set according to the ingredients included in the determined ingredient combination. [3] The method for determining the composition described above.

[12] The functional items include at least one selected from sweetness, saltiness, sourness, bitterness, umami, and spiciness; The method for determining a blending ratio according to any one of [1] to

[10] .

[0017]

[12] further comprising a cost setting step in which a desired cost value for the food product is set; A cost value is assigned to each of the materials; the blending step determines a blending of materials based on the value related to the changed function and the cost value; [1] The method for determining the composition described above.

[0018]

[13] A composition determination system that determines the composition of ingredients in food based on a user's request, An input unit for inputting a desired base composition from a plurality of base compositions preset for a specific food product and changes to values ​​related to the function items of the selected base composition; A compounding determination system comprising: a compounding determination unit that determines a compounding of materials based on values ​​related to the changed function items. [Effects of the Invention]

[0019] According to the present invention, it is possible to provide a composition determination method and a composition determination system that can easily determine the composition of foods. [Brief explanation of the drawings]

[0020] [Figure 1] 1 is a block diagram showing the configuration of a formulation determination system according to an embodiment of the present invention. [Figure 2] (a) A diagram showing a table of the materials and their amounts contained in each of multiple base formulations stored in the memory unit of the formulation determination system, (b) A diagram showing an example of the table of Figure 2(a). [Figure 3] FIG. 3(a) is a diagram showing a table of sensory evaluation values ​​set for functional items in each of a plurality of base formulations, and FIG. 3(b) is a diagram showing an example of the table of FIG. 3(a). [Figure 4] FIG. 4(a) is a diagram showing a table of sensory evaluation values ​​and potency values ​​set for each function item for each material, and FIG. 4(b) is a diagram showing an example of the table in FIG. 4(a). [Figure 5] FIG. 5(a) is a table showing the sensory evaluation values ​​and scores for functional items for each base formulation, and FIG. 5(b) is a diagram showing an example of the table of FIG. 5(a). [Figure 6] FIG. 1A is a diagram showing an example of a graph for calculating a coefficient of sweetness, and FIG. 1B is a diagram showing an example of a graph for calculating a coefficient of saltiness. [Figure 7] FIG. 10 is a schematic diagram for explaining cosine similarity used when calculating the probability of selecting a material. [Figure 8] FIG. 10 is a schematic diagram for explaining how to determine the blending amounts of selected materials. [Figure 9] FIG. 10 is a schematic diagram for explaining the process of determining the material composition so as to approach the target position. [Figure 10] A figure showing an example of a display of a formulation determined by the formulation determination system of an embodiment of the present invention. [Figure 11] 1 is a flow chart showing the control operation of a formulation determination system according to an embodiment of the present invention. [Figure 12] A block diagram showing the configuration of a formulation determination system according to a modified example of an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0021] Hereinafter, a formulation determination system and a formulation determination method according to an embodiment of the present invention will be described with reference to the drawings.

[0022] (Overview of Mixing Decision System 1) The blending determination system 1 of this embodiment determines the blending of ingredients for food based on the user's request. Examples of food include liquid foods such as curry or soup, and seasonings (powdered seasonings).

[0023] 1 is a block diagram showing the configuration of a blending determination system 1. The blending determination system 1 includes a terminal device 10 and a cloud server 20. The terminal device 10 may be, for example, a personal computer, a tablet, a smartphone, or the like.

[0024] The terminal device 10 includes a processor and a storage device. The processor is, for example, a CPU (Central Processing Unit). Alternatively, the processor may be a processor different from the CPU. The processor executes processing for formulation determination according to a program stored in the storage device. The storage device includes a non-volatile memory such as a ROM (Read Only Memory) and a volatile memory such as a RAM (Random Access Memory). The storage device may include an auxiliary storage device such as a hard disk or an SSD (Solid State Drive). The storage device is an example of a non-transitory computer-readable recording medium.

[0025] The terminal device 10 has an input unit 11, a display unit 12, and a transmission / reception unit 13. The terminal device 10 realizes the functions of the input unit 11, the display unit 12, and the transmission / reception unit 13 by executing a program stored in a storage device.

[0026] An operator operates the input unit 11 of the terminal device 10 to select a desired base formulation from a plurality of base formulations preset for a specific food product in response to a user (client) request. The input unit 11 is operated by the operator to input changes to values ​​related to functional items of the selected base formulation in response to a user (client) request. Examples of the input unit 11 include a touch panel, a keyboard, a mouse, and a switch.

[0027] The display unit 12 displays an input screen when the operator operates the input unit 11. An example of the display unit 12 is a display. If the input unit 11 is a touch panel, it may also serve as the display unit 12.

[0028] The transceiver 13 communicates signals with the transceiver 22 of the cloud server 20. The transceiver 13 transmits information about the selected base formulation input to the input unit 11 as a selection signal to the cloud server 20. The transceiver 13 transmits information about changes to values ​​related to the functional items of the selected base formulation input to the input unit 11 as a change signal to the cloud server 20. The transceiver 13 receives information about the formulation determined by the cloud server 20 as a formulation signal from the cloud server 20. Communication between the transceiver 13 of the terminal device 10 and the transceiver 22 of the cloud server 20 may be via, for example, the Internet, Bluetooth (registered trademark), a LAN, a telephone line, an in-house network, WiFi (Wireless Fidelity), or other communication lines, or a combination of these, and may be wired or wireless.

[0029] The cloud server 20 determines the ingredient combination based on the values ​​of the changed function items. The terminal device 10 displays the combination determined by the cloud server 20.

[0030] Although FIG. 1 shows a configuration in which one terminal device 10 communicates with the cloud server 20, a plurality of terminal devices 10 may communicate with the cloud server 20.

[0031] (Cloud Server 20) The cloud server 20 includes a processor and a storage device. The processor is, for example, a CPU (Central Processing Unit). Alternatively, the processor may be a processor different from the CPU. The processor executes processing for formulation determination according to a program stored in the storage device. The storage device includes a non-volatile memory such as a ROM (Read Only Memory) and a volatile memory such as a RAM (Random Access Memory). The storage device may include an auxiliary storage device such as a hard disk or an SSD (Solid State Drive). The storage device is an example of a non-transitory computer-readable recording medium.

[0032] The cloud server 20 has a memory unit 21, a transmission / reception unit 22, and a combination determination unit 23. The memory unit 21 is included in the storage device. The cloud server 20 realizes the functions of the transmission / reception unit 22 and the combination determination unit 23 by executing a program stored in the storage device.

[0033] (Storage unit 21) First, we will explain the data stored in the storage unit 21 of the cloud server 20. The storage unit 21 stores the compounding amounts of ingredients in a plurality of base formulations, the sensory evaluation values ​​of the functional items in the base formulations, the sensory evaluation values ​​and potencies of ingredients used to determine the formulations, and the coefficients of the functional items.

[0034] (Base formulation) A plurality of base formulations are set in advance. The base formulation is composed of a plurality of materials, and each material is contained in a predetermined blending amount. If the number of base formulations is n, then in FIG. 1, base formulations A1 to A nare stored in advance in the storage unit 21. The storage unit 21 stores the ingredients contained in each of the plurality of base formulations and the amounts of ingredients included.

[0035] FIG. 2(a) is a table showing the ingredients and their amounts contained in each of a plurality of base formulations. The top row shows n base formulations A i (1≦i≦n). The leftmost column shows m materials B j (1≦j≦m) Base formulation A i Ingredient B included in j The amount of α AiBj For example, the amount of material B2 in base mixture A1 is α A1B2 The blending amount α AiBj includes the value zero.

[0036] For example, if the food is curry, the base formula A1 contains ingredient B. j The ingredients B include roux, salt, sugar, curry powder, water, and adjusted lard. Base formulation A2 also includes multiple ingredients B, but is different from base formulation A1. For example, by increasing the amount of sugar in base formulation A1 compared to base formulation A2, base formulation A1 can be made into a sweeter curry. In this way, multiple base formulations A1 to A n is set in advance and stored in the storage unit 21.

[0037] Figure 2(b) is a table showing the specific example of Figure 2(a). The ingredients shown in Figure 2(b) are listed for ease of explanation. In Figure 2(b), for example, base formula A1 is composed of 84 parts roux, 12 parts sauce, 6 parts curry powder, 10 parts salt, 19 parts sugar, 6 parts MSG (monosodium glutamate), and 2 parts beef extract mixed with water. The amounts are expressed, for example, in parts per thousand.

[0038] (Sensory evaluation values ​​of functional items in base formulation) Each base composition (A1~A nThe sensory evaluation has been carried out in advance by an evaluator for each base formulation (A1 to A2), and a sensory evaluation value has been set for each function item. n ) function items. Fig. 3(a) is a diagram showing a table of the sensory evaluation values ​​set for the function items for each base formulation. The top row shows the sensory evaluation values ​​set for the q function items C k (1≦k≦q). The leftmost column shows n base formulations A i (1≦i≦n).

[0039] Function items C1 to C q Examples of the flavors include sweetness, saltiness, sourness, bitterness, umami, spiciness, oiliness, beef flavor, buttery flavor, fruity flavor, spicy flavor, and herb flavor.

[0040] Base Formula A i Function item C in k The sensory evaluation value of AiCk The sensory evaluation value of the function item C2 of the base compound A1 is β A1C2 The sensory evaluation value β AiCk is a numerical value set between 0 and 5, for example, and is determined by an evaluator. For example, in this embodiment, the sensory evaluation value is set in increments of 0.5.

[0041] FIG. 3(b) is a table showing a specific example of FIG. 3(a). In FIG. 3(b), functional items such as sweetness, saltiness, sourness, bitterness, umami, and spiciness are used as examples. The functional items (sweetness, saltiness, sourness, bitterness, umami, spiciness) in base blend A1 are evaluated and set as (sweetness 2, saltiness 4, sourness 2.5, bitterness 1.5, umami 3, spiciness 1.5). The functional items (sweetness, saltiness, sourness, bitterness, umami, spiciness) in base blend A2 are evaluated and set as (sweetness 2, saltiness 2.5, sourness 1.5, bitterness 1.5, umami 3.5, spiciness 2).

[0042] (Sensory evaluation values ​​and potency of ingredients used to determine formulation) The above-mentioned materials B1 to B mThe storage unit 21 stores the sensory evaluation values ​​of the materials B1 to B2 in advance. m The sensory evaluation values ​​of the functional items set for each of the materials B1 to B2 are stored. m The table below shows the sensory evaluation values ​​set for each functional item for each of the following materials. j Function item C in k The sensory evaluation value of γ AiCk For example, the sensory evaluation value of the function item C2 of material B1 is γ B1C2 The sensory evaluation value γ BiCk is the sensory evaluation value β AiCk Similarly, the sensory evaluation value is a value indicated by a number between 0 and 5, for example, and is determined by an evaluator. For example, the sensory evaluation value is evaluated in increments of 0.5.

[0043] In FIG. 4(a), materials B1 to B m The potency of each of the materials B1 to B2 is shown in the storage unit 21. m The titer for each of the materials is recorded. j The titer of δ Bj The potency is expressed as 100 / (concentration at the time of sensory evaluation). The potency is a value based on the concentration when setting the sensory evaluation value. The lower the concentration of the material for the sensory evaluation value, the higher the potency is set. For example, the potency can be expressed as 100 / (concentration at the time of sensory evaluation). For example, a material diluted to 1% for sensory evaluation has a potency of 100, and a material diluted to 0.5% has a potency of 200.

[0044] Figure 4(b) is a diagram showing a specific example of Figure 4(a). In Figure 4(b), for example, the roux is set to (sweetness 0, saltiness 0, sourness 0, bitterness 0, umami 1, spiciness 0) and has a potency of 7. Also, the curry powder is set to (sweetness 0, saltiness 0, sourness 0, bitterness 4, umami 0, spiciness 3) and has a potency of 33.

[0045] (Function item coefficients) The coefficient of the feature item is the feature item C k Base formulation A1~A n Sensory evaluation value β AiCk And materials B1 to Bm Sensory evaluation value γ BiCk The score is calculated in advance from the relationship between the score set from and the score set from, and is stored in the storage unit 21. n Function item C for each of k Score ε AiCk For example, base formulation A i Function item C k Score ε AiCk is base formulation A i Materials B1~B m It is calculated from the sensory evaluation value, blend amount and power factor of each of the above. i Function item C k Score ε AiCk is calculated from the following formula: (Formula 1) TIFF2025146058000002.tif19161

[0046] Figure 5(a) shows the coefficients ζ of the function items. Ck The coefficients of the function items are shown. Ck is the slope of the sensory evaluation value relative to the score. Ck is the score / 1000 on the horizontal axis and the sensory evaluation value on the vertical axis, (β A1Ck , ε A1CK / 1000), (β A2Ck , ε A2CK / 1000), (β A3Ck , ε A3CK / 1000), (β A4Ck , ε A4CK / 1000), (β AkCk , ε AkCK / 1000), (β AnCk , ε AnCK / 1000) and linearly approximate it using, for example, the least squares method.

[0047] Fig. 5(b) is a diagram showing a specific example of Fig. 5(a), in which examples are shown for six types of functional items (sweetness, saltiness, sourness, bitterness, umami, and spiciness).

[0048] As shown in Figure 2(b), base formula A1 is composed of the following ingredients: roux, sauce, curry powder, salt, sugar, MSG, and beef extract. From the table in Figure 4(b), it can be seen that of the ingredients that make up base formula A1, the only ingredient that contributes to sweetness is sugar, and the sugar potency is 10. Also, as shown in Figure 2(b), the blend amount of sugar in base formula A1 is 19, so the sweetness score of base formula A1 is 5 (sensory evaluation value of ingredients) x 10 (potency) x 19 (blended amount) = 950.

[0049] In base formula A1, the ingredients that contribute to the salty taste are sauce, salt, and beef extract, as shown in the table in Figure 4(b). Therefore, the saltiness score for base formula A1 is {1 (sensory evaluation value of sauce) × 50 (potency of sauce) × 12 (amount of sauce) + 5 (sensory evaluation value of salt) × 67 (potency of salt) × 10 (amount of salt) + 1 (sensory evaluation value of beef extract) × 50 (potency of beef extract) × 2 (amount of beef extract)} = 600 + 3350 + 100 = 4050.

[0050] As shown in FIG. 6(a), the sensory evaluation value is plotted on the vertical axis, and (score / 1000) is plotted on the horizontal axis, and the relationship between the sensory evaluation value and the score for the sweet base formulations A1 to A6 is plotted. The slope of this approximate line, 1.8648, is the coefficient for sweetness. As shown in FIG. 6(b), the sensory evaluation value is plotted on the vertical axis, and the score is plotted on the horizontal axis, and the relationship between the sensory evaluation value and the score for the salty base formulations A1 to A6 is plotted. The slope of this approximate line, 1.2447, is the coefficient for saltiness. Similarly, examples of coefficients for sourness, bitterness, umami, and pungency are shown in the table in FIG. 5(b). The approximate line can be determined by the least squares method or the like.

[0051] (Transmitter / receiver 22) The transmitting / receiving unit 22 of the cloud server 20 receives, as a selection signal, information on the base composition selected by the operator operating the input unit 11. For example, when the operator operates the input unit 11 to select base compositions A1 to A2, n When the base formula A2 is selected, the transmitting / receiving unit 22 receives information indicating that the base formula A2 has been selected.

[0052] The transmitter / receiver 22 of the cloud server 20 receives, as a change signal, information about a change to the value related to the function item of the base formulation selected by the operator operating the input unit 11. For example, if the operator operates the input unit 11 to input sweetness 3 and umami 4 for the selected base formulation A2, the transmitter / receiver 22 receives, as a change signal, information indicating that sweetness 3 and umami 4 have been input. The transmitter / receiver 22 transmits, as a formulation signal, information about the formulation of ingredients determined by the formulation determination unit 23 to the terminal device 10.

[0053] (Bixture determination department 23) The formulation determination unit 23 determines the formulation of ingredients based on the formulation amounts of ingredients in multiple base formulations stored in the memory unit 21, the sensory evaluation values ​​of the functional items in the base formulation, the sensory evaluation values ​​and potencies of the ingredients used in formulation determination, and the coefficients of the functional items, so as to match the sensory evaluation value input in the input unit 11.

[0054] The composition determination unit 23 includes a conversion unit 31, a material selection unit 32, a composition amount determination unit 33, and a transfer unit .

[0055] The conversion unit 31 converts the change in the sensory evaluation value based on the change signal received from the input unit 11 into a change in the score. i Function item C k Sensory evaluation value β AiCk and the entered function item C k From the sensory evaluation value of AiCk The change amount η AiCk For example, if you enter sweetness 3 and umami 4 for the selected base formula A2, the sensory evaluation values ​​for base formula A2 are sweetness 2 and umami 3.5, so the change amounts are sweetness 1 and umami 0.5.

[0056] Base Formula A i Function item C k Sensory evaluation value β AiCk The change amount η AiCk the coefficient ζ Ck Divide by and multiply by 1000 to get the score change λCk =(η AiCk / ζ Ck ) × 1000. The reason for multiplying by 1000 is that the score was divided by 1000 when calculating the coefficient of the function item. C1、 λ C2、··· λ Ck , ···, λ Cq ) is base formulation A i The amount of change η is expressed as a difference vector from the current position vector based on the sensory evaluation value of the vehicle to the target position vector based on the sensory evaluation value changed by the input unit 11. AiCk Since λ includes positive, negative, and zero values, Ck is a value including positive values, negative values, and zero. The difference vector can be expressed by the following equation (2).

[0057] (Formula 2) TIFF2025146058000003.tif26161

[0058] Note that the cloud server 20 does not need to specifically calculate the current position vector and the target position vector, but only calculates the difference vector. However, for ease of understanding, the current position vector and the target position vector will be described below.

[0059] Base Formula A i The current position vector for the material can be expressed as a value obtained by multiplying the score shown in FIG. 5(a) by the coefficient of the function, as shown in the following formula (3).

[0060] (Formula 3) TIFF2025146058000004.tif8161 The target position vector can be expressed as a value obtained by adding the amount of change in the score to the current position, as shown in the following equation (4).

[0061] (Formula 4) TIFF2025146058000005.tif15170

[0062] For example, if only sweetness, saltiness, sourness, bitterness, umami, and spiciness are used among the function items as described above, the sensory evaluation value of sweetness has changed from 2 to 3, so the conversion unit 31 calculates a change amount of 1 and calculates a change amount of the score as (1 / 1.865)×1000=536. Furthermore, the sensory evaluation value of umami has changed from 3.5 to 4, so the conversion unit 31 calculates a change amount of 0.5 and calculates a change amount of the score as (0.5 / 0.7391)×1000=676.

[0063] Therefore, conversion unit 31 obtains the difference vector as (536, 0, 0, 0, 676, 0). The current position vector is expressed as (950×1.865, 4050×1.2447, 600×1.1286, 200×0.8149, 2288×0.7391, 594×0.7148), and the target position vector is expressed as (950×1.865+536, 4050×1.2447, 600×1.1286, 200×0.8149, 2288×0.7391+676, 594×0.7148).

[0064] The material selection unit 32 sets a higher probability of selecting a material with a higher similarity the more similar the direction of movement from the current position based on the change in score when an ingredient is added to or subtracted from the base recipe is to the direction of movement of the difference vector.

[0065] The material selection unit 32 selects all materials B1 to B2 based on the difference vector. m The mixture determination unit 23 calculates the probability of selecting the materials B1 to B m Any one of the materials B a When the compounding amount is changed, the material change vector and the difference vector form an angle of cosθ a Calculate the probability based on the cosine similarity. a is calculated from the vector inner product formula using the following formula (5).

[0066] (Formula 5) TIFF2025146058000006.tif27161

[0067] The blending determination unit 23 determines all the materials B1 to B m Calculate cosθ for each of the materials B a The probability of selecting P a is calculated by the following formula (6). The molecule is one material B a The denominator indicates the sum of the cosine similarities of all materials.

[0068] (Formula 6) TIFF2025146058000007.tif13161 FIG. 7 is a schematic diagram for explaining the cosine similarity. In FIG. 7, c and the target position is V t For example, as described above, we will explain the case where the sensory evaluation value of sweetness is increased by +1 and the sensory evaluation value of umami is increased by +0.5 for base blend A1. As described above, the difference vector is (536, 0, 0, 0, 676, 0). For example, when salt is added, the material change vector when a blend amount of salt is added can be expressed as (0, 335, 0, 0, 0, 0) from the table in Figure 4(b). As shown in Figure 7, the angle between the material change vector due to the addition of salt and the difference vector is a right angle, and the dot product of the material change vector due to the addition of salt and the difference vector is 0, resulting in cosθ of salt = 0.

[0069] On the other hand, when MSG is added, the material change vector when MSG is added in a blending amount of 1 can be expressed as (0, 0, 0, 0, 250, 0) from the table in Figure 4(b). The angle between the material change vector and the difference vector when MSG is added in a blending amount of 1 is expressed as θa, and cosθa will have a value other than 0, and the probability that MSG will be selected can be calculated.

[0070] As described above, the composition determination unit 23 determines the composition of the materials B1 to B m Calculate cosθ for each of the above and find the probability Pa for each material.

[0071] The material selection unit 32 selects one material based on the determined probability. Note that a material with a high probability is simply more likely to be selected, but the material with the highest probability is not necessarily selected. This is because, in this embodiment, in order to be able to determine multiple types of combinations as described below, it is necessary to avoid selecting only the same material.

[0072] The blending amount determination unit 33 determines the blending amounts of the materials selected by the material selection unit 32. Here, the amount of movement due to an increase or decrease in the selected material can be determined linearly, so the amount of increase or decrease that most closely approaches the target position can be found by calculation. However, in this embodiment, the blending amount determination unit 33 determines the blending amounts of the selected materials using a probability distribution so that multiple types of blends can be determined.

[0073] FIG. 8 is a schematic diagram for explaining how to determine the blending amount of the selected material. As shown in FIG. 8, at the current position V c When material is added from the target position V, it moves in the direction of the vector. t Position V closest to d The destination is determined by a normal probability distribution in units of 1 / 1000 (0.01) based on the distance from the current position V. c From position V d By mixing the selected material, the target position V t This position V d The destination is determined by the probability of normal distribution in the direction of increasing or decreasing the blend amount based on the reference value V. d Current position V c Side position V d If the compounding amount is determined to move to position V d Current position V c Position V opposite d In some cases, the compounding amounts are determined so that the material moves to "". As will be described later, in order to be able to determine multiple types of compounding, the compounding amounts of the selected materials are set according to probability.

[0074] The moving unit 34 is located at the current position V cThe material selected from the above is moved to a position where the selected material is increased or decreased by the determined blending amount. This position movement is in the direction of the material change vector of the selected material by the determined blending amount. The material selection unit 32 described above refers to the moved position as the current position V c as the target position V t The difference vector is calculated from the target position V t is the current position V c The position is expressed as a relative position from the current position V c Therefore, the distance between the moved position and the target position V can be expressed as a relative position from the target position V without specifically calculating the current position vector and the target position vector as explained in (Equation 3) and (Equation 4). t The difference vector between these two can be calculated.

[0075] Then, the material selection unit 32 selects the current position V after the movement. c and target position V t Based on the difference vectors between m The mixing amount determination unit 33 determines the mixing amount of the selected material. The movement unit 34 moves the current position V c The selected material is moved to a position where the determined amount of material has been added.

[0076] The combination determination unit 23 sets the moved position as the current position V c As above, calculate the selection probability of all materials, select materials, determine the blending amount of the selected materials, and return to the current position V c The number of repetitions Q can be set by the input unit 11, for example.

[0077] FIG. 9 is a schematic diagram for explaining the process of determining the material composition so as to approach the target position. c0 Let the current position after the first movement be V c1 Let the current position after the second movement be V c2 Then, set it sequentially, and the current position after Q times is V cQAs shown in Figure 9, by repeating the above control, the current position V c0 From the target position V t It will gradually approach towards.

[0078] The blending determination unit 23 repeats the process for the set number of times Q to the position V cQ The amount of change in the composition of each material in Base composition A i The blending amounts of the ingredients are changed from those in the formula 1 and the blending amounts are determined. This allows the user to determine a blend that has the desired function.

[0079] If the number of types of compounding amounts determined by the compounding determination unit 23 is R, the determination of the compounding amounts of the ingredients as described above is repeated R times. This allows the compounding determination unit 23 to determine R types of compounding that will achieve the desired compounding that is modified from the base compounding. The transceiver unit 22 transmits information about the R types of compounding determined by the compounding determination unit 23 to the display unit 12 as a display command signal.

[0080] The display unit 12 displays R types of combinations determined by the cloud server 20 based on the display command signal. Fig. 10 is a diagram showing the combination amounts displayed on the display unit 12. In Fig. 10, R=10 types of combinations are displayed. In Fig. 10, the "distance" is shown for each of the 10 types of combinations, and the order is displayed in ascending order of distance. The "distance" is the distance from the current position V after Q times. cQ and target position V t The "distance" is the distance between the initial difference vector and the position V after moving Q times. cQ The closer the distance, the closer the function to the target, so the earlier it is displayed in the ranking.

[0081] Based on the recipe displayed on the display unit 12, ingredients can be mixed to create food that meets the user's needs.

[0082] (Operation of the blending decision system) Next, the operation of the blending determination system will be described with reference to the flowchart of FIG.

[0083] First, in step S11, the operator uses the input unit 11 to input the base formulations A1 to A2. n The input unit 11 selects a base composition from the selected base composition A. i The information relating to the above is transmitted as a selection signal to the cloud server 20, and the selection signal is received by the cloud server 20. Step S11 corresponds to an example of a selection step.

[0084] Next, in step S12, the operator uses the input unit 11 to input the function item C k Sensory evaluation value β AiCk to a desired value. The input unit 11 transmits information on the changed sensory evaluation value as a change signal to the cloud server 20, and the cloud server 20 receives the change signal. Step S11 corresponds to an example of a change input step.

[0085] Next, in step S13, the conversion unit 31 of the cloud server 20 converts the sensory evaluation value β AiCk The change amount η AiCk Calculate the change amount η AiCk The score change amount λ Ck Step S13 corresponds to an example of a conversion step.

[0086] Next, in step S14, the combination determination unit 23 starts up a new combination. In steps S15 to S19 following step S14, one type of combination is determined.

[0087] Next, in step S15, the material selection unit 32 of the cloud server 20 calculates the score change amount λ Ck The initial difference vector obtained from the material B1~B m The probability of selection for each material is calculated from the material change vector when the blending amount of each material is changed. a Calculate.

[0088] Next, in step S16, the blending amount determination unit 33 determines the blending amount of the selected material based on the probability of the normal distribution shown in FIG.

[0089] Next, in step S17, the movement unit 34 changes the selected material by the determined blending amount, thereby moving the position from the current position by the blending amount, and sets the moved position as the current position.

[0090] Next, in step S18, the blending determination unit 23 determines whether the number of times of positional movement has reached the set number Q. If it is determined in step S18 that the number of times Q has not been reached, control returns to step S15, and steps S15, S16, and S17 are repeated until the number of times of positional movement reaches Q.

[0091] In step S18, if it is determined that the movement of the position has been repeated the set Q times, in step S19, the combination determination unit 23 determines the position V cQ The amount of change in the composition of each material in Base composition A i The compounding amount of the ingredients is changed from that in the target position V t and position V cQ The distance between the first and second ingredients is also calculated. Steps S14 to S18 correspond to an example of a combination determination process. Steps S13 to S18 correspond to an example of a combination determination step.

[0092] Next, in step S20, the combination determination unit 23 determines whether or not R types of combinations have been determined. If the launch of a new combination in step S14 has been performed R times, the combination determination unit 23 determines that R types of combinations have been determined. Note that the combination determination unit 23 may also determine that R types of combinations have been determined if the combination determination in step S19 has been performed R times.

[0093] Next, in step S21, the transmitter / receiver 22 of the cloud server 20 transmits information about the determined combination of R types as a combination signal to the terminal device 10. The transmitter / receiver 13 of the terminal device 10 receives the combination signal transmitted from the cloud server 120.

[0094] Next, in step S22, the display unit 12 displays the determined combination of R types, as shown in Fig. 10. Step S22 corresponds to an example of a display step.

[0095] The formulation determination method of this embodiment is to determine a plurality of base formulations A1 to A n From the desired base formulation A i Select the base formulation A you have selected. i Function item C k The value of β AiCk The changes to the function are entered, and the value β for the item of the changed function is AiCk +η AiCk The material composition is determined based on the above.

[0096] (Other embodiments) Although one embodiment of the present invention has been described above, the present invention is not limited to the above embodiment, and various modifications are possible without departing from the gist of the invention.

[0097] (A) In the above embodiment, as shown in FIG. 2(a), materials B1 to B m Base composition A1~A n However, it is not limited to these, and the materials B1 to B2 used in the blending m Base composition A1~A n It may contain materials that are not included in any of the above.

[0098] (B) In the above embodiment, the base formulations A1 to A nAlthough no limit is set for the amount of change in the base material, a limit may be set. For example, as shown in Fig. 2(b), when roux is the base material for base formulations A1 to A6, a limit of 80% to 120% may be set for the increase or decrease in the amount of change. The limit may be input by the operator using the input unit 11.

[0099] (C) In the above embodiment, the formulation determination unit 23 does not impose a limit on the number of times the same material is repeatedly selected during the Q repetitions, but a limit may be imposed on the number of times of selection. For example, if the number of repetitions Q is set to 20, the number of times MSG in FIG. 2(b) is selected may be limited to, for example, 6 times. This limit on the number of times of selection may be configured so that the operator can input it using the input unit 11.

[0100] (D) In the above embodiment, no reference is made to cost when determining the composition, but the composition of materials may be determined taking cost into consideration. In this case, the storage unit 21 stores the materials B1 to B2. m The cost values ​​of each of the above are stored. Along with steps S11 and S12, a cost setting step is provided, in which a desired cost value is input using the input unit 11. The desired cost value can be, for example, an upper limit value of the cost or a target value of the cost. In this case, the composition determination unit 23 determines the composition of materials taking the upper limit value into consideration. It may also be possible to set a weight for exceeding the desired cost value.

[0101] In the above embodiment, in step S16, the blending amount determination unit 33 determines the blending amount according to a normal distribution, based on the blending amount that minimizes the distance to the target position. This distance can be considered a flavor-related distance that approaches the target position by blending the ingredients. However, when considering the cost of ingredients, the distance is set based on the distance related to the flavor as well as the distance related to the ingredient cost. The distance related to the ingredient cost is the difference between the ingredient cost before adding the ingredient selected in step S15 and the set desired cost value. Furthermore, it is preferable to set this difference as, for example, the excess over the set desired cost value. The blending amount determination unit 33 calculates the Euclidean distance to the target position from the distance related to the flavor and the distance related to the ingredient cost, and determines the blending amount according to a normal distribution, based on the blending amount that minimizes the Euclidean distance.

[0102] For example, if sweetness is "1" away from the target and sourness is "1" away, and the cost is also 1 (yen / kg), the distance to the target position taking raw material cost into consideration is the Euclidean distance (=√3). In step S16, the blending amounts of the ingredients selected in step S15 are determined according to a normal distribution based on the position where this Euclidean distance (√3) is shortest. The raw material cost is calculated based on the selected ingredients and the determined blending amounts.

[0103] Furthermore, in the above example, a deviation of 1 in flavor distance is equivalent to an excess of 1 (yen / kg) in raw material cost, but "weighting" can be used to change the value of an excess of 1 (yen / kg) in raw material cost relative to a deviation of 1 in flavor distance. For example, when "weighting" is set to 0.01, as described above, if sweetness is 1 and sourness is 1 from the target, and the cost is also 1 (yen / kg) above, the cost distance will be 0.01, which is the Euclidean distance (≒√2). The ingredient blend amounts are then determined according to a normal distribution based on the position where this Euclidean distance is shortest.

[0104] If the weighting is set high, it becomes difficult to generate a recipe that exceeds the cost price. On the other hand, if the weighting is set low, it is possible to generate a recipe that meets the requirements, even if the cost price exceeds the upper limit. In this way, the weighting of raw material costs can be changed depending on whether the cost requirements are strict or not, depending on the user.

[0105] The method of determining the material composition taking cost into consideration is not limited to the above. For example, if the composition determined in step S19 exceeds a set cost value, it may be discarded, and the flow from step S14 to step S20 may be repeated until a composition within the set cost value is determined.

[0106] (E) In the above embodiment, the blending amount and distance are displayed as shown in Fig. 10, but the difference from the sensory evaluation value desired by the user may also be displayed for each type of blend. For example, if the sweetness desired by the user is 3, the sweetness score may be calculated from the ingredients of a predetermined blend (for example, the blend ranked first) determined by the blend determination method of this embodiment, and the score may be converted to a sensory evaluation value using the relationship in Fig. 6(a), and the difference between the converted sensory evaluation value and the sensory evaluation value of 3 for the desired sweetness may be displayed.

[0107] (F) In the above embodiment, multiple types of determined combinations are displayed, but only one type of combination may be determined and displayed.

[0108] (G) In the above embodiment, the formulation determination system 1 includes a terminal device 10 and a cloud server 20, but may also be configured as a standalone device such as a personal computer or tablet. For example, the formulation determination system of the present application may be configured as the formulation determination device 100 of Figure 12. The formulation determination device 100 includes the above-mentioned input unit 11, display unit 13, memory unit 21, and formulation determination unit 23.

[0109] (H) In the above embodiment, the coefficient ζ of the function item Ckis a fixed numerical value stored in advance in the storage unit 21, but may be updatable. For example, a prototype is actually made of the composition displayed on the display unit 12 as shown in FIG. 10, and a sensory evaluation is performed on the prototype. This makes it possible to determine the sensory evaluation value for each functional item as shown in FIGS. 3(a) and 3(b) for the prototype. The sensory evaluation value for each functional item is input using the input unit 11 and transmitted to the cloud server 20. The cloud server 20 also determines a score for each functional item from the composition of the prototype. This determines the sensory evaluation value and score for the functional item for the composition of the prototype. The storage unit 21 of the cloud server 20 stores, for example, the sensory evaluation values ​​and scores for the functional items for each of the base compositions shown in FIG. 5(a). The cloud server 20 adds the sensory evaluation value and score of the functional item for the composition of the prototype to the sensory evaluation value and score, and calculates a new functional item coefficient ζ for each functional item. Ck The cloud server 20 calculates the coefficient ζ of the calculated function item. Ck is stored in the storage unit 21, and the coefficients of the previous function items are updated. As a result, the coefficients of the updated function items can be used when determining the next combination.

[0110] For example, using Figure 5(b), if the sensory evaluation score for sweetness of a prototype is 3.5 and the score is 1500, add the sensory evaluation score of 3.5 and the score of 1500 to the sensory evaluation score and score data for base formulations A1 to A6, plot the sensory evaluation score against score / 1000 as shown in Figure 6(a), and calculate the slope of the approximate line. This allows the coefficient for the sweetness function item to be updated to 1.969. In this way, new plots can be added and the coefficient can be corrected.

[0111] (I) In the above embodiment, the cloud server 20 is used as the server used in the blending determination system 1, but it is not limited to a virtual server such as a cloud server, and may be a physical server. [Industrial Applicability]

[0112] According to the method for determining the composition of the present invention, the composition of food can be easily determined, and can be used to determine the composition of liquid foods such as curry or soup, or seasonings (powdered seasonings), for example. [Explanation of symbols]

[0113] 11: Input section 12: Controller 13: Display section 21: Storage section 22: Transmitter / receiver 23: Mixture determination section 31: Conversion section 32: Material selection section 33: Blend amount determination section 34: Moving section

Claims

1. A method for determining a food ingredient composition based on a user's request, comprising: a selection step in which a desired base formulation is selected from a plurality of pre-defined base formulations for a given food product; A change input step in which changes to values ​​related to the selected function items of the base formulation are input; and a blending step of determining a blending of materials based on values ​​relating to the changed function items.

2. A sensory evaluation value is determined in advance for each function item of the base formulation by sensory evaluation, the change input step includes inputting a change to the sensory evaluation value. The method for determining a formulation according to claim 1.

3. The blending determination step includes: A conversion process for converting a change in the sensory evaluation value into a change in score based on the relationship between a score set according to the ingredients included in the base formulation and the sensory evaluation value in a plurality of the base formulations for each function item of the base formulation; A blending determination step of determining a blending of the materials in accordance with the change in the score. The method for determining a blending ratio according to claim 2.

4. a sensory evaluation value is determined in advance for each of the functional items of the material through sensory evaluation; The score of the base formulation is set based on the sensory evaluation value, formulation amount, and potency of the ingredients. The method for determining a blending ratio according to claim 3.

5. the change in the score is represented as a difference vector from a current position based on the sensory evaluation value of the base formulation to a target position based on a target score corresponding to the changed sensory evaluation value; The blending determination step includes: the probability of selecting the ingredient is set higher as the similarity between the direction of movement from the current position based on the change in the score when the ingredient is added to or subtracted from the base recipe and the direction of the difference vector becomes higher; The method for determining a blending ratio according to claim 3.

6. The blending determination step includes: determining an amount of the ingredient selected according to the set probability to be added or subtracted from the ingredient by a probability distribution centered on the amount that matches the target position; The method for determining a blending ratio according to claim 5.

7. The blending determination step includes: a position based on the score of a formulation changed by adding or subtracting the determined amount of the selected ingredient to or from the base formulation is set as the current position, and the higher the similarity between the direction of a difference vector from the current position to the target position and the direction of movement from the current position when the ingredient is added to or subtracted from the changed formulation, the higher the probability of selecting the ingredient is set; The method for determining a blending ratio according to claim 6.

8. The blending determination step includes: determining the amount of the material to be added or dropped selected according to the set probability based on a probability distribution centered on the amount that coincides with the target position; the current position is determined to be a position based on the score of the composition changed by adding or subtracting the determined amount of the selected ingredient to or from the changed composition, and the probability of selecting the ingredient is set higher the higher the similarity between the direction of the difference vector from the current position to the target position and the direction of movement from the current position when the ingredient is added to or subtracted from the changed composition; and the composition of the ingredients is determined by repeating this process a predetermined number of times. The method for determining a blending ratio according to claim 7.

9. a display step of displaying the determined combinations of the ingredients in order of similarity to the target score corresponding to the changed sensory evaluation value; The method for determining a blending ratio according to claim 3.

10. The relationship is updated using a sensory evaluation value determined by a sensory evaluation for each function of the ingredient combination determined in the ingredient combination determination step, and a score set according to the ingredients included in the determined ingredient combination. The method for determining a blending ratio according to claim 3.

11. The functional items include at least one selected from sweetness, saltiness, sourness, bitterness, umami, and spiciness. The method for determining a blending ratio according to any one of claims 1 to 10.

12. a cost setting step in which a desired cost value for the food product is set; A cost value is assigned to each of the materials; the blending step determines a blending of materials based on the value related to the changed function and the cost value; The method for determining a formulation according to claim 1.

13. A compounding system that determines the compounding of ingredients in food based on a user's request, An input unit for inputting a desired base composition from a plurality of base compositions preset for a specific food product and changes to values ​​related to the function items of the selected base composition; A compounding determination system comprising: a compounding determination unit that determines a compounding of materials based on values ​​related to the changed function items.

Images