CN112397156B - Digital flavoring method based on K-means clustering - Google Patents
Digital flavoring method based on K-means clustering Download PDFInfo
- Publication number
- CN112397156B CN112397156B CN201910704921.9A CN201910704921A CN112397156B CN 112397156 B CN112397156 B CN 112397156B CN 201910704921 A CN201910704921 A CN 201910704921A CN 112397156 B CN112397156 B CN 112397156B
- Authority
- CN
- China
- Prior art keywords
- fragrance
- raw material
- raw materials
- aroma
- clustering
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
- G16C20/50—Molecular design, e.g. of drugs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
- G16C20/70—Machine learning, data mining or chemometrics
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
- G16C20/90—Programming languages; Computing architectures; Database systems; Data warehousing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- Computing Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Crystallography & Structural Chemistry (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Medicinal Chemistry (AREA)
- Pharmacology & Pharmacy (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Medical Informatics (AREA)
- Fats And Perfumes (AREA)
- Cosmetics (AREA)
Abstract
The invention discloses a digital fragrance blending method based on K-means clustering, which is characterized in that fragrance raw materials with higher similarity in a fragrance raw material information base are clustered by using a K-means clustering algorithm through the established fragrance raw material information base, fragrance raw materials in a fragrance creation target experience formula list are automatically identified based on a clustering division result to obtain and replace related fragrance raw materials with strong correlation, fragrance raw material information updating is completed, and finally dosage information of the fragrance raw materials is adjusted through the difference of fragrance ratio intensity values to generate and output a new formula list. Because the fragrance raw materials have partially similar properties but different performances, the formula created by the method can show new characteristics while maintaining the main characteristics of the original formula.
Description
Technical Field
The invention belongs to the technical field of essence and fragrance raw materials, and particularly relates to a digital fragrance blending method based on K-means clustering.
Background
The long-term accumulated fragrance raw material information (materialization information, sensory information, smell profile and the like), essence formula and various fragrance blending experience models are huge wealth, and modern advanced computer and artificial intelligence technologies become keys for opening the wealth door, which establishes a bridge between fragrance blenders and embodies the experiences of the fragrance blenders in the creation of the fragrance formula.
Chinese patent (CN109726815A) discloses a Pareto-based optimal digital flavoring method, which comprises the steps of establishing a flavor raw material description information base including sensory evaluation of common flavor raw materials, flavor ratio intensity values, flavor retention values, flavor values, steam pressure, simplicity classification types (top flavor, body flavor, base flavor), source classification (natural and synthetic) and other information, then simultaneously using a three-value theory, a dimension-division theory and a resonance theory as optimization targets according to a flavoring target, using total number of flavor raw materials, material preference, flavor creation subjects, quantity ratio of natural and synthetic flavor raw materials and the like as constraint conditions, seeking an optimization solution set by using a genetic algorithm, and finally obtaining a candidate formula list. Although the method can realize the digital incense creating process according to the characteristics of the incense raw materials, the introduction of the experience of an incense burner is lacked, so that the rationality and the industrial applicability of the obtained formula are uncertain. Therefore, in order to enable the formula list obtained by digitally creating the fragrance to be closer to the actual application requirement and effect, the fragrance creation is carried out on the basis of the empirical formula list, so that the formula list obtained by digitally creating the fragrance can better meet the actual requirement.
Disclosure of Invention
The invention aims to provide a mode for creating fragrance on the basis of an empirical formula list, and fragrance raw materials in the empirical formula list are replaced by K-means clustering to obtain a new formula list, so that efficient digital fragrance blending is realized, and actual requirements are met.
The invention provides a digital flavoring method based on K-means clustering, which comprises the following steps:
s1: acquiring a fragrant raw material information base, wherein the fragrant raw material information in the fragrant raw material information base at least comprises a fragrant score and a fragrance ratio value of fragrant raw materials, and the fragrant score of each fragrant raw material forms a feature vector of the fragrant raw material;
s2: clustering and dividing various fragrant raw materials in a fragrant raw material information base by adopting a K-means clustering algorithm based on the characteristic vector of each fragrant raw material;
s3: acquiring an experience formula sheet matched with a fragrance creating target, and selecting fragrance raw materials in the experience formula sheet as fragrance raw materials to be replaced;
s4: acquiring and replacing relevant perfume raw materials of the perfume raw materials to be replaced based on the clustering division result of the perfume raw materials in the perfume raw material information base in the step S2;
identifying a cluster set to which the incense raw materials to be replaced belong, and if no repeated incense raw materials exist, selecting the non-repeated incense raw materials with the strongest correlation with a cluster center in the cluster set as related incense raw materials; if the repeated fragrance raw materials exist, selecting the next strong-correlation and non-repeated fragrance raw materials as the related fragrance raw materials in the clustering set according to the sequence from strong correlation to weak correlation with the clustering center;
s5: and adjusting the dosage of the raw materials based on the difference of the fragrance ratio to obtain a new formula of the creative fragrance target.
Preferably, the note score of the flavor raw material comprises the scores of flue-cured tobacco flavor, sun-cured tobacco flavor, faint scent, fruity flavor, spicy flavor, costustoot, green flavor, flowery flavor, herbal flavor, bean flavor, cocoa flavor, milk flavor, paste flavor, baking flavor and sweet flavor.
Preferably, the perfume raw material information of the perfume raw material further includes basic characteristic information of the perfume raw material, and the basic characteristic information includes a perfume value, a vapor pressure, and a usage range of the perfume raw material.
Preferably, the correlation between the two perfume materials in step S4 is based on the euclidean distance, mahalanobis distance, cosine similarity or pearson correlation coefficient based on the feature vector.
Preferably, the process of clustering and dividing the various types of fragrant raw materials in the fragrant raw material information base by using the K-means clustering algorithm based on the feature vector of each type of fragrant raw material in step S2 is as follows:
s21: manually setting a category number K, and randomly selecting K types of incense raw materials in the incense raw material information base as initial clustering centers, wherein K is a positive integer;
s22: calculating the Euclidean distance from each incense raw material in the incense raw material information base to each clustering center, and dividing each incense raw material into the category corresponding to the clustering center closest to the incense raw material based on the Euclidean distance;
s23: judging whether an iteration termination condition is reached, if not, updating the positions of all the clustering centers, and returning to the step S22; and if so, obtaining stable K clustering division results.
Preferably, the implementation process of step S5 is as follows:
firstly, calculating the aroma ratio strength value of the essence in the empirical formula according to the aroma ratio strength value and the using amount of the raw materials in the empirical formula;
then, based on the fragrance ratio of the essence with the empirical formula, the dosage of the fragrance raw materials is adjusted according to the fragrance ratio of the fragrance raw materials in the new formula, so that the fragrance ratio of the essence with the new formula is the same as that of the essence with the empirical formula.
Advantageous effects
The invention provides a digital flavoring method based on K-means clustering, which clusters fragrant raw materials with higher similarity in a raw material library through a K-means clustering algorithm, selectively replaces the fragrant raw materials in an empirical formula list based on a clustering division result, wherein the fragrant raw materials with strongest correlation are identified for replacement to complete raw material information updating. In addition, the method of the invention creates fragrance on the basis of the empirical formula, has simple mode and higher efficiency, and ensures that the formula obtained by digitally creating fragrance better meets the actual requirement.
Drawings
FIG. 1 is a schematic flow diagram of the present invention.
Detailed Description
The present invention will be further described with reference to the following examples.
The invention provides a digital flavoring method based on K-means clustering, which is used for creating fragrance on the basis of an empirical formula. The embodiment of the invention specifically explains the invention by taking the flavoring in the tobacco industry as an example. The flavoring method provided by the embodiment of the invention comprises the following steps:
s1: and acquiring a perfume raw material information base. In this embodiment, the fragrance raw material information base includes 344 kinds of fragrance raw materials, and the fragrance raw material information of each kind of raw fragrance raw material includes 15 kinds of scores of fragrance notes (flue-cured tobacco fragrance, air-cured tobacco fragrance, faint scent, fruit fragrance, spicy fragrance, radix aucklandiae, green scent, flower fragrance, herb fragrance, bean fragrance, cocoa fragrance, milk fragrance, cream fragrance, baking fragrance, and sweet fragrance), a fragrance specific strength value, a fragrance retention value, a fragrance value, steam pressure, and a dosage range. In other possible embodiments, the perfume material information may be a combination of the above or contain other types of values, and the present invention is not limited to this. The following table 1 shows the perfume raw material information of several perfume raw materials in this example:
TABLE 1 perfume materials description information base (part)
Wherein, if the partial information of the partial incense raw material is null, the value is regarded as 0.
The aroma score of each aroma raw material in the aroma raw material information base is represented by one characteristic vector, if N kinds of aroma raw material information are stored in the aroma raw material information base, and the aroma score of each aroma raw material is represented by one characteristic vector p multiplied by 1, the aroma score of the aroma raw materials in the whole aroma raw material information base can be represented by one N multiplied by p matrix, wherein each row corresponds to the aroma score information of one aroma raw material. The fragrance value, vapor pressure and dosage range of the fragrance raw materials form basic characteristic information of the fragrance raw materials, the basic characteristic information can be used for assisting in adjusting the dosage of the fragrance raw materials in the formula, clustering results can also be subsequently restrained, and the rationality of the fragrance creating results is judged and optimized through the basic characteristic information, for example, if the dosage of a certain fragrance raw material in an output formula list exceeds the dosage range in common use, the fragrance raw material needs to be manually adjusted, so that the rationality of the formula structure is ensured.
S2: and (3) clustering and dividing various fragrant raw materials in the fragrant raw material information base by adopting a K-means clustering algorithm based on the characteristic vector of each fragrant raw material. The K-means clustering algorithm is one of the most representative algorithms in the prior numerous prototype clustering algorithms, is an iterative solution clustering analysis algorithm, and comprises the steps of randomly selecting K objects as initial clustering centers, then calculating the distance between each object and each seed clustering center, and allocating each object to the nearest clustering center. The cluster centers and the objects assigned to them represent a cluster, and the cluster centers of the clusters are recalculated based on the objects existing in the cluster, for each sample assigned. This process will be repeated until some termination condition is met. The termination condition may be that no (or minimum number) objects are reassigned to different clusters or that no (or minimum number) cluster centers are changed again or that the sum of squared errors is locally minimal. The method is applied to blending fragrance, and clustering division is carried out on the fragrance raw materials in the fragrance raw material information base through the algorithm.
The specific implementation process of step S2 is as follows:
s21: and randomly selecting K kinds of incense raw materials in the incense raw material information base as initial clustering centers, wherein K is a positive integer. Wherein if the incense raw material information base contains N kinds of incense raw materials, each incense raw material X i (i ═ 1,2, …, N) is represented by a p × 1 feature vector. C for clustering center k And (K ═ 1,2, …, K).
S22: and calculating the Euclidean distance from each fragrance raw material to each clustering center in the fragrance raw material information base, and dividing each fragrance raw material into the category corresponding to the clustering center closest to the fragrance raw material based on the Euclidean distance.
Specifically, each incense raw material X in the incense raw material information base i Calculating its Euclidean distance d to each cluster center ik ＝||X i -C k And (i is 1,2, …, N; K is 1,2, …, K), and the fragrance raw materials are divided into categories corresponding to the nearest clustering centers.
S23: judging whether an iteration termination condition is reached, if not, updating the positions of all the clustering centers, and returning to the step S22; and if so, obtaining stable K clustering division results.
The iteration termination condition may be that the set iteration number or the clustering result is not changed or the clustering center is not changed or the moving distance of the clustering center is smaller than a preset threshold or the error square sum is locally minimum. The position updating of the clustering center is a conventional means of the existing K-means clustering, and is not described in detail herein.
In this embodiment, the manual setting of the number of categories is selected to be 8, then 8 kinds of fragrant raw materials are randomly selected from the database as initial clustering centers, and the clustering centers are updated by using the above method to obtain 8 clustering division results.
S3: and acquiring an experience formula sheet matched with the fragrance creating target, and selecting the fragrance raw materials in the experience formula sheet as the fragrance raw materials to be replaced. The empirical formula is preferably created by adjusting and verifying by a perfumer and using the perfume raw materials in the above perfume raw material information base, and 3 empirical formulas with the subjects of floral, fragrant and fruity are shown in the chart 2:
TABLE 2 empirical formula
In this embodiment, the recipe list data structure includes five columns, the first column is a serial number, the second column is a name (or index number) of the used flavor material, the third column is a specification (i.e., dilution degree) of the flavor material, the fourth column is a flavor material proportion, and the fifth column is an option description.
In this embodiment, the aim of creating a fragrance is to blend the essence formula theme with fruity fragrance, so the fruity fragrance formula list in table 2 above is used as the experience formula list. In other embodiments, matching empirical recipes are found based on the theme of the creation, quality improvement goals, and the like.
S4: acquiring and replacing relevant perfume raw materials of the perfume raw materials to be replaced based on the clustering division result of the perfume raw materials in the perfume raw material information base in the step S2; selecting a perfume raw material which belongs to the same cluster as the perfume raw material to be replaced and is selected as a related perfume raw material of the perfume raw material to be replaced according to the strength of the correlation with the cluster center, if no repeated perfume raw material exists, selecting the perfume raw material with the strongest correlation, and if the perfume raw material with the strongest correlation is repeated, selecting the next unrepeated perfume raw material according to the strength relationship. Specifically, the euclidean distance, the mahalanobis distance, the cosine similarity, or the pearson correlation coefficient may be used as the correlation determination basis, and the euclidean distance is selected as the determination basis in this embodiment.
S5: and adjusting the dosage of the raw materials based on the difference of the fragrance ratio to obtain a new formula of the creative fragrance target. The process is as follows:
firstly, the aroma ratio strength value of the essence in the empirical formula is calculated according to the aroma ratio strength value of the raw materials used in the empirical formula, then the obtained aroma ratio strength value of each raw material used in the creative essence is used as a value interval, the dosage of different raw materials is automatically adjusted by using the dosage range of each raw material as a value interval through a computer, so that the aroma ratio strength value of the creative essence is consistent with the aroma ratio strength value of the essence in the original empirical formula, and finally the creative essence formula can be output. (the aroma specific strength value of the essence is calculated by the formula of (the amount of the aroma raw material 1 + the amount of the aroma raw material 2 + … + the amount of the aroma raw material n + the aroma specific strength value of the aroma raw material n)/(the amount of the aroma raw material 1 + the amount of the aroma raw material 2 + … + the amount of the aroma raw material n)).
In this embodiment, for all used fragrance raw materials in the selected fruit fragrance experience formula sheet, it is determined respectively that the clustering information to which the fragrance raw materials belong obtains the respective cluster set to which the fragrance raw materials belong, and the fragrance raw materials are selected as the relevant fragrance raw materials in the corresponding cluster set according to the strong and weak relationship with the cluster center for replacement, specifically, it is determined first whether the fragrance raw material with the strongest correlation with the cluster center is the repeated fragrance raw material in the formula, if not, the fragrance raw material is used as the replacement fragrance raw material for replacing the corresponding fragrance raw material in the experience formula, if repeated, the next unrepeated fragrance raw material is selected as the relevant fragrance raw material in the order from strong to weak according to the correlation for replacement, and finally, the original information update is completed, and then the dosage information of the fragrance raw materials is adjusted according to the difference of the fragrance ratio strength value, and the formula sheet shown in table 3 is output.
Table 3 shows the output recipe list of creating the fragrance and having the theme of fruit fragrance
The essence formula which is automatically generated and has the fragrance creating subject of fruit fragrance is prepared through tests, and a fragrance adding test is carried out, so that the formula is prominent in fruit fragrance, obvious in fragrance and rich in sweet flavor through inspection, plays a certain role in increasing the fragrance quality of cigarettes and reducing miscellaneous gases, and can effectively improve the fragrance quality and richness of the fragrance of the cigarettes. The process greatly reduces the time and energy requirements of modulation completely by experience, and can effectively improve the efficiency of creating fragrance.
It should be emphasized that the examples described herein are illustrative and not restrictive, and thus the invention is not to be limited to the examples described herein, but rather to other embodiments that may be devised by those skilled in the art based on the teachings herein, and that various modifications, alterations, and substitutions are possible without departing from the spirit and scope of the present invention.
Claims (6)
1. A digital flavoring method based on K-means clustering is characterized in that: the method comprises the following steps:
s1: acquiring a fragrant raw material information base, wherein the fragrant raw material information in the fragrant raw material information base at least comprises a fragrant score and a fragrance ratio value of fragrant raw materials, and the fragrant score of each fragrant raw material forms a feature vector of the fragrant raw material;
s2: clustering and dividing various fragrant raw materials in a fragrant raw material information base by adopting a K-means clustering algorithm based on the characteristic vector of each fragrant raw material;
s3: acquiring an experience formula sheet matched with a fragrance creating target, and selecting fragrance raw materials in the experience formula sheet as fragrance raw materials to be replaced;
s4: acquiring and replacing relevant perfume raw materials of the perfume raw materials to be replaced based on the clustering division result of the perfume raw materials in the perfume raw material information base in the step S2;
identifying a cluster set to which the incense raw materials to be replaced belong, and if no repeated incense raw materials exist, selecting the non-repeated incense raw materials with the strongest correlation with a cluster center in the cluster set as related incense raw materials; if the repeated fragrance raw materials exist, selecting the next strong-correlation and non-repeated fragrance raw materials as the related fragrance raw materials in the clustering set according to the sequence from strong correlation to weak correlation with the clustering center;
s5: and adjusting the dosage of the raw materials based on the difference of the fragrance ratio to obtain a new formula of the creative fragrance target.
2. The method of claim 1, wherein: the grading of the aroma of the raw materials comprises grading of flue-cured tobacco aroma, sun-cured tobacco aroma, faint scent, fruit aroma, spicy aroma, costustoot, green essence aroma, flower aroma, herb aroma, bean aroma, cocoa aroma, milk aroma, paste aroma, baking aroma and sweet aroma.
3. The method of claim 1, wherein: the incense raw material information of the incense raw material also comprises basic characteristic information of the incense raw material, wherein the basic characteristic information comprises a fragrance retaining value, a fragrance value, steam pressure and a dosage range of the incense raw material.
4. The method of claim 1, wherein: the correlation between the two perfume materials in step S4 is based on the euclidean distance, mahalanobis distance, cosine similarity or pearson correlation coefficient based on the feature vector.
5. The method of claim 1, wherein: the process of clustering and dividing the various fragrant raw materials in the fragrant raw material information base by adopting a K-means clustering algorithm based on the characteristic vector of each fragrant raw material in the step S2 is as follows:
s21: manually setting a category number K, and randomly selecting K types of incense raw materials in the incense raw material information base as initial clustering centers, wherein K is a positive integer;
s22: calculating the Euclidean distance from each incense raw material in the incense raw material information base to each clustering center, and dividing each incense raw material into the category corresponding to the clustering center closest to the incense raw material based on the Euclidean distance;
s23: judging whether an iteration termination condition is reached, if not, updating the positions of all the clustering centers, and returning to the step S22; and if so, obtaining stable K clustering division results.
6. The method of claim 1, wherein: the implementation process of step S5 is as follows:
firstly, calculating the aroma ratio strength value of the essence in the empirical formula according to the aroma ratio strength value and the using amount of the raw materials in the empirical formula;
then, based on the fragrance ratio of the essence with the empirical formula, the dosage of the fragrance raw materials is adjusted according to the fragrance ratio of the fragrance raw materials in the new formula, so that the fragrance ratio of the essence with the new formula is the same as that of the essence with the empirical formula.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910704921.9A CN112397156B (en) | 2019-07-31 | 2019-07-31 | Digital flavoring method based on K-means clustering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910704921.9A CN112397156B (en) | 2019-07-31 | 2019-07-31 | Digital flavoring method based on K-means clustering |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112397156A CN112397156A (en) | 2021-02-23 |
| CN112397156B true CN112397156B (en) | 2022-08-16 |
Family
ID=74601224
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910704921.9A Active CN112397156B (en) | 2019-07-31 | 2019-07-31 | Digital flavoring method based on K-means clustering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112397156B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114431504B (en) * | 2022-01-20 | 2023-06-02 | 红云红河烟草(集团)有限责任公司 | Tobacco leaf group dosing scheme planning algorithm based on formula bill analytic calculation |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102830196A (en) * | 2012-09-25 | 2012-12-19 | 广东中烟工业有限责任公司 | Method for detecting quality of tobacco flavor |
| CN107341613A (en) * | 2017-07-07 | 2017-11-10 | 云南中烟工业有限责任公司 | A kind of method for aiding in tobacco leaf formulation balance to replace |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013035070A1 (en) * | 2011-09-07 | 2013-03-14 | Yeda Research And Development Co. Ltd. | Olfactory signature and odorant mixture having the same |
| US20150220592A1 (en) * | 2014-01-27 | 2015-08-06 | Foodpairing Nv | Method and system for substituting a food or drink or one or more ingredients thereof |
-
2019
- 2019-07-31 CN CN201910704921.9A patent/CN112397156B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102830196A (en) * | 2012-09-25 | 2012-12-19 | 广东中烟工业有限责任公司 | Method for detecting quality of tobacco flavor |
| CN107341613A (en) * | 2017-07-07 | 2017-11-10 | 云南中烟工业有限责任公司 | A kind of method for aiding in tobacco leaf formulation balance to replace |
Non-Patent Citations (3)
| Title |
|---|
| 《GC/MS结合计算机辅助调香系统在烟用香精配方解析中的应用》;杨蕊;《中国优秀博硕士学位论文全文数据库(硕士)工程科技Ⅰ辑》;20140515(第5期);第B014-397页 * |
| 《QSPR analysis for the retention index of flavors and fragrances on a OV-101 column》;Cristian Rojas Villa，et al;《chemometrics and intelligent laboratory systems》;20150131;第126-132页 * |
| 《聚类分析技术在造纸法再造烟叶原料配方中的应用》;郝建辉，等;《中国烟草科学》;20100831;第31卷(第4期);第72-75页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112397156A (en) | 2021-02-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108509534B (en) | Personalized music recommendation system based on deep learning and implementation method thereof | |
| CN104914720B (en) | Electronic nose intelligence baking control system and control method with autolearn feature | |
| CN106841494B (en) | Plant otherness metabolin rapid screening method based on UPLC QTOF | |
| CN101414183B (en) | Cigarette working procedure quality overall evaluation system and method based on gray correlation analysis | |
| Chapman et al. | Using crop simulation to generate genotype by environment interaction effects for sorghum in water-limited environments | |
| CN109100321A (en) | A kind of cigarette recipe maintenance method | |
| CN112397156B (en) | Digital flavoring method based on K-means clustering | |
| CN107341613B (en) | Method for assisting balance replacement of leaf group formula | |
| CN106644983B (en) | Spectral wavelength selection method based on PLS-VIP-ACO algorithm | |
| CN109447167B (en) | Intelligent cigarette formula maintenance method based on non-negative matrix factorization | |
| CN104568825B (en) | The method of cured tobacco leaf beating and double roasting composed piece | |
| CN106529584A (en) | Flue-cured tobacco aroma type and quality judgment intelligent evaluation method | |
| CN107784452A (en) | A kind of objective integrated evaluating method of tobacco style characteristic similarity | |
| CN104820724A (en) | Method for obtaining prediction model of knowledge points of text-type education resources and model application method | |
| CN110132879A (en) | A kind of flue-cured tobacco note discrimination method based near infrared spectrum | |
| CN103070465A (en) | Tobacco leaf composition blending method based on compatibility | |
| CN106202999A (en) | Microorganism high-pass sequencing data based on different scale tuple word frequency analyzes agreement | |
| CN105092789A (en) | Analysis method for tobacco leaf style characteristic variation trend in lamina aging process | |
| Kao et al. | A modified Roger’s distance algorithm for mixed quantitative–qualitative phenotypes to establish a core collection for Taiwanese vegetable soybeans | |
| CN1115112C (en) | Method for creating fuzzy-neural network expert system for evaluating sensing quality of cigarette | |
| CN106645530B (en) | A method of the multi-model based on tobacco leaf aroma component evaluates raw tobacco material similarity | |
| CN107179379A (en) | A kind of appraisal procedure of cigarette mainstream flue gas composition sense organ style index | |
| CN108520276B (en) | Characterization method for internal sensory quality of tobacco leaf raw material | |
| CN108872142B (en) | Multi-parameter selection optimization method in wavelength selection algorithm | |
| CN111999443A (en) | Evaluation method of backup variety resources of cigarette industry base |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |