Automatic cooker, automatic cooking method, computer program, and storage medium

Abstract

Capable of controlling the texture of food to be cooked on the basis of the odour of the food to be cooked, this automatic cooker (10), automatic cooking method, computer program and storage medium can be used for automatic cooking. The automatic cooker (10) includes a cooking unit (11), a storage unit (12), an odour sensor (13) and a control unit (14). The cooking unit (11) cooks the food to be cooked. The storage unit (12) stores an odour reference relating to the texture of the food to be cooked. The odour sensor (13) detects odour information relating to the odour of the food to be cooked. The control unit (14) acquires the odour information by means of the odour sensor (13). On the basis of the odour reference and the odour information, the control unit (14) sets parameters for a cooking operation of the cooking unit (11) so as to change the texture of the food to be cooked, and controls the cooking unit (11) to cook the food to be cooked.

Description

Automatic cooking device, automatic cooking method, computer program, and storage medium

[0001] The present disclosure relates to an automatic cooker, an automatic cooking method, a computer program, and a storage medium, and more particularly to an automatic cooker, an automatic cooking method, a computer program, and a storage medium that automatically cook food.

[0002] In recent years, technologies have been developed for automatic cooking appliances such as microwave ovens that automatically heat and cook food using microwaves or heaters. Technologies for controlling automatic cooking using odor sensors such as gas sensor arrays have also been developed.

[0003] Special Publication No. 2008-525755

[0004] Patent Literature 1 discloses a cooking device equipped with a gas sensor array, which analyzes the atmosphere in the cooking chamber of the automatic cooking device measured by the gas sensor array to prevent over-burning during automatic cooking, and performs the cooking process according to the analysis results.

[0005] However, in automatic cooking, there is room for improvement in the texture, seasoning, and heating level of the food.

[0006] An object of the present disclosure is to provide an automatic cooker, an automatic cooking method, a computer program, and a storage medium that are capable of controlling the texture of a food based on the smell of the food.

[0007] In order to solve the above-mentioned problems, the automatic cooker, automatic cooking method, computer program, and storage medium of the present disclosure have the following aspects.

[0008] An automatic cooker according to one aspect of the present disclosure includes a cooking unit, a memory unit, an odor sensor, and a control unit. The cooking unit cooks the food. The memory unit stores odor standards related to the texture of the food. The odor sensor detects odor information related to the odor of the food. After cooking begins, the control unit acquires the odor information using the odor sensor. The control unit sets parameters for the cooking operation of the cooking unit based on the odor standards and the odor information to change the texture of the food. The cooking unit uses the parameters to control the cooking unit and cook the food.

[0009] Another aspect of the automatic cooking method according to the present disclosure is a method for automatically cooking food using an automatic cooker, which includes the steps of acquiring odor information about the odor of the food using an odor sensor, setting parameters for cooking operations of the automatic cooker so as to change the texture of the food based on the odor criteria and the odor information about the texture of the food, and controlling the automatic cooker using the parameters to cook the food.

[0010] A computer program according to another aspect of the present disclosure is a computer program for causing an automatic cooker to execute the automatic cooking method described above.

[0011] Another aspect of the present disclosure is a non-transitory computer-readable storage medium having a computer program stored therein, the computer program implementing the aforementioned automatic cooking method when executed by a processor.

[0012] According to the automatic cooker, automatic cooking method, computer program, and storage medium disclosed herein, the texture of the food can be controlled based on the smell of the food.

[0013] 1. A block diagram showing the general configuration of an automatic cooker. 2. A schematic diagram of an example of an automatic cooker that is an autocooker. 3. A schematic diagram of an example of an automatic cooker that is a home bakery. 4. A flow chart of an example of an automatic cooking method. 5. An example of aroma components detected from food cooked based on the recipe "soybean" or the recipe "bread". 6. A flow chart of an example of step S200. 7. An example of a standard control sequence for the recipe "soybean" and a graph showing the association of smells detected from food. 8. An example of a standard control sequence for the recipe "bread" and a graph showing the association of smells detected from food. 9. A flow chart of an example of an automatic cooking method. 10. An example of preference information. 11. An example of a graph showing the association of smells detected from food. An example of a graph showing the relationship between aromas and sensory evaluations for a recipe "soybean" An example of learned data of cooking conditions for the recipe "soybean" An example of learned data of sensory evaluations of dishes cooked under different cooking conditions for the recipe "soybean" An example of learned data of aroma components detected from dishes cooked under different cooking conditions for the recipe "soybean" An example of the relationship between aroma components and sensory evaluations for a recipe "soybean" An example of learned data of cooking conditions for the recipe "bread" An example of learned data of sensory evaluations of dishes cooked under different cooking conditions for the recipe "bread" An example of the relationship between aroma components and sensory evaluations for the recipe "bread" A flowchart of an example of an automatic cooking method

[0014] First, various aspects of the automatic cooker, automatic cooking method, computer program, and storage medium of the present disclosure will be described.

[0015] An automatic cooker according to a first aspect of the present disclosure includes a cooking unit, a memory unit, an odor sensor, and a control unit. The cooking unit cooks the food. The memory unit stores odor standards related to the texture of the food. The odor sensor detects odor information related to the odor of the food. After cooking begins, the control unit acquires the odor information using the odor sensor. The control unit sets parameters for the cooking operation of the cooking unit based on the odor standards and the odor information to change the texture of the food. The cooking unit uses the parameters to control the cooking unit and cook the food.

[0016] In the automatic cooker of the second aspect according to the present disclosure, in the first aspect, the odor information may include aroma components that form the odor of the cooked food and the concentrations of the aroma components, and the odor standard may include aroma components and the concentrations of the aroma components associated with a specific texture.

[0017] In the automatic cooker of the third aspect of the present disclosure, in the first or second aspect, the memory unit may further store a standard control sequence corresponding to a specific recipe, and the control unit may set parameters to change the texture of the cooked food based on the odor criteria, the odor information, and the standard control sequence.

[0018] In the automatic cooker of the fourth aspect of the present disclosure, in the third aspect, the odor standard may include aroma components associated with a particular texture, and may also include the concentration of the aroma components at the timing when the parameters of the cooking operation are changed in the standard control sequence.

[0019] In the automatic cooker of the fifth aspect of the present disclosure, in any one of the first to fourth aspects, the parameters may include at least one of the amount of heat by the heating element that heats the food, the heating temperature by the heating element, the heating time by the heating element, the pressure inside the automatic cooker, the rotation speed of the rotating element that stirs the food in the cooking section, the rotation time of the rotating element, whether or not the steam generating element for heating or humidifying the food with steam supplies steam to the food, the amount of steam supplied by the steam generating element to the food, and the time for which steam is supplied by the steam generating element.

[0020] In the automatic cooker of a sixth aspect according to the present disclosure, in any one of the first to fifth aspects, the memory unit may further store a parameter standard value corresponding to the odor standard. The control unit may compare the odor information with the odor standard, and if it determines that the odor information satisfies the odor standard, may set parameters or terminate cooking based on the parameter standard value corresponding to the odor standard.

[0021] In the automatic cooker of a seventh aspect according to the present disclosure, in any one of the first to sixth aspects, the storage unit may further store preference information representing a user's preference for food texture. The odor standard can be set based on the preference information.

[0022] In the automatic cooker of an eighth aspect according to the present disclosure, in the seventh aspect, the odor standard may include aroma components and concentrations of aroma components associated with the texture expressed by the preference information.

[0023] In the automatic cooker of a ninth aspect according to the present disclosure, in the seventh or eighth aspect, the automatic cooker may further include a communication unit or an input unit. The control unit may receive the preference information via the communication unit or the input unit.

[0024] In the automatic cooker of the tenth aspect of the present disclosure, in the fourth aspect, the standard control sequence, odor standard, or standard parameter values ​​corresponding to the odor standard can be set for the recipe based on a trained model obtained by machine learning.

[0025] In an automatic cooker according to an eleventh aspect of the present disclosure, in the tenth aspect, the trained model may be one that has learned the relationship between cooking conditions, sensory evaluations, and detected aroma components and concentrations of the aroma components using input data including different cooking conditions, sensory evaluations of foods cooked under the cooking conditions, and aroma components and concentrations of the aroma components detected from the foods. The cooking conditions may include at least one cooking operation.

[0026] In the automatic cooker of a twelfth aspect according to the present disclosure, in the eleventh aspect, when specific cooking conditions are input, the trained model may estimate a sensory evaluation of a dish cooked under the cooking conditions, or aroma components and concentrations of the aroma components detectable in the dish. When the specific sensory evaluation is input, the trained model may estimate cooking conditions or aroma components and concentrations of the aroma components for obtaining the specific sensory evaluation. Alternatively, when the specific aroma components and concentrations of the aroma components are input, the trained model may estimate cooking conditions for obtaining the specific aroma components and concentrations of the aroma components, or a sensory evaluation perceived from the specific aroma components and concentrations of the aroma components.

[0027] In the automatic cooker of a thirteenth aspect of the present disclosure, in any one of the first to twelfth aspects, the odor standard may include a concentration of at least one of methylbutanol, acetoin, and glutamic acid.

[0028] In a fourteenth aspect of the present disclosure, the automatic cooker of any one of the first to thirteenth aspects may further include a communication unit. The control unit may receive data for adding or updating a standard control sequence, an odor standard, or preference information via the communication unit.

[0029] A fifteenth aspect of the present disclosure provides an automatic cooking method for automatically cooking an item using an automatic cooker, the automatic cooking method including the steps of: acquiring odor information about the odor of the item using an odor sensor; setting parameters for the cooking operation of the automatic cooker to change the texture of the item based on the odor criteria and the odor information; and controlling the automatic cooker using the parameters to cook the item.

[0030] A computer program according to a sixteenth aspect of the present disclosure is a computer program for causing an automatic cooker to execute the automatic cooking method according to the fifteenth aspect.

[0031] A seventeenth aspect of the present disclosure provides a storage medium that is a non-transitory computer-readable storage medium storing a computer program that, when executed by a processor, realizes the automatic cooking method of the fifteenth aspect.

[0032] <Technical Concept> In addition to the five tastes of sweetness, sourness, saltiness, bitterness, and umami, texture also has a significant impact on the deliciousness of cooked food. Texture refers to the sensations felt in the mouth and throat when food is placed in the mouth, such as the feel on the teeth, tongue, and throat. Because texture involves the sense of touch, it is also referred to as a physical taste. By controlling texture, the automatic cooking device and automatic cooking method disclosed herein can make cooked food more palatable to everyone or to suit the tastes of specific users. For example, a bread maker, which is an example of an automatic cooking device, controls the kneading time, dough temperature, and baking temperature to make bread with a chewy texture. An autocooker, which is an example of an automatic cooking device, controls the temperature and pressure in a cooking pot to make beef stew with a tender meat texture.

[0033] The automatic cooking device and automatic cooking method according to the present disclosure can also determine the texture of a food based on odor information detected from the food during cooking. The odor information includes aroma components that form the odor of the food and the concentration of those aroma components. Aroma components are volatile compounds that humans perceive as smelling. Each dish (recipe) has its own unique odor; for example, bread dough emits the odor of fermenting dough, and baked bread emits a fragrant odor. The inventors of the present disclosure have discovered that specific odors and aroma components are associated with texture. In this disclosure, machine learning is performed on the relationship between aroma components detectable in the food and the odor and texture for each recipe. The automatic cooking device and automatic cooking method use a trained model to detect the odor of the food while setting parameters such as heating time to change the texture of the food.

[0034] Each of the embodiments described below is an example of the present disclosure. The numerical values, shapes, configurations, steps, and step orders shown in the following embodiments are examples and do not limit the present disclosure.

[0035] In each of the embodiments described below, certain elements may be modified, and other elements may be appropriately combined with any configuration, and the combined configurations will provide the respective effects. In the embodiments, the respective combinations of the respective modified configurations will provide the respective effects of the respective modified configurations.

[0036] In the following detailed description of the embodiments, terms such as "first," "second," etc. are used for descriptive purposes only and should not be understood as expressing or implying relative importance or ranking of technical features. Features qualified as "first" and "second" expressly or imply the inclusion of one or more features.

[0037] Hereinafter, embodiments of an automatic cooker, an automatic cooking method, a computer program, and a storage medium according to the present disclosure will be described in detail with reference to the accompanying drawings as appropriate.

[0038] <<First Embodiment>> Fig. 1A is a block diagram showing the schematic configuration of an automatic cooking device. The automatic cooking device 10 cooks food by performing cooking operations such as heating on food contained in its internal cooking space. The automatic cooking device 10 may be, for example, a microwave oven, a microwave oven, an automatic cooker (automatic cooking device), an electric pressure cooker, a bread maker, a rice cake maker, a coffee maker, a waterless pot, a fryer, a toaster, an air fryer, or a low-temperature cooker. The food to be cooked by the automatic cooking device 10 may be ingredients or seasonings that have not yet been cooked, ingredients or seasonings that are in the middle of being cooked, or partially cooked food or semi-finished dishes.

[0039] The automatic cooker 10 includes a cooking unit 11, a memory unit 12, an odor sensor 13, and a control unit 14. The automatic cooker 10 may further include an input unit 15 that receives user input. The automatic cooker 10 may further include a communication unit 16 that is communicatively connected to a terminal device 20 or a server 30. The automatic cooker 10 may further include a display unit that displays the startup status, cooking status, recipe options, remaining time for automatic cooking, cooking temperature, etc. The automatic cooker 10 may include a container for containing the food to be cooked, such as a cooking pot, a heating chamber, a heating tank, a steamer, a bread case, or a frying basket.

[0040] <Cooking Unit 11> The cooking unit 11 can perform various cooking operations on the food in the cooking space. The cooking operation may be heating or cooling the temperature in the cooking space to a specific temperature, or increasing or decreasing the pressure in the cooking space to a specific pressure. The cooking operation may be stirring the food in the container, kneading the food, crushing the food, or mashing the food. The cooking operation may be supplying steam to the food. The cooking operation may be, for example, adding seasoning to a container containing the food.

[0041] The cooking unit 11 does not need to perform all of these cooking operations, but only needs to perform at least one specific cooking operation depending on the type of automatic cooker 10, the purpose of the automatic cooker 10, or the recipe to be cooked. For example, a microwave oven can perform a heating cooking operation, an oven range can perform a heating and steam supply cooking operation, a pressure cooker can perform a heating and pressurizing cooking operation, and an auto cooker can perform a heating, pressurizing, and stirring cooking operation.

[0042] The cooking unit 11 may include various elements for performing cooking operations. For example, the cooking unit 11 includes a heating element 111. The heating element 111 is a device that generates energy for heating food, and is, for example, a microwave generator, an induction heater (IH heater), a radiant heater, a sheath heater, or a far-infrared heater.

[0043] The cooking unit 11 may also include a rotating element 112 that rotates the entire food or rotates on the food. The rotating element 112 includes a motor, a rotating shaft, and a rotating body, and the rotating body may be a rotating plate, a rotating roller, a rotating blade, a rotating blade, or a mill blade. The rotating element 112 can rotate the rotating plate on which the food is placed, or stir, knead, cut, scrape, crush, or grind the food.

[0044] The cooking unit 11 may also include a steam generating element 113 for heating or humidifying the food with steam. The steam generating element 113 generates steam (water vapor or superheated steam) by heating water and supplies the steam via a nozzle or the like. The steam generating element 113 may supply steam toward the food or away from the food. The steam generating element 113 may include a water tank, a heater, or a nozzle.

[0045] The cooking unit 11 may further include other elements. For example, the cooking unit 11 may include a storage element and an input element for adding seasonings, yeast, or ingredients to the food. The cooking unit 11 may include a valve mechanism or an opening for supplying steam generated during cooking to the outside of the automatic cooker 10. The cooking unit 11 may also include a filter element for filtering the food and coffee grounds.

[0046] The control unit 14 of the automatic cooker 10 sets cooking operation parameters to control the cooking unit 11 and cook the food. The controllable parameters for the cooking unit 11 include at least one of the amount of heat, heating temperature, and heating time by the heating element 111, the pressure in the cooking space, the rotation speed and rotation time of the rotating element 112, and whether or not, the amount of steam supplied, and the supply time of steam supplied by the steam generating element 113. The amount of heat and heating temperature by the heating element 111 can be controlled by controlling the frequency or heater power of the microwave generator.

[0047] <Storage Unit 12> The storage unit 12 is a recording medium that records various information and control programs, and may be a memory that functions as a work area for the control unit 14. The storage unit 12 is realized, for example, by a flash memory, a RAM, an SSD (Solid State Device), a hard disk, or other storage device, or an appropriate combination thereof. The storage unit 12 stores at least one odor standard related to the texture of a cooked food.

[0048] The memory unit 12 may further store criteria or thresholds for executing the automatic cooking function, information obtained from the odor sensor 13 or other sensors, or information obtained from the terminal device 20 or server 30. For example, the memory unit 12 may store odor criteria used for automatic cooking (described below) or thresholds included in the odor criteria. This information can be stored in the form of a table, database, or individual file, and is read by the control unit 14 when automatic cooking is performed. The memory unit 12 may also store recipes. A recipe specifies the execution of at least one cooking operation for at least one ingredient. The automatic cooker 10 can cook food based on a recipe selected by a user.

[0049] The storage unit 12 may also store a program for causing the control unit 14 to execute an automatic cooking method.

[0050] <Odor Sensor 13> The odor sensor 13 can detect multiple aroma components and their concentrations from air containing the odor of cooking food. Furthermore, the odor sensor 13 can determine (classify) multiple odor patterns from the air using machine learning. More specifically, the odor sensor 13 includes a sensor array consisting of multiple sensor elements, each with a sensitive membrane corresponding to a specific aroma component. With this configuration, the multiple sensor elements constituting the odor sensor detect specific volatile components assigned to each sensor element, convert them electrically, and output the detected volatile components. The sensor then generates an odor pattern composed of a composite of these outputs. The sensitive membrane is selective for the aroma components. For example, in a MEMS (Micro Electro Mechanical Systems)-type odor sensor, when the sensitive membrane adsorbs a selected aroma component, distortion occurs in the sensor element. The type and concentration of the odor are detected by detecting changes in electrical resistance due to distortion in the sensor element. Generally, the more sensor elements there are, the more aroma components can be detected and odor patterns can be determined, and the higher the detection accuracy. The number of sensor elements included in the odor sensor 13 may be, for example, 9, 16, 25, or 64. In the present disclosure, the concentration of an aroma component is an area value measured by a gas chromatogram, which is a dimensionless number and therefore has no unit.

[0051] A regression model (trained model) is obtained by labeling the smell pattern of a cooked food using human sensory evaluations, adding labels to the detected aroma components and concentrations, performing machine learning, and then performing regression analysis. The trained model learns the relationship between the sensory evaluation and the detected aroma components and their concentrations. For each smell pattern, the prediction accuracy can be calculated as the degree of prediction by regression analysis. In general, a trained model is considered to be able to effectively predict smell patterns with high prediction accuracy. For example, the coefficient of determination (R ２ When using R as the prediction accuracy, 0.5 is often used as a guideline. ２It is believed that for odor patterns with a value higher than 0.5, predictions can generally be made correctly.

[0052] If machine learning is performed using sensory evaluation on items related to the texture of cooked foods, the trained model will have learned the relationship between texture and detected aroma components, etc. If machine learning is performed using sensory evaluation on the smell, taste, and texture that humans perceive from cooked foods, the trained model will have learned the relationship between the smell, taste, and texture and the detected aroma components, etc. In other words, the odor sensor 13 of the automatic cooker 10 can use the trained model to detect the smell, taste, and texture that humans perceive from cooked foods, and can reproduce the results of sensory evaluations by humans.

[0053] Note that machine learning can be performed by the server 30, rather than the automatic cooker 10. In this case, the automatic cooker 10 downloads the results of machine learning, such as a trained model, from the server 30 via the communication unit 16, and executes the automatic cooking method using the trained model.

[0054] In addition to the odor sensor 13, the automatic cooker 10 may further include sensors for acquiring various information from inside the automatic cooker 10 in order to execute the functions of the automatic cooker 10. For example, the automatic cooker 10 may include a temperature sensor that detects the temperature inside the cooking space that contains the food to be cooked, or a pressure sensor that detects the pressure inside the cooking space. Information detected by these sensors is also stored in the memory unit 12 and later used by the control unit 14 or transmitted to the terminal device 20 or the server 30.

[0055] <Control Unit 14> The control unit 14 is a controller that controls the entire automatic cooker 10. The control unit 14 includes a general-purpose processor such as a CPU or MPU that executes programs to achieve predetermined functions. The control unit 14 invokes and executes control programs stored in the storage unit 12 to achieve various controls in the automatic cooker 10. The control unit 14 is not limited to a device that achieves predetermined functions through the cooperation of hardware and software, but may also be a hardware circuit designed specifically to achieve predetermined functions. In other words, the control unit 14 can be realized by various processors, such as a CPU, MPU, GPU, FPGA, DSP, or ASIC.

[0056] The control unit 14 acquires various data and information from the storage unit 12 or the sensors and controls the automatic cooker 10. The control unit 14 controls the automatic cooker 10 to automatically cook the food, based on the acquired odor information and odor criteria, in accordance with the automatic cooking method described below, so as to change the texture of the food.

[0057] <Input Unit 15> The input unit 15 is an input interface that accepts input commands from a user, and may include a touch panel, a touch pad, a switch, a button, a mouse, a keyboard, etc. The input unit 15 converts the input commands and operation contents accepted from the user into electrical signals and transmits them to the control unit 14.

[0058] <Communication Unit 16> The communication unit 16 can communicate with the terminal device 20 or the server 30, for example, by sending and receiving internet packets. The automatic cooker 10 can receive user input, such as a recipe selected by the user or information about the user's preferences, from the terminal device 20 via the communication unit 16. The automatic cooker 10 can download new odor standards, trained models, or control programs from the server 30 via the communication unit 16. The control unit 14 may also cooperate with the terminal device 20 or the server 30 via the communication unit 16. The communication unit 16 may communicate and send and receive data between the automatic cooker 10, the terminal device 20, and the server 30 in accordance with standards such as Wi-Fi (registered trademark), IEEE 802.2, IEEE 802.3, 3G, and LTE. The communication unit 16 may communicate via the Internet, an intranet, an extranet, a LAN, ISDN, a VAN, a CATV communication network, a virtual private network, a telephone line network, a mobile communication network, a satellite communication network, infrared, or Bluetooth (registered trademark).

[0059] <Terminal Device 20> The terminal device 20 may be, for example, a controller for the automatic cooker 10 or a controller capable of managing and controlling multiple types of home appliances. The terminal device 20 may also be an information terminal capable of data communication with the automatic cooker 10, such as a smartphone, mobile phone, tablet, wearable device, or computer equipped with a dedicated related application. A user can input the recipe to be cooked and preferences regarding texture and the like to the automatic cooker 10 via the terminal device 20 and the communication unit 16 of the automatic cooker 10. The automatic cooker 10 can transmit the cooking status of the automatic cooker, recipe options, remaining time for automatic cooking, and the like to the terminal device 20 via the communication unit 16.

[0060] <Server 30> The server 30 may be a management server of a manufacturer of the automatic cooker 10 for managing at least one automatic cooker 10, managing information about users of the automatic cooker 10, or collecting data. The server 30 may also be an application server that cooperates with related applications of the terminal device 20. As described above, the server 30 can perform machine learning to learn the relationship between the sensory evaluation and the detected aroma components and concentrations. Furthermore, as described below, the server 30 can perform machine learning to learn the relationship between cooking conditions, including at least one cooking operation, and the sensory evaluation, and the relationship between cooking conditions and the aroma components and concentrations. The server 30 may use the trained model to set odor standards used for automatic cooking by the automatic cooker 10. The server 30 may transmit recipes, trained models, odor standards, etc. to the automatic cooker 10.

[0061] FIG. 1B is a schematic diagram of an automatic cooker, an example of an automatic cooker 10. The cooking section 11 of the automatic cooker 10 includes a heating element 111, a rotating element 112, a container 114, and a stirring element 115. In the example of FIG. 1B, the heating element 111 is a heater, the rotating element 112 includes a motor, etc., the container 114 is a cooking pot, and the stirring element 115 is a stirring blade. The cooking section 11 also includes a valve mechanism (not shown) for releasing steam generated during cooking. The odor sensor 13 is located between the container 114 and an outlet 17 of the valve mechanism. The automatic cooker can perform cooking operations set by the user and can cook ingredients placed in the container 114 according to a specific menu until the dish is complete. The control unit 14 is located inside the automatic cooker 10 and may be located, for example, next to the container 114 in a position where the influence of the heating element 111 is minimal.

[0062] FIG. 1C is a schematic diagram of a bread machine, an example of an automatic cooker 10. The cooking section 11 of the automatic cooker 10 includes a heating element 111, a rotating element 112, a container 114, and a stirring element 116. In the example shown in FIG. 1C, the heating element 111 is a heater, the rotating element 112 includes a motor 112a and a rotating shaft 112b, and the stirring element 115 is a stirring blade. The bread machine automatically mixes, kneads, and ferments ingredients such as flour, water, and yeast contained in a container 114 to produce bread dough, and then bakes the dough to produce bread. Some bread machines can use rice flour or rice itself as an ingredient instead of wheat flour. For example, the bread machine may automatically grind rice grains into rice flour using a cutting and kneading element 116, and then add yeast to the rice flour to produce dough. The cooking section 11 includes a fan 19 to expel gas generated during cooking through an intake 18 in the cooking space. The odor sensor 13 is provided downstream of the fan 19, for example, between the fan 19 and the air outlet 17. The control unit 14 is provided inside the automatic cooker 10, and may be provided, for example, next to the container 114 in a position where the influence of the heating element 111 is small.

[0063] Up to this point, we have provided an overview of the configuration and operation of the automatic cooker 10 according to the present disclosure. From here on, we will explain the features of the automatic cooking method, computer program, and storage medium performed by the automatic cooker 10 according to the present disclosure.

[0064] <Automatic Cooking Method> The control unit 14 executes the automatic cooking method of the automatic cooker 10. According to the automatic cooking method, the texture of the food to be cooked can be controlled based on the smell of the food to be cooked.

[0065] FIG. 2 is a flowchart of an example of an automatic cooking method, which includes steps S100 to S300. As one example, a user inputs a cooking start command via the input unit 15. When the cooking start command is input, the control unit 14 of the automatic cooker 10 starts automatic cooking. As another example, the user inputs a designated time to start cooking via the input unit 15. The automatic cooker 10 starts automatic cooking at the designated time. In this way, the automatic cooking method according to the present disclosure can be executed at the start of automatic cooking and during automatic cooking. Furthermore, during cooking, the control unit 14 can change cooking operation parameters as necessary and execute automatic cooking while monitoring odor information acquired by the odor sensor 13 in accordance with the automatic cooking method according to the present disclosure.

[0066] The control unit 14 acquires odor information relating to the odor of the food being cooked using the odor sensor 13 (step S100).

[0067] A recipe specifies at least one cooking step for cooking ingredients to create a dish. A cooking step is a step in which the automatic cooker 10 performs at least one cooking operation on at least one ingredient. As an example of a cooking step, the recipe "Boiled Soybeans" (hereinafter abbreviated as "Soybeans") specifies the cooking step of "heating soybeans at 100°C for 70 minutes." As another example, the recipe "Soybeans" specifies the cooking step of "increasing the pressure inside the pot to 0.2 MPa and heating soybeans at 120°C for 50 minutes."

[0068] Smell information refers to information about the smell of the food being cooked in the cooking space and can be detected by the smell sensor 13. The smell information includes aroma components that form the smell of the food being cooked and may further include the concentration of these aroma components. The smell information may also include information representing an odor pattern determined using the detection results of the smell sensor 13 and a trained model. The odor pattern includes items related to texture, such as "hardness," "smoothness," "chewy texture," and "good texture." The odor pattern may further include items related to smell or taste, such as "sweet aroma," "grassy aroma," "umami," and "salty." The information representing the odor pattern may be the name of the odor pattern or a corresponding identification number. "Good texture" means that the cooked food has a good texture. Good texture may be a general, comprehensive sensory evaluation of texture, or may be a texture preferred by a specific user. In the case of a texture preferred by a specific user, the automatic cooker 10 may set information related to texture based on information input by the user into the automatic cooker 10.

[0069] Since multiple recipes specify different ingredients or cooking operations, it is conceivable that different recipes will produce different aroma components from the cooked food. Therefore, a trained model is generated for each recipe. Accordingly, each recipe and trained model corresponds to different odor pattern items and aroma components. Based on sensory evaluations and experimental results for various cooked food, aroma components related to texture include, for example, at least one of methyl butanol, acetoin, and glutamic acid.

[0070] FIG. 3 shows examples of aroma components detected from food cooked based on the recipe "soybeans" or the recipe "bread." Machine learning and regression analysis revealed that, among the various aroma components detected from cooked soybeans, the presence and concentration of 2-methyl-1-butanol / 3-methyl-1-butanol (2-Methyl-1-butanol / 3-Methyl-1-butanol) affect the "taste," "hardness," "chewy texture," "unpleasant odor," and "grassy aroma" of the food. Furthermore, machine learning and regression analysis revealed that the presence and concentration of thiophene affect the "smoothness" and "moistness" of the food. Furthermore, through machine learning and regression analysis, it was found that, in cooked soybeans, methylbutanol affects a "grassy aroma," pentane affects a "sweet aroma," "sweetness," and "umami," isobutane affects a "savory aroma" and "umami," and butyraldehyde and 2-methylbutanal / 3-methylbutanal affect a "savory aroma."

[0071] In addition to 2-methyl-1-butanol / 3-methyl-1-butanol, other butanal aroma components may also affect the texture. For example, components that have the same effect as methylbutanol may also affect the texture.

[0072] Next, the control unit 14 sets the parameters of the cooking operation of the automatic cooker 10 so as to change the texture of the food based on the odor criteria related to the texture of the food and the odor information acquired in step S100 (step S200).

[0073] Then, the control unit 14 controls the automatic cooker 10 to cook the food using the parameters set in step S200 (step S300).

[0074] The odor standard is a standard used to determine the timing of setting or changing parameters based on odor information. As an example, the odor standard includes aroma components associated with a particular texture. As another example, the odor standard further includes the concentration of the aroma components. When the odor standard includes the concentration of an aroma component, the odor standard may include, for example, the concentration of at least one of methylbutanol, acetoin, and glutamic acid. As another example, the odor standard includes an aroma component corresponding to a particular odor pattern and the concentration of the aroma component. In this case, for example, the odor standard may be "thiopene is 20." As another example, the odor standard may include the relative relationship between the concentrations of multiple aroma components. In this case, for example, the odor standard may be "the concentration of thiopene is equal to or less than the concentration of isobutane."

[0075] As described above, different aroma components are produced in the food cooked for each recipe. As an example, the aroma components included in the odor standard are determinable by the odor sensor 13 and correspond to at least one of the odor pattern items related to texture. In this case, the odor standard may include the relative relationship of the concentrations of aroma components related to the determinable odor pattern. For example, the odor standard may be "the concentration of the aroma component related to 'sweetness' is equal to or lower than the concentration of the aroma component related to 'fragrantness'."

[0076] In step S200, the control unit 14 sets cooking operation parameters based on the result of comparing the odor information with the odor standard. Figure 4 is a flowchart of an example of step S200. In step S200, the control unit 14 compares the odor information with the odor standard (step S210) and determines whether the odor information satisfies the odor standard (step S220).

[0077] When the odor standard is the concentration of a specific odor component, the odor information satisfying the odor standard refers to the corresponding odor component reaching the concentration specified in the odor standard. For example, when the corresponding odor component in the odor standard increases and exceeds the odor standard, or when the corresponding odor component in the odor standard decreases and falls below the odor standard, the control unit 14 determines that the odor standard is satisfied. For example, when the odor standard is "thiopene is 20," the control unit 14 compares the odor information with the odor standard to extract the concentration of the odor component "thiopene" from the odor information and determines whether the concentration exceeds 20.

[0078] If the control unit 14 determines that the odor information satisfies the odor standard, it sets cooking operation parameters based on the parameter standard values ​​corresponding to the odor standard (step S230). The parameter standard values ​​correspond to at least one cooking operation parameter and are stored in the storage unit 12.

[0079] As an example, the parameter standard value is a value that should be set for a corresponding parameter when the odor criterion is met. The control unit 14 sets the parameter corresponding to the met odor criterion to the parameter standard value corresponding to that odor criterion. For example, the odor criterion "the concentration of thiopene is equal to or lower than the concentration of isobutane" corresponds to the cooking operation "heating" and the parameter standard value "40°C." As another example, for the same odor criterion, the control unit 14 sets the parameter, the heating temperature of the heating element 111, to 40°C. As another example, for the same odor criterion, the control unit 14 sets the parameter, the heater power, so that the control space becomes 40°C.

[0080] In step S300, the control unit 14 controls the cooking unit 11 using the parameters set in this way to proceed with cooking.

[0081] For example, the odor criterion is a criterion for finishing a dish. When the control unit 14 determines that the odor criterion is satisfied during the final heating or finishing stage of cooking, the control unit 14 terminates the cooking. For example, the control unit 14 may set the heating amount of the heating element 111 to zero, control the valve to open so that the pressure in the cooking space is 0.1 MPa, or set the rotation speed of the rotating element 112 to 0 rpm.

[0082] <Cooking Using a Standard Control Sequence> As an example, the control unit 14 sets parameters to change the texture of the food based on the odor criteria, odor information, and standard control sequence. The standard control sequence is a control sequence that predefines the parameters to be set along a time axis or the flow of operations to be performed in order to control each element of the cooking unit 11. In other words, the standard control sequence directly or indirectly predefines the parameters of the cooking unit 11 so that the automatic cooker 10 automatically cooks food according to a recipe. Each element of the cooking unit 11 operates using the predefined parameters in accordance with the standard control sequence. A standard control sequence is generated in advance for each recipe and stored in the memory unit 12.

[0083] Figure 5 is an example of a graph showing the relationship between the standard control sequence for the recipe "soybeans" and the smell detected from the food being cooked. Figure 5 also shows graphs showing the time-dependent changes in the temperature inside the pot, the pressure inside the pot, and the concentration of ingredients when the standard control sequence for the recipe "soybeans" is executed.

[0084] The graph of component concentrations shown at the bottom of Figure 5 shows the change over time for each type of odor detected from the cooked food. The example shown in Figure 5 shows the change over time for (1) the component concentration of an odor associated with "sweetness," (2) the component concentration of an odor associated with "fragrantness" and "umami," and (3) the component concentration of an odor associated with "good texture." The odor associated with "sweet aroma" and "sweetness" corresponds to pentane, etc., the odor associated with "fragrantness" corresponds to isobutane, etc., the odor associated with "umami" corresponds to pentane / isobutane, etc., and the odor associated with "good texture" such as "smoothness" and "moistness" corresponds to thiopene, etc.

[0085] This recipe "soybeans" is cooked using a pressure cooker or autocooker, which is an example of an automatic cooker 10. The standard control sequence corresponding to the recipe "soybeans" specifies parameters for the temperature and pressure inside the container 114, which serves as a cooking pot. For example, the standard control sequence specifies that the temperature inside the pot is heated at 40°C for approximately 18 minutes (time t1 to time t2), then heated to 120°C, and further heated to 130°C. The standard control sequence then specifies that heating is stopped and the temperature inside the pot is lowered when the concentration of aroma components associated with the "sweet" odor pattern falls below the concentration of aroma components associated with the "savory" odor pattern (time t3), and cooking is completed just before the thiopene concentration begins to decrease (time t4). In an airtight controlled space, the pressure inside the pot changes with the temperature inside the pot, even without intentional control. Therefore, the pressure inside the pot can be controlled by controlling the temperature inside the pot or the amount of heat.

[0086] Controlling the pressure inside the pot based on the temperature inside the pot is not limited to controlling the pressure inside the pot; the pressure inside the pot can also be controlled directly using a pressure sensor that measures the pressure inside the pot. In this case, the program is set to pressurize the pot to 0.2 MPa approximately 25 minutes after the start of cooking (time t2), then further pressurize it to 0.3 MPa, and then reduce the pressure inside the pot once the concentration of aroma components associated with the "sweet" odor pattern falls below the concentration of aroma components associated with the "savory" odor pattern (time t3), and end cooking when the concentration of thiopene begins to decrease (time t4).

[0087] For example, the odor criterion may include an aroma component associated with a particular texture and the concentration of that aroma component at the timing when the cooking operation parameters are changed in the standard control sequence. For example, in the example of FIG. 5 , one odor criterion is "the thiopene concentration has begun to decrease." When the odor criterion is met (time t4), the control unit 14 stops heating and pressurizing. That is, the control unit 14 sets the heating amount of the heating element 111 to zero and the pressure inside the pot to 0.1 MPa.

[0088] As another example, the standard control sequence incorporates control for texture. For example, the standard control sequence of Fig. 5 specifies two-stage heating at 120°C and 130°C and two-stage pressure application at 0.2 MPa and 0.3 MPa (time t2 to time t3) to reduce the hardness of the food.

[0089] As another example, the standard control sequence incorporates control for taste or smell. For example, the standard control sequence in Figure 5 incorporates low-temperature cooking at 40°C (time t1 to time t2). This low-temperature cooking causes enzymatic decomposition of soybean protein, extracting the sweetness of the soybeans. Also, high-temperature cooking at 120°C to 130°C is specified. This high-temperature cooking promotes the Maillard reaction, enhancing the fragrant flavor of the cooked food.

[0090] 6 is an example of a graph showing the relationship between the standard control sequence for the recipe "bread" and the odor detected from the food being cooked. This recipe "bread" is cooked by a bread maker, which is an example of an automatic cooker 10. FIG. 6 shows graphs representing the changes over time in the temperature inside the pan, heater power, rotation speed, and ingredient concentration when the standard control sequence for the recipe "bread" is executed.

[0091] The graph of component concentrations shown at the bottom of Figure 6 shows the change over time for each type of odor detected from cooked food. The example shown in Figure 6 shows the change over time for (1) the component concentration of an odor associated with an "unpleasant odor," (2) the component concentration of an odor associated with a "fragrant" flavor, and (3) the component concentration of an odor associated with a "chewy texture." The odor associated with an "unpleasant odor" corresponds to 2-methyl-1-butanol / 3-methyl-1-butanol, etc., the odor associated with a "fragrant" flavor corresponds to 2-methylbutanal / 3-methylbutanal, etc., and the odor associated with a "chewy texture" corresponds to 2-methyl-1-butanol / 3-methyl-1-butanol, etc.

[0092] The standard control sequence corresponding to the recipe "bread" defines parameters such as the heating temperature for the container 114, the heater power to reach the heating temperature, and the rotation speed of the rotating element 112 that mixes / kneads the ingredients and dough.

[0093] For example, the heating control specifies that the temperature be maintained slightly below 35°C during the fermentation stage (time t5 to time t6), and that the baking stage (time t6 to time t7) be heated to 180°C and baked at 180 to 190°C for a while, before lowering the heating temperature to 170°C. To achieve the specified heating temperature, the heater power is first increased to 300 W during the kneading stage (up to time t5) before the fermentation stage, raising the heating temperature to 35°C. Then, during the fermentation stage, the heater power is maintained at 150 W and then intermittently increased to 300 W to maintain the temperature. When the baking stage begins, the heater power is set to 400 W to reach 180°C.

[0094] The rotating element 112 rotates at 300 rpm during the kneading stage to create dough, and rotates intermittently during the fermentation stage to release gases from within the dough.

[0095] 6, one odor criterion is that "the concentration of aroma components associated with a 'chewy texture' becomes 10 or less during the baking stage." When this odor criterion is met (time t7), the control unit 14 sets the heater power to zero and heats (bakes) the bread to completion using residual heat in the control space. Furthermore, by setting the odor criterion to "the concentration of aroma components associated with a 'chewy texture' and aroma components associated with an 'unpleasant odor' becomes 10 or less during the baking stage," both texture and odor can be controlled.

[0096] The standard control sequence for "bread" also incorporates controls for texture and odor. For example, during the fermentation stage, the temperature at which yeast fermentation becomes active is 35°C, but if the temperature rises above 35°C, unpleasant odors such as alcohol and stuffiness will occur. Therefore, by maintaining the temperature slightly below 35°C during the fermentation stage and stirring the dough intermittently, it is possible to promote fermentation while reducing unpleasant odors, and to prevent the texture from becoming too chewy.

[0097] Furthermore, in the early stage of the baking process, the heating temperature is raised to 180°C in one go, causing the water vapor inside the bread to evaporate in one go. This reduces unpleasant odors and adjusts the moisture content, optimizing the chewy texture. In the middle stage of the baking process, the heater power is slightly reduced, preventing the surface of the bread from burning too much, further reducing unpleasant odors and optimizing the chewy texture. Furthermore, in the later stage of the baking process, the heater is turned on to perform additional heating, evaporating the moisture near the surface and enhancing the fragrant aroma due to the Maillard reaction.

[0098] As described above, the automatic cooker 10 can control the texture of the cooked food using such a standard control sequence and feedback control based on odor information. The automatic cooker 10 can also control the taste and odor of the cooked food based on odor information related to the odor or taste. The taste and odor of the cooked food can also be controlled by incorporating control for the odor or taste into the standard control sequence. Control of texture, odor, and taste can be freely combined.

[0099] This completes the automatic cooking process for the automatic cooker 10. According to this automatic cooker 10 and automatic cooking method, various elements of the cooking unit 11 can be controlled to control the texture of the food based on its odor. Furthermore, the odor sensor 13 of the automatic cooker 10 can detect aroma components emitted from the food and their concentrations. Based on the aroma components, their concentrations, and combinations of aroma components, for example, using a trained model based on this information, the automatic cooker 10 can also determine the texture, odor, or taste perceived by humans from the food.

[0100] As an example, the automatic cooker 10 includes a computer program used to execute the above-described automatic cooking method. The computer program causes the control unit 14 of the automatic cooker 10 to execute the automatic cooking method.

[0101] As another example, the automatic cooker 10 includes a non-transitory computer-readable storage medium storing a computer program. The automatic cooking method of the present disclosure is realized when the computer program is executed by the processor (control unit 14). The storage medium may be the same as the storage unit 12 of the automatic cooker 10, may be included in the storage unit 12, or may be a component separate from the storage unit 12.

[0102] Second Embodiment Control Based on User Preference Information In the second embodiment, the automatic cooker 10 can cook food with a texture, smell, or taste that matches the user's preferences. For example, the automatic cooker 10 can automatically cook food by controlling the heating temperature, heating time, pressure in the cooking space, whether or not to stir, the supply of steam, and the like, so that the finished food approaches the user's preferences.

[0103] FIG. 7 is a flowchart of an example of an automatic cooking method executed by the automatic cooker 10 of the second embodiment.

[0104] 7, the automatic cooker 10 of the second embodiment executes step S400 and then executes steps S100 to S300. Steps S100 to S300 are the same as steps S100 to S300 of the first embodiment in FIG. 2, and detailed description thereof will be omitted.

[0105] In the second embodiment, before performing automatic cooking, the control unit 14 receives preference information via the input unit 15 or the communication unit 16 (step S400) and stores the preference information in the memory unit 12. The user inputs the preference information via the input unit 15 of the automatic cooker 10 or the terminal device 20. When the user inputs the preference information into the terminal device 20, the automatic cooker 10 receives the preference information from the terminal device 20 or the server 30 via the communication unit 16.

[0106] The preference information refers to information that represents a user's preferences. For example, the preference information may be information that represents the user's preferences regarding texture.

[0107] The odor standard or cooking operation parameters can be set based on preference information by the control unit 14 of the automatic cooker 10, the terminal device 20, or the server 30. For example, the server 30 may change a predetermined odor standard based on preference information. If the odor standard includes aroma components and their concentrations related to the preference information, the server 30 may change the concentrations specified in the odor standard based on the preference information. Also, for example, the control unit 14 may change parameters specified in the standard control sequence based on preference information.

[0108] The following mainly illustrates a case where the control unit 14 sets or changes the odor criteria or cooking operation parameters, but these operations can also be performed by the terminal device 20 or the server 30. When the server 30 performs the setting or change, the automatic cooker 10 or the terminal device 20 transfers the received preference information to the server 30.

[0109] In the second embodiment, the preference information corresponds to a specific recipe and represents the user's preferences for that recipe. The preference information may be related to aroma components and their concentrations defined in the odor standard for the specific recipe. For example, the preference information may include the content of "hardness -1" and "smoothness" corresponding to the recipe "soybean," and the content of "preferred texture is 'chewy'" corresponding to the recipe "bread." In other words, the preference information corresponds to at least one of the items of an odor pattern that can be determined or controlled for a specific recipe, or corresponds to the aroma components associated with that odor pattern.

[0110] If the preference information includes a user's preferred texture, the control unit 14 may set the preferred texture as a "good texture" category to be monitored using the odor sensor 13 during automatic cooking. For example, the preference information may include content such that "the preferred texture is a 'chewy' texture." In this case, the control unit 14 monitors odor information related to the "chewy texture" and sets cooking operation parameters to change the texture of the cooked food based on the odor information related to the "chewy texture." If the preference information includes multiple preferred textures, at least one preferred texture may be set as a "good texture," or all preferred textures may be set as "good textures" (e.g., categorized as "first good texture" and "second good texture").

[0111] When the preference information corresponds to a specific recipe, the control unit 14 may change the odor standard that defines the aroma components corresponding to the preference information. For example, the preference information may include "hardness -1," and the odor standard for the recipe "soybeans" may be set as follows: "After heating at 40°C for 20 minutes, the concentration of the aroma components corresponding to the odor pattern 'hardness' exceeds the first concentration threshold." When this odor standard is met, the control unit 14 heats the pot so that the temperature inside the pot reaches 120°C. Because soybeans become soft when cooked at a low temperature of 40°C, the control unit 14 sets the first concentration threshold higher to extend the low-temperature cooking time based on the preference information "hardness -1."

[0112] The control unit 14 can also set cooking operation parameters based on the preference information. For example, the control unit 14 sets the parameter "cooking time" for cooking at 40°C to be longer based on the preference information "hardness -1" so as to extend the low-temperature cooking time.

[0113] As an example, the preference information includes content unrelated to a specific recipe. In other words, the preference information corresponds to at least one item of an odor pattern, or corresponds to an aroma component associated with that odor pattern, regardless of the recipe. In this case, the control unit 14 determines whether there is an odor pattern or an aroma component corresponding to both the preference information and the recipe to be cooked. If there is, the control unit 14 may change the odor standard that defines the odor pattern or aroma component corresponding to the preference information. For example, as described above, the control unit 14 may change the concentration threshold defined in the odor standard based on the preference information. Note that a comparison table can be created, stored, and used for each recipe.

[0114] FIG. 8 shows an example of preference information. In the example shown in FIG. 8, preference information A to C representing the preferences of users A to C for the recipe "soybeans" is stored in the storage unit 12. The preference information A to C includes items such as "hardness" and the intensity of the preferences that users A to C have for each item. For example, preference information A includes the following: "hardness -1.5," "sweetness +1," "moistness +0.5," and "smoothness -0.5." The intensity of preference is a human sensory evaluation of each item. For example, the sensory evaluation for the item "hardness" ranges from +3 to -3, with the closer the sensory evaluation to +3, the harder the cooked food is, and the closer the sensory evaluation to -3, the softer the cooked food is. The sensory evaluation for the item "sweetness" ranges from +3 to -3, with the closer the sensory evaluation to +3, the sweeter the cooked food is, and the closer the sensory evaluation to -3, the bitterer and astringent the cooked food is. The sensory evaluation for the item "moistness" is given from +3 to -3, with the closer the sensory evaluation to +3 the moister the cooked food is, and the closer it is to -3 the drier the cooked food is. The sensory evaluation for the item "smoothness" is given from +3 to -3, with the closer the sensory evaluation to +3 the smoother the cooked food is (i.e., it melts easily in the mouth), and the closer it is to -3 the rougher the cooked food is.

[0115] As an example, the storage unit 12 may further store a comparison table between preference information and changes to odor standards or parameters. For example, the comparison table includes changes to the odor pattern "hardness" corresponding to preference information "hardness -3," "hardness -2," "hardness -1," "hardness +1," "hardness +2," and "hardness +3." As an example, the change corresponding to preference information "hardness -1" is a value of the odor standard or parameter set so that the sensory evaluation of the "hardness" item for the finished dish becomes "-1." For example, if the "hardness" estimated based on the concentration of the detected aroma components is "-3," it is possible to reduce the temperature inside the pot, the heating time, or the pressure inside the pot so that the final hardness is raised by two levels to "-1." Thus, the change corresponding to "hardness -1" may be, for example, "set the temperature inside the pot to 125°C," "set the heating time to 20 minutes," or "set the pressure inside the pot to 0.25 MPa." As another example, the change may be an adjustment to an odor standard or parameter. For example, the change corresponding to "hardness -1" may be, "reduce the temperature inside the pot by 5°C," "reduce the heating time by 5 minutes," or "reduce the pressure inside the pot by 0.2 MPa."

[0116] As an example, the odor criteria are set based on preference information and a standard control sequence. In this example, the standard control sequence is universal and is not changed based on preference information during cooking. On the other hand, the odor criteria or default settings of parameter standard values ​​based on the standard control sequence can be changed based on preference information. The control unit 14 may change the odor criteria based on the preference information and then set the parameters of the cooking operation based on the changed odor criteria.

[0117] The odor criteria can be changed by the control unit 14, the server 30, or a user of the server 30. For example, the server 30 or a user of the server 30 may change the odor criteria using a trained model that has learned the "relationship between cooking conditions and sensory evaluation" or the "relationship between cooking conditions and aroma components and the concentrations of those aroma components," which will be described later.

[0118] FIG. 9 is an example of a graph showing the relationship between the control sequence for the "soybean" recipe and the odor detected from the food. Like FIG. 5, FIG. 9 also shows graphs showing the time-dependent changes in the temperature inside the pot, the pressure inside the pot, and the concentration of ingredients when the standard control sequence for the "soybean" recipe is executed. The bottom graph of ingredient concentration in FIG. 9 also shows the time-dependent changes in the same type of odor as in FIG. 5. In the example shown in FIG. 9, for example, preference information "sweetness -3" is entered for the "soybean" recipe. The control unit 14 compares the preference information with the comparison table and obtains information about the change "reduce the heating time at 40°C by 20 minutes." If the standard control sequence specifies a heating time at 40°C of approximately 18 minutes, the control unit 14 changes the heating time at 40°C to zero. In other words, the low-temperature cooking process for increasing sweetness is omitted, and when automatic cooking begins, the food is heated to 120°C (time t8). This reduces the "sweetness" of the food, resulting in a food that matches the user's preferences.

[0119] FIG. 10 is an example of a graph showing the relationship between the control sequence for the "soybean" recipe and the odor detected from the food being cooked. Like FIG. 5, FIG. 10 also shows graphs showing the time-dependent changes in the temperature inside the pot, the pressure inside the pot, and the concentration of ingredients when the standard control sequence for the "soybean" recipe is executed. The graph showing the concentration of ingredients at the bottom of the figure also shows the same type of odor change over time as in FIG. 5. In the example shown in FIG. 10, for example, preference information of "moistness +3" and "smoothness +1" is input for the "soybean" recipe, resulting in a "good texture." The control unit 14 compares the preference information with the comparison table and obtains information regarding the change "increase the pressure inside the pot to 5 MPa." Although the standard control sequence specifies the pressure inside the pot to be 0.2 MPa and 0.3 MPa, the control unit 14 automatically sets the pressure inside the pot to approximately 5 MPa based on the preference information and performs automatic cooking. By cooking under pressure of 5 MPa, the starch properties of the food change, giving it a smooth and moist texture, and creating a food that meets the user's tastes.

[0120] For example, the standard control sequence is generated with the aim of creating a delicious dish that will appeal to everyone. On the other hand, the automatic cooker 10 or the server 30 changes the odor standard, parameter standard values, or parameters during automatic cooking based on the preference information so as to match the user's preferences.

[0121] This completes the process based on the preference information. In this way, it is possible to prepare food that has a texture, smell, or taste that matches the user's preferences.

[0122] Third Embodiment Setting by Machine Learning In the third embodiment, the standard control sequence, odor criteria, or standard parameter values ​​are set for a recipe based on a trained model obtained by machine learning. More specifically, the trained model learns the relationships between cooking conditions, sensory evaluation, and aroma components through machine learning, and the standard control sequence, odor criteria, or standard parameter values ​​can be set using the trained model. In this way, it is possible to perform automatic cooking that reflects the relationships between the cooking state, sensory evaluation, and aroma components, and to control the texture, odor, and taste of the cooked food.

[0123] The machine learning of the relationships among cooking conditions, sensory evaluations, and aroma components may be performed separately for learning the relationship between cooking conditions and sensory evaluations, learning the relationship between cooking conditions and aroma components, and learning the relationship between sensory evaluations and aroma components. Alternatively, two or more relationships may be learned together. As a result, trained models corresponding to the three relationships may be generated individually, or trained models that train two or three relationships together, such as a trained model that combines multiple trained models into a single data set, may be generated.

[0124] Since each recipe has different ingredients and cooking actions, and each recipe has its own unique smell that humans perceive from the cooked food, in the present disclosure, machine learning is performed for each recipe to generate a trained model. For example, for one recipe, trained models corresponding to the relationships between the three items described above may be generated, or a single trained model that learns all of the relationships may be generated.

[0125] Hereinafter, we will explain the machine learning of the relationships between these three items and their use, but it is possible to set a standard control sequence or perform automatic cooking without learning all of these relationships. For example, as described in the first embodiment, automatic cooking can be performed so as to control the texture of the food by using only a trained model that has learned the relationship between sensory evaluation and aroma components.

[0126] <Example of Recipe "Soybeans"> Hereinafter, machine learning according to the present disclosure will be described with reference to FIGS. 11 to 14, using the recipe "soybeans" cooked by an auto cooker as an example.

[0127] <Relationship between Cooking Conditions and Sensory Evaluation> Input data used in machine learning of the relationship between cooking conditions and sensory evaluation includes training data of cooking conditions and training data of sensory evaluation. Training data is also called a machine learning dataset. The training data of cooking conditions includes multiple different cooking conditions.

[0128] The cooking conditions include at least one cooking operation. Figure 11 shows an example of learning data for cooking conditions for the recipe "soybeans." For example, cooking condition A2 is "heat at 200°C for 25 minutes," and cooking condition A5 is "heat at 130°C and a gauge pressure of 0.2 MPa for 5 minutes."

[0129] As an example, cooking conditions include at least one cooking operation and the ingredients or food to be cooked that are the subject of the cooking operation. For example, cooking conditions for soybeans that are the subject of a cooking operation may include "dried soybeans" or "soaked soybeans." In this case, "dried soybeans" may be combined with cooking condition A3 "heating at 200°C for 40 minutes" to form a single cooking condition, "heating dried soybeans at 200°C for 40 minutes." Alternatively, "soaked soybeans" may be combined with cooking condition A2 "heating at 200°C for 25 minutes" to form a single cooking condition, "heating soaked soybeans at 200°C for 25 minutes."

[0130] The sensory evaluation training data includes sensory evaluations of the cooked dishes cooked under each of the cooking conditions described above. FIG. 12 is an example of sensory evaluation training data for the recipe "soybeans" cooked under different cooking conditions. The sensory evaluation training data shown in FIG. 12 is a human sensory evaluation of the odor pattern items "hardness," "moistness," "smoothness," and "overall evaluation" for soybeans cooked under cooking conditions A1 to A13 shown in FIG. 11. For example, the sensory evaluation for the item "hardness" is rated from +3 to -3, with the closer the sensory evaluation to +3, the harder the cooked dish is, and the closer it is to -3, the softer the cooked dish is. The sensory evaluation for the item "moistness" is rated from +3 to -3, with the closer the sensory evaluation to +3, the moister the cooked dish is, and the closer it is to -3, the drier the cooked dish is. The sensory evaluation for the item "smoothness" is rated from +3 to -3, with the closer the sensory evaluation to +3, the smoother the cooked dish is (i.e., melts easily in the mouth), and the closer it is to -3, the rougher the cooked dish is. By using sensory evaluation as labels and adding them to cooking conditions through machine learning, and then performing regression analysis, a trained model that has learned the relationship between cooking conditions and sensory evaluation can be obtained.

[0131] If machine learning is performed using sensory evaluations that include items related to the texture of cooked foods, the trained model will have learned the relationship between texture and cooking conditions.If machine learning is performed using sensory evaluations of the smell, taste, and texture of cooked foods that humans perceive, the trained model will have learned the relationship between smell, taste, texture, and cooking conditions.

[0132] A trained model that has learned the relationship between cooking conditions and sensory evaluations can estimate (predict) the sensory evaluation of a food cooked under specific cooking conditions when those cooking conditions are input. Furthermore, when a specific sensory evaluation is input, the trained model can estimate the cooking conditions required to obtain that specific sensory evaluation. In other words, the trained model can be used to estimate how food should be cooked to obtain a specific texture.

[0133] Therefore, a standard control sequence, odor criteria, or parameter standard values ​​can be set to control texture, odor, or taste based on the trained model. For example, to obtain a texture of "hardness ±0" for the recipe "soybeans," cooking conditions including the cooking operation of "heating at 40°C for 20 minutes" are estimated. By incorporating this cooking operation into the standard control sequence, the server 30 or the like can generate a standard control sequence corresponding to the texture of "hardness ±0." When the automatic cooker 10 cooks soybeans according to this standard control sequence, it is believed that a cooked product with a texture of "hardness ±0" will be obtained.

[0134] <Relationship Between Cooking Conditions and Aroma Components> Input data used in machine learning of the relationship between cooking conditions and aroma components includes training data of cooking conditions and training data of aroma components. As described above, the training data of cooking conditions includes a plurality of different cooking conditions. As one example, the training data of aroma components includes aroma components detected from foods cooked under each cooking condition. As another example, the training data of aroma components includes aroma components detected from foods cooked under each cooking condition and the concentrations of the aroma components.

[0135] FIG. 13 shows an example of training data of aroma components detected from food cooked under different cooking conditions for the recipe "soybeans." In the cooking experiment shown in FIG. 13, the aroma components "pentane / isobutane" with a concentration of "0.013" were detected from food cooked under cooking condition A1. Furthermore, the aroma components "butanal / 2-methylbutanal / 3-methylbutanal" with a concentration of "0.107" were detected from food cooked under cooking condition A2. By adding the detected aroma components as labels to the cooking conditions and conducting machine learning, a trained model that has learned the relationship between the aroma components and sensory evaluation can be obtained. By adding the detected aroma components and their concentrations as labels to the cooking conditions and conducting machine learning, a trained model that has learned the relationship between the aroma components, their concentrations, and sensory evaluation can be obtained.

[0136] If machine learning is performed on aroma components related to the texture of cooked foods, the trained model will have learned the relationship between aroma components related to texture and cooking conditions.If machine learning is performed on aroma components related to texture, smell, and taste odor patterns, the trained model will have learned the relationship between aroma components related to texture, smell, and taste and cooking conditions.

[0137] A trained model that has learned the relationship between cooking conditions and aroma components and their concentrations can, when specific cooking conditions are input, estimate the aroma components and their concentrations detectable in food cooked under those cooking conditions. Furthermore, when a specific aroma component and its concentration are input, the trained model can estimate the cooking conditions required to obtain that specific aroma component and its concentration. In other words, the trained model can be used to estimate how cooking should be done to obtain a specific aroma component or a specific concentration of an aroma component.

[0138] Therefore, a standard control sequence, odor criteria, or parameter standard values ​​can be set based on the trained model to control texture, odor, or taste. For example, to increase the aroma component "thiopene" that affects the texture "smoothness" of a recipe for "soybeans," cooking conditions including two cooking operations, "heating at 120°C" and "heating at 120°C followed by heating at 130°C," are estimated. The server 30 or the like can generate a standard control sequence that increases the aroma component "thiopene" by incorporating these cooking operations into the standard control sequence. It is believed that when the automatic cooker 10 cooks soybeans according to the standard control sequence, a cooked product that emits more of the aroma component "thiopene" will be obtained.

[0139] <Relationship between Sensory Evaluation and Aroma Components> Input data used in machine learning of the relationship between sensory evaluation and aroma components includes training data for sensory evaluation and training data for aroma components. As described above, the training data for sensory evaluation is the sensory evaluation of the dishes cooked under each of the cooking conditions described above. The training data for aroma components may include the aroma components detected from the dishes cooked under each of the cooking conditions and the concentrations of the aroma components, and may also include the concentrations of the aroma components.

[0140] The machine learning of the relationship between sensory evaluation and aroma components, and the trained model that has learned this relationship, have been described together with the odor sensor 13. As described above, if machine learning is performed using sensory evaluation of the smell, taste, and texture of a cooked dish, the trained model will have learned the relationship between the sensory evaluation of the smell, taste, and texture and the aroma components (and their concentrations).

[0141] When a specific sensory evaluation is input, the trained model can estimate the aroma components required to obtain the specific sensory evaluation, or can estimate the aroma components and their concentrations. Furthermore, when a specific aroma component (and its concentration) is input, the trained model can estimate the sensory evaluation perceived from the specific aroma component and its concentration.

[0142] Figure 14 shows an example of the relationship between aroma components and sensory evaluation for the recipe "soybean," and can also be considered output data of a trained model that has learned this relationship. The values ​​in the table in Figure 14 are regression coefficients when aroma components and sensory evaluation are correlated, and the bottom column shows the prediction accuracy R ２ The prediction accuracy R for texture "hardness," "moistness," and "smoothness" is ２ is 0.5 or more, which indicates that the estimation (prediction) results by the trained model are reliable.

[0143] Therefore, based on the trained model, it is possible to identify, from among a vast number of aroma components, those associated with a particular texture, smell, or taste. The trained model can be used for or together with the odor sensor 13.

[0144] In summary, the above-described machine learning uses input data including different cooking conditions, sensory evaluations of the dishes cooked under the cooking conditions, and aroma compounds detected from the dishes and their concentrations. The trained model generated by such machine learning learns the relationship between the cooking conditions, the sensory evaluations, and the detected aroma compounds and their concentrations. The trained model can estimate the remaining item based on two items from among the cooking conditions, the sensory evaluations, and the aroma compounds.

[0145] <Example of Recipe "Bread"> Hereinafter, machine learning according to the present disclosure will be described with reference to FIGS. 15 to 18, using the recipe "bread" cooked by a home bakery as an example.

[0146] 15 shows an example of learning data of cooking conditions for the recipe "bread." For example, cooking condition B1 is "stir for 40 minutes, then raise the dough temperature to 34°C, let it rise once for 45 minutes, and then heat for 30 minutes." Also, cooking condition B6 is "stir for 109 minutes, then raise the dough temperature to 34°C, let it rise twice for 157 minutes, and then heat for 40 minutes."

[0147] FIG. 16 is an example of training data of sensory evaluations of food items cooked under different cooking conditions for the recipe "bread." The numerical values ​​in the table of FIG. 16 are sensory evaluations of food items cooked under each cooking condition. A positive sensory evaluation indicates a positive evaluation of the smell pattern item, i.e., a delicious feeling. A negative sensory evaluation indicates a negative evaluation of the smell pattern item, i.e., a less delicious feeling. The sensory evaluation indicates the degree of the impression. For example, it is shown that a person who performed the sensory evaluation perceived bread cooked under cooking condition B6 to be highly delicious in the items "chewy," "fragrant," and "taste."

[0148] The server 30 can perform machine learning using training data such as those shown in FIGS. 15 and 16 . Furthermore, training data containing aroma components detected in dishes cooked under cooking conditions B1 to B8 in an experiment for the "bread" recipe can also be used as input data. FIG. 17 shows an example of training data of aroma components detected in dishes cooked under different cooking conditions for the "bread" recipe. The example in FIG. 17 shows the amounts of the aroma components "detector A_2-methyl-1-butanol / detector A_3-methyl-1-butanol," "detector B_2-methyl-1-butanol / detector B_3-methyl-1-butanol," and "acetoin" detected in dishes cooked under cooking conditions B1 to B8 in an experiment. For example, in an experiment, the aroma component "acetoin" with a concentration of "2393" was detected in dishes cooked under cooking condition B1. Using this training data, the server 30 performs machine learning of the relationships between cooking conditions, sensory evaluations, and aroma components for the recipe "bread" and generates a trained model corresponding to these relationships. Note that detectors A and B are different analyzers.

[0149] Figure 18 shows an example of the relationship between food components, including aroma components, and sensory evaluation for the recipe "bread," and can also be considered output data of a trained model that has learned this relationship. Of the four food components shown in Figure 18, "glutamic acid" is a flavor component, and the remaining three are aroma components. ★ indicates the degree to which each food component influences the sensory experience, with × indicating a strong negative impact and ◎ indicating a strong positive impact. Figure 18 also shows how and to what extent several aroma components affect the "chewy" texture. For example, the aroma components "detector A_2-methyl-1-butanol / detector A_3-methyl-1-butanol" have a strong negative impact on the "chewy" texture, while the umami component "glutamic acid" has a positive impact.

[0150] As described above, texture, odor, or taste control can be incorporated into a standard control sequence. Furthermore, by incorporating appropriate cooking operations into the standard control sequence based on the trained model, a standard control sequence that increases or decreases specific aroma components or texture can be generated. For the generated standard control sequence, the timing at which cooking operation parameters change, the values ​​of the parameters before and after the change, and the relationship between the concentrations of multiple aroma components at that timing can be analyzed to set odor standards and parameter standard values ​​corresponding to the standard control sequence.

[0151] As an example, a base sequence for a particular recipe is defined by a human. The server 30 automatically generates a standard control sequence by incorporating appropriate cooking actions into the base sequence using the trained model. When generating the standard control sequence, the server 30 also automatically generates odor criteria and standard parameter values ​​corresponding to the standard control sequence. In this case, the server 30 can automatically output the standard control sequence, odor criteria, and standard parameter values ​​for each recipe.

[0152] As another example, a user of the server 30 may perform various estimations using the trained model and manually generate a standard control sequence, odor criteria, and standard parameter values ​​based on the estimation results. For example, the user of the server 30 may make adjustments to the standard control sequence, odor criteria, and standard parameter values ​​generated by the server 30 before finalizing them.

[0153] In this way, a trained model can be generated that has learned the relationship between cooking conditions, sensory evaluation, and aroma components, and the trained model can be used to set a standard control sequence, odor standard, or parameter standard value that can control the texture of the cooked food, etc.

[0154] <Fourth Embodiment> <Adding or Updating Data> In the fourth embodiment, data used for controlling and setting automatic cooking can be added or updated. By adding or updating data, more dishes can be cooked, or the food can be cooked to suit the user's preferences more closely.

[0155] FIG. 19 is a flowchart of an example of an automatic cooking method executed by the automatic cooker 10 of the fourth embodiment.

[0156] 19, the automatic cooker 10 of the fourth embodiment executes steps S100 to S300 and then executes step S400. Steps S100 to S300 are the same as steps S100 to S300 of the first embodiment in FIG. 2, and detailed description thereof will be omitted.

[0157] The automatic cooker 10 receives data for adding or updating recipes, standard control sequences, odor standards, or preference information (hereinafter referred to as additional data or updated data) via the communication unit 16 (step S500). The control unit 14 stores the received additional data and updated data in the storage unit 12 and uses them for automatic cooking.

[0158] As an example, the automatic cooker 10 downloads a new recipe and the trained model, standard control sequence, odor standard, and standard parameter values ​​corresponding to the new recipe from the server 30. Then, the automatic cooker 10 can perform automatic cooking according to the new recipe.

[0159] As another example, the automatic cooker 10 downloads from the server 30 at least one of a trained model, a standard control sequence, an odor standard, and standard parameter values ​​corresponding to a recipe already stored in the memory unit 12. The control unit 14 can use the downloaded data to update the automatic cooking control of the already stored recipe.

[0160] Receiving additional or updated data can be performed when automated cooking is running or not. Updating recipes, standard control sequences, odor criteria, or preference information can be performed when automated cooking is not running.

[0161] This completes the process of adding or updating automatic cooking data. Adding data allows you to cook more delicious dishes or more dishes. Updating data also allows you to cook dishes that more closely match the user's preferences.

[0162] It should be noted that the above-described embodiments are merely specific embodiments of the present disclosure, and the scope of protection of the present disclosure is not limited thereto. The present disclosure includes the contents described above in the drawings and the specific embodiments described above, but the present disclosure is not limited thereto. Modifications that do not deviate from the functional and structural principles of the present disclosure are within the scope of the claims. Furthermore, in each of the above-described embodiments or some examples thereof, further additional technical features and the effects brought by these technical features are described.

[0163] REFERENCE SIGNS LIST 10 Automatic cooker 11 Cooking section 12 Memory section 13 Smell sensor 14 Control section 15 Input section 16 Communication section 17 Air outlet 18 Intake port 19 Fan 111 Heating element 112 Rotating element 112a Motor 112b Rotating shaft 113 Steam generating element 114 Container (cooking pot) 115 Stirring element 116 Cutting / kneading element

Claims

1. An automatic cooker including: a cooking unit that cooks an item; a memory unit that stores odor standards related to the texture of the item; an odor sensor that detects odor information related to the odor of the item; and a control unit, wherein after cooking begins, the control unit: acquires the odor information using the odor sensor; sets parameters for the cooking operation of the cooking unit so as to change the texture of the item based on the odor standards and the odor information; and controls the cooking unit using the parameters to cook the item.

2. The automatic cooker of claim 1, wherein the odor information includes aroma components that form the odor of the cooked food and the concentrations of the aroma components, and the odor standard includes aroma components associated with a specific texture and the concentrations of the aroma components.

3. The automatic cooker of claim 1 or 2, wherein the memory unit further stores a standard control sequence corresponding to a specific recipe, and the control unit sets the parameters so as to change the texture of the food based on the odor standard, the odor information, and the standard control sequence.

4. The automatic cooker according to claim 3, wherein the odor standard includes an aroma component associated with a specific texture and includes a concentration of the aroma component at a timing when a parameter of a cooking operation is changed in the standard control sequence.

5. The automatic cooker according to any one of claims 1 to 4, wherein the parameters include at least one of the amount of heat applied by a heating element that heats the food, the heating temperature applied by the heating element, the heating time applied by the heating element, the pressure inside the automatic cooker, the rotation speed of a rotating element that stirs the food in the cooking section, the rotation time of the rotating element, whether or not a steam generating element for heating or humidifying the food with steam supplies steam to the food, the amount of steam supplied by the steam generating element to the food, and the supply time of the steam by the steam generating element.

6. An automatic cooking appliance as described in any one of claims 1 to 5, wherein the memory unit further stores a parameter standard value corresponding to the odor standard, and the control unit compares the odor information with the odor standard, and if it determines that the odor information satisfies the odor standard, sets the parameters or terminates cooking based on the parameter standard value corresponding to the odor standard.

7. The automatic cooker according to any one of claims 1 to 6, wherein the memory unit further stores preference information representing a user's preferences for food texture, and the odor standard is set based on the preference information.

8. The automatic cooker according to claim 7, wherein the odor standard includes aroma components related to the texture represented by the preference information and concentrations of the aroma components.

9. The automatic cooker according to claim 7 or 8, further comprising a communication unit or an input unit, and the control unit receives the preference information via the communication unit or the input unit.

10. The automatic cooker according to claim 4, wherein the standard control sequence, the odor standard, or a standard parameter value corresponding to the odor standard is set for the recipe based on a trained model obtained by machine learning.

11. The automatic cooker of claim 10, wherein the trained model learns the relationship between the cooking conditions, the sensory evaluation, and the detected aroma components and concentrations of the aroma components using input data including different cooking conditions, sensory evaluations of food cooked under the cooking conditions, and aroma components detected from the food and concentrations of the aroma components, and the cooking conditions include at least one cooking operation.

12. The automatic cooker of claim 11, wherein the trained model, when specific cooking conditions are input, estimates a sensory evaluation of a dish cooked under the cooking conditions, or aroma components and the concentrations of the aroma components detectable in the dish, when a specific sensory evaluation is input, estimates cooking conditions or aroma components and the concentrations of the aroma components for obtaining the specific sensory evaluation, or, when a specific aroma component and its concentration are input, estimates cooking conditions for obtaining the specific aroma component and its concentration, or a sensory evaluation perceived from the specific aroma component and its concentration.

13. The automatic cooker according to any one of claims 1 to 12, wherein the odor standard includes a concentration of at least one of methylbutanol, acetoin, and glutamic acid.

14. The automatic cooker according to any one of claims 1 to 13, further comprising a communication unit, and wherein the control unit receives data for adding or updating a standard control sequence, an odor standard, or preference information via the communication unit.

15. An automatic cooking method for automatically cooking an ingredient using an automatic cooker, comprising the steps of: acquiring odor information regarding the odor of the ingredient using an odor sensor; setting parameters for the cooking operation of the automatic cooker so as to change the texture of the ingredient based on odor standards regarding the texture of the ingredient and the odor information; and controlling the automatic cooker using the parameters to cook the ingredient.

16. A computer program for causing an automatic cooker to execute the automatic cooking method according to claim 15.

17. A non-transitory computer-readable storage medium having a computer program stored thereon, the computer program implementing the automatic cooking method of claim 15 when executed by a processor.

Images