KR20220020086A - Quality prediction system for finished coffee products - Google Patents

Abstract

An objective of the present invention is to provide a quality prediction system for a complete product of coffee, capable of predicting quality of complete products of coffee that vary depending on a mixing ratio and roasting degree of a plurality of raw green coffee beans. According to an embodiment of the present invention, the quality prediction system for a complete product of coffee comprises: a green bean database that stores an evaluation result of raw green coffee beans; a single origin database that stores results of characteristics for each roasting degree of the raw green coffee beans based on an evaluation table for a plurality of characteristics of the raw green coffee beans; and a product characteristic prediction module that derives a quality prediction value of a complete coffee product according to a blending ratio between a plurality of the raw coffee beans and a roasting degree of each of the raw coffee beans based on the single origin database.

Description

Quality prediction system for finished coffee products

The present invention relates to a system for predicting the quality of finished coffee products according to the blending ratio of raw green coffee beans and the characteristics of each roasting degree.

Coffee is a beverage that is consumed by harvesting green beans from a coffee tree, roasting them through a processing process, and extracting them. In general, coffee can be divided into single origin roasted coffee beans from a single origin and blended coffee blended with green coffee beans from multiple origins. In the case of single origin coffee, it is easy to predict the quality of the final product because the characteristics of raw coffee beans are distinct depending on the country and region of production, processing method, variety, and grade. It is difficult to predict the quality of a product. In addition, the product characteristics of coffee beans vary greatly depending on the degree of roasting, which is a processing method, in addition to raw raw coffee beans. In general, the degree to which raw raw coffee beans are roasted is called roasting. These roasted coffee beans are called roasted coffee beans, and various processed products are manufactured using them. Since green coffee beans have a unique aroma and taste depending on the degree of roasting, even the same coffee beans have different characteristics depending on the degree of roasting.

Currently, the industry uses subjective criteria such as light, medium, and dark to express the degree of roasting rather than numerically clear values. It is not easy to predict the quality of the final product before roasting due to the diversity of raw raw coffee beans and unclear standards of roasting degree.

In addition, to check coffee quality, generally check the screen size, moisture content, and density of raw raw coffee beans, or use the Cupping Form used by the Specialty Coffee Association (SCA) after roasting to obtain Fragrance (the flavor of ground coffee). /Aroma (flavor of wet coffee), Flavor (taste), aftertaste, acidity, texture, taste uniformity, and blemishes are evaluated in a total of 10 items. But. In the case of SCA cupping foam, it is an evaluation table used by Q-graders with expertise to evaluate specialty green coffee beans. do.

The SCA cupping foam used for cupping, which is a conventional coffee evaluation method, is used for the purpose of evaluating the quality of specialty green beans by coffee bean testers named Q Grader. It's not an evaluation form for coffee. In addition, cupping is tested by slurping in which coffee and air are inhaled together and sprayed all over the mouth. Coffee beans for cupping test are generally roasted light, so products produced differently from the roasting degree of the product to be produced are tested. It may differ from the actual sensory evaluation results felt when drinking. In addition, since SCA cupping foam uses technical terms and there is no conceptual explanation for it, there is a problem that it is difficult to understand and feels unfamiliar to general consumers.

It is an object of the present invention to provide a system for predicting the quality of a finished coffee product that can predict the quality of a finished coffee product that varies depending on a blending ratio and roasting degree of a plurality of raw green coffee beans.

It provides a system for predicting the quality of finished coffee products according to an embodiment of the present invention. The system for predicting the quality of finished coffee products includes a green coffee database for storing evaluation results of raw raw coffee beans, a single origin database for storing results of characteristics for each roasting degree of raw raw coffee beans based on an evaluation table for a plurality of characteristics of the raw raw coffee beans; and a product characteristic prediction module for deriving a quality prediction value of a finished coffee product according to a blending ratio between a plurality of green raw coffee beans and a roasting degree of each of the raw raw coffee beans, based on the single origin database.

In one example, the evaluation table includes a preference item, an intensity item, a fragrance item, and a sour taste type item.

As an example, the predicted quality value is determined as the sum of values obtained by multiplying the blending ratio by the score for each item according to the roasting degree setting for each raw green coffee bean to be blended.

In one example, the product characteristic prediction module derives the quality prediction value based on the type, roasting degree, and blending ratio of the raw green coffee beans to be blended determined by the user, and the blended raw raw coffee beans include first raw green coffee and a second raw coffee bean, wherein the quality predicted value is a first sub-prediction value obtained by multiplying the score for each item according to the roasting degree setting of the first raw green coffee bean by a blending ratio, and a score for each item according to the roasting degree setting of the second raw green coffee bean It is determined as a value obtained by adding the second sub-prediction value multiplied by the blending ratio.

According to an example, the quality prediction value is expressed as a score for each item of the evaluation table.

According to an example, a product database storing the actual quality value of the finished coffee product, and a comparison module for comparing the actual quality value stored in the product database with the predicted quality value, wherein the comparison module comprises the actual quality value and the quality The reliability of the quality prediction value is evaluated based on the difference between the prediction values.

As an example, when the quality prediction value evaluated by the comparison module is not included in the stunned range, the comparison module recognizes the quality prediction value as erroneous data.

In one example, the comparison module updates the single origin database based on a result of comparing the predicted quality value with the actual quality value and the predicted quality value, not error data.

According to an example, the comparison module stores the new data derived by adding the difference between the predicted quality value and the actual quality value to the score for each item of the raw green coffee beans in the single origin database, and the single origin database is An average value is derived based on the scores for each item according to the roasting degree of the raw green coffee beans and the new data, and the derived average value is determined as the score for each item according to the roasting degree of the raw green coffee beans.

In one example, the green coffee database stores information on production area evaluation, processing method evaluation, moisture content evaluation, number of defects, grade evaluation according to screen size, cupping evaluation, and unit price confirmation.

In one example, the product characteristic prediction module derives the manufacturing cost of the finished coffee product according to a blending ratio of the plurality of raw green coffee beans based on data stored in the green coffee database.

According to an embodiment of the present invention, since the quality evaluation method for evaluating the actual sensory for general consumers is used, the actual sensory result of the product is reflected, and the cost of the final product before the product is produced without the product producer going through the process of roasting, Characteristics and preferences can be predicted.

According to an embodiment of the present invention, the single origin database can be periodically updated by comparing the quality predicted value with the actual quality value, thereby confirming the change in consumer preference for each single origin and blended coffee and understanding the coffee bean market trend. can In addition, as the single origin database is updated based on new data, the reliability of the quality prediction value derived by the product characteristic prediction module may be improved.

1 is a block diagram illustrating a system for predicting quality of finished coffee products according to an embodiment of the present invention.
2 is an evaluation table for evaluating the characteristics of raw green coffee beans after roasting according to an embodiment of the present invention.
3 is a diagram illustrating characteristics of raw green coffee beans stored in a single origin database according to an exemplary embodiment of the present invention by roasting degree.
4 is a diagram illustrating a quality predicted value by a product characteristic prediction module according to an embodiment of the present invention.
5 is a view showing the actual quality value of the coffee beans product stored in the product database according to an embodiment of the present invention.
6 is a flowchart illustrating a method for predicting the quality of a finished coffee product according to an embodiment of the present invention.

Advantages and features of the present invention, and a method for achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only this embodiment serves to complete the disclosure of the present invention, and to obtain common knowledge in the technical field to which the present invention pertains. It is provided to fully inform the possessor of the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numerals refer to like elements throughout.

Terms such as "...unit", "...unit", "...module", etc. described in the specification mean a unit that processes at least one function or operation, which is hardware or software or hardware and software. It can be implemented as a combination.

In addition, the reason that the names of the components are divided into the first, the second, etc. in the present specification is to distinguish the names of the components in the same relationship, and the order is not necessarily limited in the following description.

The detailed description is illustrative of the invention. In addition, the above description shows and describes preferred embodiments of the present invention, and the present invention can be used in various other combinations, modifications, and environments. That is, changes or modifications are possible within the scope of the concept of the invention disclosed herein, the scope equivalent to the described disclosure, and/or within the scope of skill or knowledge in the art. The described embodiment describes the best state for implementing the technical idea of the present invention, and various changes required in specific application fields and uses of the present invention are possible. Accordingly, the detailed description of the present invention is not intended to limit the present invention to the disclosed embodiments. Also, the appended claims should be construed to include other embodiments.

1 is a block diagram illustrating a system for predicting the quality of a finished coffee product according to an embodiment of the present invention, and FIG. 2 is an evaluation table for evaluating characteristics of raw green coffee beans according to an embodiment of the present invention after roasting.

1 and 2, the coffee finished product quality prediction system 1 includes a green coffee database 100, a single origin database 200, a product characteristic prediction module 300, a product evaluation module 400, and a product database ( 500) and a comparison module 600 . The coffee finished product quality prediction system 1 can predict the characteristics of the final extracted coffee product when raw green coffee beans with different characteristics for each production region are blended, and the coffee derived through the predicted characteristics and actual sensory evaluation The single origin database 200 may be updated by comparing characteristics of finished products.

The green coffee database 100 may store the result of evaluating raw green coffee beans. The result of evaluating raw green coffee beans may include information related to the production location, processing method, moisture content, number of defects and screen size-based grades, cupping evaluation results, and cost of the received green coffee beans. The number of defects determines how many green coffee beans have defects between them, and the screen size may mean the size of the green coffee beans. The cupping evaluation result may mean the quality of green coffee beans evaluated by the existing SCA cupping evaluation method.

The single origin database 200 may store data in which characteristics of a plurality of raw green coffee beans, which vary according to the roasting degree, are scored based on the evaluation table 50 . A plurality of scores for one raw green coffee may exist according to a plurality of evaluations, and the single origin database 200 may store an average value of the plurality of scores. After roasting, the roasting degree can be divided into specified numerical values using a device (Color Track R-100B Laser Color-Analyzer, Fresh Roast Systems, Inc., USA).

The evaluation table 50 may be divided into large items of a preference item, an intensity item, a fragrance item, and a sour taste item. Specifically, the evaluation table 50 includes 8 preference items (fragrance, sour taste, bitter taste, softness, body feeling, aftertaste, balance, overall taste), 2 intensity items (sour taste, bitter taste), fragrance type (flower flavor, fruit flavor) , sweet, chocolate, nutty, grassy, etc.) and sour taste (sour, fresh, fresh and astringent, bitter, soft and sweet, sour, weak acidity). Sour taste and bitter taste may be included in both the preference and intensity items. All these sensory results can be displayed on a 10-point scale, and the scored data can be displayed to the user. A point of 1 means a score given when a negative preference is felt in the sub-item, and a score of 10 means a score given when a positive preference is felt in the sub-item. The score for each item of the evaluation table 50 can be divided into three sections and displayed in different colors, thereby facilitating the identification of scores for each score.

The intensity is divided into sour and bitter, and evaluated on a 10-point scale. If the intensity of each item is absent or weak, 1 point can be given, and when the intensity is felt strong, 10 points can be given. The intensity item was divided into three sections and color was given to give ease of classification.

Fragrance can be divided into a total of 7 sub-items: floral, fruity, sweet, chocolate, nutty, grassy, and others. For example, the floral scent may include rose scent, jasmine scent, and acacia scent. The fruit flavor may include a berry flavor, a citrus flavor, a raisin flavor, a peach flavor, a grape flavor, and an apple flavor. Chocolate flavor may include milk chocolate flavor, dark chocolate flavor, and cacao flavor. The nutty flavor may include a peanut flavor, a walnut flavor, an almond flavor, and a hazelnut flavor. Other flavors may include cinnamon flavor, pepper flavor, spice flavor, grain flavor, burnt flavor, rubber flavor, earth flavor, wood flavor, and the like. The evaluator checks the flavor characteristics felt when drinking the sample, and then marks it in the detailed items.

Sour taste (sour taste in the preference category) is a major item that distinguishes the types of sour taste felt when sensuality. It can be divided into seven subcategories: soft and sweet, sour such as vinegar and acetic acid, and weak acidity with little or no acidity.

Unlike the above-described example, evaluation items of the evaluation table 50 may not be particularly limited.

In addition to the sensory evaluation of the evaluator, a score for each item of the evaluation table 50 may be given by using a measuring device.

For example, by performing quantitative analysis of aroma and acidity through a gas chromatograph-mass spectrometer (GC-MS), the main aromatic components for each functional group, such as Floral, Fruity, Caramelic, Chocolate, Nutty, Woody, Bready, etc. can be classified as The classification result of fragrance and acidity may match flower fragrance, fruit fragrance, sweet fragrance, chocolate fragrance, nutty fragrance, grassy fragrance, etc. on the evaluation table 50 .

As an example, 8 preference items, 2 intensity items, fragrance types, and sour taste types may be classified using an electronic tongue and coffee acidity component analysis results. Scores may be assigned to the classified characteristics of raw green coffee beans according to items in the evaluation table 50 , and data on characteristics of the scored raw green coffee beans may be stored in the single origin database 200 .

The product characteristic prediction module 300 may derive a quality prediction value of the finished coffee product based on the data stored in the single origin database 200 . The product characteristic prediction module 300 may derive a quality prediction value that varies according to the mixing ratio between the plurality of raw green coffee beans and the setting of the roasting degree of each of the raw green coffee beans. The product characteristic prediction module 300 may predict how the quality of the finished coffee will be derived when a plurality of raw green coffee beans are roasted in a blended state. When one raw green coffee bean (ie, single origin) is roasted, coffee having characteristics similar to data stored in the single origin database 200 will be extracted. However, when coffee is extracted from a plurality of raw coffee beans in a blending state, it is difficult to predict what characteristics the extracted coffee will have. The user can input the type of raw coffee beans used in the product to be predicted, the mixing ratio, and the roasting degree, and the product characteristic prediction module 300 calls the data stored in the green coffee database 100 and the single origin database 200 . If coffee is extracted based on raw green beans blended with , it is possible to derive a quality prediction value that can predict which characteristics of coffee will be extracted. In this case, the quality prediction value may be expressed as a score for each item of the evaluation table 50 .

The quality prediction value may be determined as the sum of values obtained by multiplying the mixing ratio by the score for each item according to the roasting degree setting for each green coffee bean to be blended. For example, when green raw coffee beans to be blended include the first raw raw green coffee and the second raw raw coffee, the quality predicted value is the first sub-prediction value obtained by multiplying the score for each item according to the roasting degree setting of the first raw raw green coffee by the blending ratio, and the second raw coffee bean. 2 It may be determined as a value obtained by adding the second sub-prediction value obtained by multiplying the mixing ratio to the score for each item according to the roasting degree setting of the raw green coffee beans.

Unlike the above example, the raw green beans to be blended may be blended into three or more types.

The product characteristic prediction module 300 may predict the manufacturing cost of the blended coffee finished product based on data stored in the green coffee database 100 . The type of raw green coffee beans to be blended and the blending ratio of the raw green coffee beans are determined by the user, and data on the cost of each of the raw green coffee beans is stored in the green coffee database 100 . Accordingly, the product characteristic prediction module 300 may derive the manufacturing cost of the finished coffee product to be finally produced based on the data on the cost of each of the raw raw coffee beans, the types of raw raw coffee beans, and the blending ratio of the raw raw coffee beans.

The product evaluation module 400 may check the characteristics of the final product produced through the process of evaluating the produced product, and the characteristics of fragrance and acidity. For example, the product evaluation module 400 may collect data obtained through sensory evaluation of evaluators or measurement devices such as electronic tongue and GC-MS. The product evaluation module 400 may store the acquired data in the product database 500 . The product evaluation module 400 may store the acquired data according to the evaluation table 50 for each item, and then store the characteristics, fragrance, and acidity characteristics of the final product to which the score is given in the product database 500 . That is, the product database 500 may store the actual quality value of the finished coffee product.

The comparison module 600 may compare the actual quality value stored in the product database with the predicted quality value predicted by the product characteristic prediction module 300 . The comparison module 600 may evaluate the reliability of the predicted quality value based on the difference between the actual quality value and the predicted quality value. For example, when there are more than three items that are not included within ±1.0 of the range where the quality prediction value evaluated by the comparison module is stunned, the comparison module 600 may recognize the predicted quality prediction value as erroneous data. The comparison module 600 may discard the quality prediction value determined as erroneous data. In addition, the comparison module 600 may update the single origin database 200 based on the result of comparing the quality predicted value with the actual quality value of the actual quality value when the reliability of the quality predicted value is recognized. The single origin database 200 stores characteristics of raw green coffee beans having different characteristics for each roasting degree, but does not store characteristics of blended green coffee beans. Therefore, by updating the single origin database 200 in consideration of the actual quality value according to the characteristics of the finished coffee product, the reliability of the quality prediction value can be improved.

As an example, the comparison module 600 may store the new data derived by adding the difference between the predicted quality value and the actual quality value to the score for each item of raw green coffee beans in the single origin database 200 . The single origin database 200 may derive an average value based on the new data and the scores for each item according to the roasting degree of the existing green raw coffee beans, and determine the derived average value as the score for each item according to the freshness of the new raw green coffee beans. . In other words, the single origin database 200 may store the average value of the premise data obtained by adding the new data to the scores for each item according to the roasting degree of the existing green coffee beans as the score for each item of the new raw green coffee beans.

For example, if the first raw raw green coffee had a score of 6, the second raw green coffee had a sour score of 8, and the first raw green coffee and second raw raw coffee were blended in a 50:50 ratio, the comparison module (600 ) is +1, the comparison module 600 adds +1 to the score for the sour taste of the first raw green coffee and +1 to the score for the sour taste of the second raw green coffee. can be derived as new data. In this case, the single origin database 200 may determine and store an average value of 6 points, which is the sour score of the existing first raw green coffee beans, and 7 points, which is the new data, as the new score for the first raw raw coffee beans, and store them. 2 The average value of 8 points, which is the sourness score of the raw green coffee beans, and 8.5, which is the average value of 9 points, which is the new data, can be determined and stored as the new score of the second raw coffee beans. Unlike the above-described example, a method of deriving new data by the comparison module 600 and the single origin database 200 and a method of updating the single origin database 200 may not be particularly limited.

According to an embodiment of the present invention, since the quality evaluation method for evaluating the actual sensory for general consumers is used, the actual sensory result of the product is reflected, and the cost of the finished coffee product before the product is produced without the product producer going through the process of roasting, Characteristics and preferences can be predicted.

According to the embodiment of the present invention, the single origin database 200 can be periodically updated by comparing the quality predicted value with the actual quality value, so that the trend of consumer preference changes for each single origin and blend coffee can be checked and the coffee bean market You can see the trend flow. In addition, as the single origin database 200 is updated based on new data, the reliability of the quality prediction value derived by the product characteristic prediction module 300 may be improved.

3 is a diagram illustrating characteristics of raw green coffee beans stored in a single origin database according to an exemplary embodiment of the present invention by roasting degree.

3 is a graph showing the characteristics of Brazil N.Y.2 Fine Cup green coffee beans based on an evaluation table. That is, FIG. 3 shows the characteristics of Brazil N.Y.2 Fine Cup green beans stored in the single origin database. Characteristics, aroma, and acidity characteristics of raw green coffee beans after roasting Brazil N.Y.2 Fine Cup green beans to a roasting degree of 65.0 may be displayed as shown in FIG. 3 . In addition, the single origin database can store the roasting degree, green bean cost, sensory total score, sample mixing ratio, concept, recommended extraction method, and the name of the applied product.

4 is a diagram illustrating a quality predicted value by a product characteristic prediction module according to an embodiment of the present invention.

4 shows the quality prediction values calculated through the product characteristic prediction module. The blending ratio of raw green beans for evaluation prediction is Brazil N.Y.2 Fine Cup 40%, Colombia Excelso 30%, Indonesia Robusta AP-1 G3 30%, and the roasting value is 65.0. In this case, the product characteristic prediction module may retrieve data on raw green coffee beans stored in the single origin database. The Cup Profile and Cup Profile (strength) of each raw green coffee stored in the single origin database can be scored as shown in Table 1 below.

division distribution diagram Cup Profile Cup Profile
(burglar) Scent Sour taste bitterness booth
love body aftertaste sense of balance whole taste Sour taste bitterness Brazil
N.Y.2
Fine Cup 65.0 5.0 5.5 5.8 6.2 5.5 5.4 6.1 5.3 4.0 5.0 Colombia Excelso 5.8 4.5 5.4 5.0 6.5 5.1 5.0 5.3 5.0 4.5 Indonesia Robusta AP-1 G3 3.9 3.6 4.6 3.4 6.1 3.3 4.1 3.4 2.8 6.9

5 is a view showing the actual quality value of the coffee beans product stored in the product database according to an embodiment of the present invention.

5 is a view of a product database accumulated through a product evaluation module after blending Brazil NY2 Fine Cup, Colombia Excelso, and Indonesia Robusta AP-1 G3 green coffee at a ratio of 4:3:3 and roasting to a value of 65.0. the results are shown. The data stored in the product database can be scored according to the evaluation table in the same way as the data stored in the single origin database. However, in the case of the product database, the type of raw material, the mixing ratio, and the roasting degree are written in an abbreviation at the top, so you can check the formulation simply by the name. The abbreviations for each country are BR for Brazil, CO for Colombia, CR for Costa Rica, ET for Ethiopia, GT for Guatemala, HN for Honduras, ID for India, IN for Indonesia, JM for Jamaica, KE for Kenya, TZ for Tanzania, Rwanda. may be written as RW, and Uganda as UG.

The results of calculating the values of the single origin database as much as the mixing ratio and comparing the summed quality predicted values with the actual quality values are shown in Table 2 below, and it can be seen that the quality predicted values and the actual quality values are similar.

division distribution diagram Cup Profile Cup Profile
(burglar) Scent Sour taste bitterness booth
love body aftertaste sense of balance whole taste Sour taste bitterness predicted value 65.0 4.9 4.6 5.3 5.0 6.0 4.7 5.2 4.7 3.9 5.4 measured value 4.8 4.8 5.2 4.7 5.8 4.2 4.9 4.4 4.3 5.3

At this time, the score for the fragrance among the quality prediction values may be derived through GC-MS. The product characteristic prediction module analyzes the main aromatic components of each raw green coffee bean through GC-MS, classifies the areas for each functional group, and matches them with floral, fruity, sweet, chocolate, nutty, grassy, and others on the evaluation table. can be classified as Table 3 below shows the scores for each green bean flavor derived through GC-MS.

division functional group Floral Fruity Caramelic,
Sweet Chocolate Nutty Herbal,
Vegetables Other Total
Area
(%) L-CFC
Scent floral scent fruit flavor sweet scent chocolate flavor nutty flavor grass scent Other Brazil N.Y.2
Fine Cup 0.8 4.1 13.2 3.1 13.4 3.5 59.9 98.0 Colombia Excelso 0.7 4.2 16.5 3.0 12.5 3.4 58.3 98.6 Indonesia Robusta AP-1 G3 0.5 2.9 7.8 2.8 14.9 2.2 66.6 97.7 predicted value 0.7 3.8 15.5 3.0 13.6 3.1 58.5 98.1

In FIG. 5, among the quality prediction values of the blended products of Brazil NY2 Fine Cup, Colombia Excelso, and Indonesia Robusta AP-1 G3 green coffee beans, the characteristics of fragrance are characteristic of sweet fragrance belonging to Caramelic and Sweet functional groups, and nutty fragrance corresponding to Nutty. indicates that Each of the quality predicted value for the fragrance or the actual quality value for the fragrance may be stored in the single origin database and the product database, and the predicted quality value for the fragrance may be displayed as shown in FIG. 4 , and the actual quality value for the fragrance is shown in FIG. 5 can be displayed as

6 is a flowchart illustrating a method for predicting the quality of a finished coffee product according to an embodiment of the present invention. For brevity of description, description of overlapping content is omitted.

3 to 6 , when raw green coffee beans are received in a factory, the evaluator may store an evaluation result of the raw green coffee beans in a green coffee database. Data on the characteristics of each unique number of raw green coffee beans may be stored in the green coffee database. The characteristics of raw green coffee beans stored in the green coffee database may include information related to the production region, processing method, moisture content, grade based on the number of defects and screen size, cupping evaluation result, and unit price (S100 and S200).

The results of the quality of raw green coffee beans by roasting degree can be stored in the single origin database. After the characteristics of raw raw coffee beans are stored in the green coffee database, the raw raw coffee beans may be roasted for each roasting degree. After the roasted raw green coffee beans are ground in a grinder at a specific grinding degree (eg, grinding degree 2), the sample dish can be filled. After that, when the average of the values measured 10,000 times per second with a laser using the Color Track R-100B Laser Color-Analyzer device is between 57 and 71, the quality prediction value can be derived using the product characteristic prediction module. (S300).

When the user sets the mixing ratio of the raw green coffee beans and the roasting degree of each of the raw green coffee beans, the product characteristic prediction module may derive a quality prediction value. For example, if you blend Brazil NY2 Fine Cup, Colombia Excelso, and Indonesia Robusta AP-1 G3 green beans in a 4:3:3 ratio, the aroma is 4.9, the sour (preference score) is 4.6, and the bitter (preference score) ) is 5.3, softness is 5.0, body is 6.0, aftertaste is 4.7, balance is 5.2, overall taste is 4.7, sour (strength score) is 3.9, and bitter (strength score) is 5.4. 2 and 4). The product characteristic prediction module may derive a quality prediction value based on data stored in a single origin database as shown in FIG. 3 . The quality prediction value for the aroma can be derived through GC-MS, and Table 3 and FIG. 4 show that the sweet and nutty flavor, which accounts for a high percentage of the scores for the main aroma components, are the main characteristics of the blended coffee.

In addition, the product characteristic prediction module may derive the manufacturing cost of the finished coffee product to be finally produced based on the type and blending ratio of raw raw coffee set by the user and the cost of each raw raw coffee stored in the green coffee database.

In addition, the product characteristic prediction module can recommend which extraction method is appropriate using the blended green coffee beans based on the recommended extraction method for each of the raw green coffee beans, the blending ratio of the raw green coffee beans, and the quality predicted value. For example, when hand drip is set as the recommended extraction method for all raw green coffee beans, the product characteristic prediction module can derive that hand drip is an appropriate extraction method for the blended product. For example, when the quality prediction values of raw raw coffee beans are high body score (preference), high bitterness score (preference), and low acidity score (strength) less than 7, the product characteristic prediction module is It can be deduced that the extraction method is appropriate. Likewise, when the quality prediction values of raw green coffee beans are high fragrance score (preference), high sourness score (preference), and low bitterness score (strength) less than 7, the product characteristic prediction module uses hand drip extraction for products blended with the corresponding raw materials. It can be deduced that the method is appropriate. If both items are applicable, both hand drip and espresso shall be listed. However, such a recommendation method is derived according to a criterion set in advance by the designer, and the product characteristic prediction module may recommend an appropriate extraction method in consideration of the preset criterion (S400).

The user may derive an actual quality value according to the sensory evaluation, and the product evaluation module may give a score for each item of the finished coffee product based on the evaluation table and store it in the product database (S500).

The comparison module can compare the quality predicted value derived by the product characteristic prediction module with the actual quality value stored in the product database. The predicted quality value and the actual quality value may be displayed to the user, and the user may intuitively determine the reliability of the predicted quality value ( S600 ).

Furthermore, the comparison module may evaluate the reliability of the predicted quality value according to a difference value between the predicted quality value and the actual quality value. The comparison module may evaluate the reliability of the predicted quality value based on whether a difference between the predicted quality value and the actual quality value is within a preset range. The preset range may be any range that can be set by the user ( S700 ).

If the difference value is outside the preset range, the comparison module may recognize the derived quality prediction value as error data and discard it (S800).

If the difference value is within a preset range, the comparison module may update the single origin database based on the difference value. In this embodiment, predicted quality values for sour taste and bitter taste were derived as 3.9 and 5.4, respectively, and actual quality values for sour and bitter taste were derived as 4.3 and 5.3, respectively. The comparison module may derive new data based on +0.4 and -0.1, which are difference values for sour and bitter taste, and update the single origin database based on the derived new data. However, the method for the comparison module to update the single origin database may not be particularly limited (S900).

In the above, embodiments of the present invention have been described with reference to the accompanying drawings, but those of ordinary skill in the art to which the present invention pertains can practice the present invention in other specific forms without changing its technical spirit or essential features. You can understand that there is Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive.

Claims (11)

a green coffee database for storing raw material green coffee evaluation results;
a single origin database for storing results of characteristics for each roasting degree of the raw green coffee beans based on an evaluation table for a plurality of characteristics of the raw green coffee beans; and
A product characteristic prediction module for deriving a quality prediction value of a finished coffee product according to a blending ratio between a plurality of raw raw coffee beans and a roasting degree of each of the raw raw coffee beans based on the single origin database,
Quality prediction system for finished coffee products. According to claim 1,
The evaluation table includes a preference item, an intensity item, a fragrance item, and a sour taste type item,
Quality prediction system for finished coffee products. According to claim 1,
The quality prediction value is determined as the sum of values obtained by multiplying the blending ratio by the score for each item according to the roasting degree setting for each raw green coffee bean to be blended;
Quality prediction system for finished coffee products. 4. The method of claim 3,
The product characteristic prediction module derives the quality prediction value based on the type, roasting degree, and mixing ratio of the raw green coffee beans to be blended determined by the user;
The blended raw green coffee beans include a first raw green coffee and a second raw green coffee,
The quality prediction value is a first sub-prediction value obtained by multiplying the score for each item according to the roasting degree setting of the first raw green coffee bean by the blending ratio, and a second sub-prediction value obtained by multiplying the score for each item according to the roasting degree setting of the second raw green coffee bean by the blending ratio determined by adding the predicted values,
Quality prediction system for finished coffee products. According to claim 1,
The quality prediction value is expressed as a score for each item in the evaluation table,
Quality prediction system for finished coffee products. According to claim 1,
A product database that stores the actual quality values of finished coffee products, and
Comprising a comparison module for comparing the actual quality value and the quality predicted value stored in the product database,
The comparison module evaluates the reliability of the predicted quality value based on the difference between the actual quality value and the quality predicted value,
Quality prediction system for finished coffee products. 7. The method of claim 6,
When the quality prediction value evaluated by the comparison module is not included in the stunned range, the comparison module recognizes the quality prediction value as error data,
Quality prediction system for finished coffee products. 8. The method of claim 7,
The comparison module updates the single origin database based on a result of comparing the predicted quality value with the actual quality value and the predicted quality value, not error data,
Quality prediction system for finished coffee products. 9. The method of claim 8,
the comparison module stores the new data derived by adding the difference between the predicted quality value and the actual quality value to the score for each item of the raw green coffee beans in the single origin database;
The single origin database derives an average value based on the new data and the existing scores for each item according to the roasting degree of the raw green coffee beans, and determines the derived average value as the score for each item according to the roasting degree of the raw green coffee beans,
Quality prediction system for finished coffee products. According to claim 1,
The green coffee database stores information on production area evaluation, processing method evaluation, moisture content evaluation, number of defects, grade evaluation according to screen size, cupping evaluation, and unit price confirmation,
Quality prediction system for finished coffee products. According to claim 1,
The product characteristic prediction module is configured to derive the manufacturing cost of the finished coffee product according to a blending ratio of a plurality of raw green coffee beans based on data stored in the green coffee database;
Quality prediction system for finished coffee products.

Images