A dynamic control system and method for cigar flavor
By combining flavor analysis and data mining techniques with GC-MS, electronic nose, and electronic tongue, personalized cigar recipes are generated, solving the problem of flavor customization in traditional cigar manufacturing and improving consumer satisfaction and production efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHENZHEN TIANJI ELECTRONIC TECH CO LTD
- Filing Date
- 2025-06-17
- Publication Date
- 2026-05-26
AI Technical Summary
Current cigar manufacturing technology cannot achieve personalized flavor customization and cannot meet the diverse needs of different consumers.
By employing a flavor analysis module, a demand collection module, a consumer habit database, a recipe generation module, and a feedback adjustment module, combined with GC-MS, electronic nose, and electronic tongue, a cigar flavor fingerprint spectrum is established. Personalized cigar recipes are generated through data mining and multi-objective optimization algorithms, and real-time feedback adjustments are made.
It enables the generation of personalized cigar recipes based on consumer information and habits, improving the consumer experience and production efficiency, and ensuring that products continue to meet consumer expectations.
Smart Images
Figure CN120707183B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of cigar production technology, and in particular to a dynamic control system and method for cigar flavor. Background Technology
[0002] The traditional cigar manufacturing industry has long relied on experience and craftsmanship, which, while ensuring the unique appeal of its products, has proven insufficient to meet the increasingly diverse demands of consumers. With technological advancements and consumers' pursuit of personalized, high-quality products, the cigar industry urgently needs a new technology that can precisely control flavor and quickly respond to market changes.
[0003] A search revealed Chinese patent application CN202111632686.2, which discloses a machine vision-based cigar appearance quality inspection and sorting device and method. The device includes a cigar sampling system, a roller feeding system, a cigar sampling platform, a cigar storage system, an image acquisition and detection system, and a control system. The method utilizes the co-rotation of two rollers in the sampling platform to rotate the cigar sample. A CCD industrial camera captures an image of the cigar surface, which is then processed and stitched together by a computer image processing system to obtain a panoramic image of the cigar surface. The image is then analyzed for impurities and color variations, and the length and width of the cigars are calculated to identify defective and qualified cigars. Finally, the control system controls the direction and speed of the two rollers to push the defective and qualified cigars into different storage boxes. However, the sorting device in this patent has the following shortcomings: while it enables intelligent production, it cannot achieve personalized flavor customization and cannot meet the individual needs of different consumers. Summary of the Invention
[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a dynamic control system and method for cigar flavor.
[0005] To achieve the above objectives, the present invention adopts the following technical solution:
[0006] A dynamic flavor control system for cigars, comprising:
[0007] The flavor analysis module analyzes the volatile components in cigar tobacco leaves, performs sensory evaluation of the overall flavor of cigars, and establishes a cigar flavor fingerprint spectrum.
[0008] The demand collection module is used to collect current consumer demand, including consumers' personal information and cigar consumption habits data.
[0009] A consumer habit database stores consumers' personal information and cigar consumption habits. Through data mining and analysis techniques, a correlation model between consumer habits and cigar flavor characteristics is established.
[0010] The recipe generation module generates personalized cigar recipes based on the association model in the consumer habit database and the specific information and preferences of the current consumers.
[0011] The production management module manages production based on the generated personalized cigar recipes.
[0012] The feedback and adjustment module performs flavor testing and consumer feedback evaluations after the cigars are produced, and then makes adjustments based on the feedback.
[0013] Preferably, the flavor analysis module analyzes volatile components using the following method:
[0014] GC-MS was used to analyze the volatile components in cigar tobacco leaves; let the content of a certain flavor compound in the cigar tobacco leaf sample be _____. Where i = 1, 2, ..., n, and n is the number of flavor compounds, the peak area of the compound is obtained by GC-MS detection. The content was calculated using the standard curve method; the equation of the standard curve is: ,in and It is a coefficient obtained by measuring a standard with a known concentration.
[0015] Preferably, the flavor analysis module uses the following method when performing sensory evaluation:
[0016] Sensory evaluation of the overall flavor of cigars was performed using an electronic nose and electronic tongue. The sensor array based on the electronic nose showed different response values to different odor molecules. Let sensor j be denoted as j = 1, 2, ..., m, where m is the number of sensors. The response value to the cigar sample is... Principal component analysis was used to process the response values to obtain the principal component scores characterizing the aroma features of cigars. Principal component analysis was used to obtain the principal component scores of the electronic tongue's representation of taste characteristics. Where k=1,2,…,p, and p is the number of electronic tongue sensors; by combining data from GC-MS, electronic nose, and electronic tongue, a cigar flavor fingerprint spectrum is established.
[0017] Preferably, when the demand collection module collects consumers' personal information and cigar consumption habit data, it does so as follows:
[0018] Consumer information includes age, gender, smoking duration, and basic regional information. Let age be Y, gender coefficient be G, smoking duration be X, and regional coefficient be D.
[0019] Cigar consumption habit data includes preferred cigar brands, commonly used sizes, daily smoking volume, and preference ratings for descriptive terms for different flavors.
[0020] Preferably, when establishing a correlation model between consumer habits and cigar flavor characteristics using the aforementioned consumer habit database, the specific details are as follows:
[0021] Let the preference rating for concentration be... The preference for light flavor is rated as follows: The sweetness preference rating is The spiciness preference rating is Through data mining and analysis techniques, a correlation model between consumer habits and cigar flavor characteristics is established; a multiple linear regression model is used to establish the degree of consumer preference for a certain flavor. Relationship with consumer information and consumption habit data:
[0022] ;
[0023] in: , , , , , , , , For regression coefficients, This is the random error term.
[0024] Preferably, the recipe generation module generates personalized cigar recipes in the following way:
[0025] Let the vector of tobacco raw material types in the generated formula be . The additive type ratio vector is The recipe generation is based on the following optimization objective function:
[0026] ;
[0027] Where q represents the number of types of tobacco raw materials, and r represents the number of types of additives. This represents the actual content of the i-th flavor compound in a cigar produced according to the formula. The expected content is determined based on consumer preferences; and Principal component scores for the actual and expected electronic nose sensor j, respectively; and λ and μ are the principal component scores of the actual and expected electronic tongue sensor k, respectively; λ and μ are weighting coefficients used to balance the importance of chemical component content and sensory evaluation indicators; by solving this optimization problem, the optimal formula that meets the personalized needs of consumers is obtained.
[0028] Preferably, the feedback adjustment module specifically includes:
[0029] After cigar production is completed, flavor testing and consumer sampling evaluations are conducted. The test results are then compared with the expected values in the recipe generation module to calculate the flavor deviation. :
[0030] ;
[0031] At the same time, feedback from consumers who tried the product was collected, including satisfaction ratings for the flavor. Suggestions for improvement: Adjust and optimize the parameters of the recipe generation module based on flavor deviation and consumer feedback.
[0032] Preferably, the feedback adjustment module, when making feedback adjustments, further includes establishing a mathematical model of flavor score and entry depth through a trial sucking experiment and data analysis methods, and adjusting the strategy accordingly, as follows:
[0033] Through consumer inhalation tests, samples were collected at different inhalation depths. Below is consumer rating data for various cigar flavors; let the consumer's rating for the k-th flavor at different depths of entry be... , where k=1,2,…,s, and s is the number of flavor types;
[0034] Using data analysis methods, a mathematical model of flavor score and entry depth was established. Multiple regression analysis was then employed to obtain the following results:
[0035] ;
[0036] in, , , , , For regression coefficients, This represents the random error term; the model is used to understand the influence of different inlet depths on the perception of various flavors.
[0037] Preferably, when the feedback adjustment module performs feedback adjustment, it adjusts the formula according to the consumer's entry depth; and combines personal information and preferences in the consumer habit database with the correlation model between entry depth and flavor to formulate a personalized formula adjustment strategy.
[0038] Preferably, the control method of the system includes the following steps:
[0039] S1: Data acquisition and analysis;
[0040] S11: Collect flavor data of cigars from different brands and origins, analyze the data using GC-MS, electronic nose and electronic tongue devices in the flavor analysis module, and establish a cigar flavor database.
[0041] S12: Collect consumers' personal information and cigar consumption habits data, clean, organize and analyze the data, and build a consumer habit database;
[0042] S2: Formulation design and production;
[0043] S21: The demand collection module collects consumers' personalized customization needs and inputs relevant consumer information and preference data into the system;
[0044] S22: The recipe generation module generates personalized cigar recipes based on the association model in the consumer habit database and the optimization objective function;
[0045] S23: The production management module produces corresponding products based on the generated personalized cigar recipes;
[0046] S3: Quality Inspection and Feedback;
[0047] S31: After production is completed, flavor testing is conducted on the customized cigars, including GC-MS analysis, electronic nose and electronic tongue evaluation, and consumers are organized to try them out.
[0048] S32: The feedback adjustment module calculates the flavor deviation and satisfaction score based on the test results and consumer feedback. If the product does not meet the requirements, the relevant information is fed back to the formula generation module, the parameters are adjusted and production is restarted until the flavor effect that satisfies consumers is achieved.
[0049] The beneficial effects of this invention are as follows:
[0050] 1. The system of the present invention can generate personalized cigar recipes based on consumers' personal information, consumption habits and specific preferences, thereby meeting consumers' personalized needs for flavor and enhancing the consumption experience.
[0051] 2. This invention combines analytical techniques such as GC-MS, electronic nose, and electronic tongue to comprehensively analyze the volatile components and sensory characteristics of cigars, establishes a precise flavor fingerprint spectrum, and provides a scientific basis for formula design.
[0052] 3. This invention utilizes big data analytics to uncover the correlation model between consumer habits and cigar flavor characteristics, enabling intelligent and precise recipe generation, thereby improving production efficiency and product quality.
[0053] 4. Through post-production quality inspection and consumer feedback, the system can calculate flavor deviation in real time and dynamically adjust the formula and production process based on the feedback results to ensure that the product continuously meets consumer expectations. By using a mathematical model of entry depth and flavor perception, combined with individual consumer habits, the system can further refine the formula adjustment strategy, enhance consumer participation and satisfaction, and improve brand image.
[0054] 5. In the process of formula generation, this invention comprehensively considers the content of chemical components, sensory evaluation indicators and consumer preferences, and achieves comprehensive optimization of product flavor and quality through a multi-objective optimization algorithm. Attached Figure Description
[0055] Figure 1 This is a framework diagram of a dynamic control system for cigar flavor proposed in this invention;
[0056] Figure 2 This is a flowchart of a method for dynamically controlling cigar flavor proposed in this invention. Detailed Implementation
[0057] The technical solution of the present invention will be further described in detail below with reference to specific embodiments. Example
[0058] A dynamic flavor control system for cigars, comprising:
[0059] The flavor analysis module analyzes the volatile components in cigar tobacco leaves, performs sensory evaluation of the overall flavor of cigars, and establishes a cigar flavor fingerprint for subsequent flavor analysis and control.
[0060] The demand collection module is used to collect current consumer demand, including consumers' personal information and cigar consumption habits data.
[0061] A consumer habit database stores consumers' personal information and cigar consumption habits. Through data mining and analysis techniques, a correlation model between consumer habits and cigar flavor characteristics is established.
[0062] The recipe generation module generates personalized cigar recipes based on the association model in the consumer habit database and the specific information and preferences of the current consumers.
[0063] The production management module manages production based on the generated personalized cigar recipes.
[0064] The feedback and adjustment module performs flavor testing and consumer feedback evaluations after the cigars are produced, and then makes adjustments based on the feedback.
[0065] The flavor analysis module analyzes volatile components using the following method:
[0066] GC-MS (Gas Chromatography-Mass Spectrometry) was used to analyze the volatile components in cigar tobacco leaves; let the content of a certain flavor compound in the cigar tobacco leaf sample be _____. (i=1,2,…,n, where n is the number of flavor compounds), the peak area of the substance was obtained by GC-MS detection. The content was calculated using the standard curve method; the equation of the standard curve is: ,in and It is a coefficient obtained by measuring a standard with a known concentration.
[0067] The flavor analysis module uses the following method when performing sensory evaluation:
[0068] Sensory evaluation of the overall flavor of cigars was performed using an electronic nose and electronic tongue. The sensor array based on the electronic nose showed different response values to different odor molecules. Let the response value of sensor j (j=1,2,…,m, where m is the number of sensors) to a cigar sample be... The response values were processed using methods such as principal component analysis (PCA) to obtain the principal component scores characterizing the aroma features of cigars. Similarly, with electronic tongues, principal component scores representing taste characteristics are obtained. (k=1,2,…,p, where p is the number of electronic tongue sensors); By combining data from GC-MS, electronic nose, and electronic tongue, a cigar flavor fingerprint is established for subsequent flavor analysis and regulation.
[0069] Specifically, when the demand collection module collects consumers' personal information and cigar consumption habit data, it does so as follows:
[0070] Consumer information includes basic information such as age, gender, smoking duration, and region. Let age be Y, gender coefficient (1 for males and 0 for females) be G, smoking duration be X, and region coefficient (divided according to different regions, such as 1 for North China and 2 for South China) be D.
[0071] Cigar consumption habit data includes preferred cigar brands, commonly used sizes, daily smoking volume, and preference ratings for different flavor descriptions (such as strong, mild, sweet, spicy, etc.) (rating range 1-5 points).
[0072] Specifically, when establishing the correlation model between consumer habits and cigar flavor characteristics using the aforementioned consumer habit database, the following is a detailed explanation:
[0073] Let the preference rating for concentration be... The preference for light flavor is rated as follows: The sweetness preference rating is The spiciness preference rating is Through data mining and analysis techniques, a correlation model between consumer habits and cigar flavor characteristics is established; a multiple linear regression model is used to establish the degree of consumer preference for a certain flavor. Relationship with consumer information and consumption habit data:
[0074] ;
[0075] in: For regression coefficients, This is the random error term.
[0076] The recipe generation module generates personalized cigar recipes in the following way:
[0077] Let the vector of tobacco raw material types in the generated formula be . (q represents the number of tobacco raw material types), the additive type ratio vector is: (r is the number of additive types); the formulation generation is based on the following optimization objective function:
[0078] ;
[0079] in, This represents the actual content of the i-th flavor compound in a cigar produced according to the formula. The expected content is determined based on consumer preferences; and Principal component scores for the actual and expected electronic nose sensor j, respectively; and λ and μ are the principal component scores of the actual and expected electronic tongue sensor k, respectively; λ and μ are weighting coefficients used to balance the importance of chemical component content and sensory evaluation indicators; by solving this optimization problem, the optimal formula that meets the personalized needs of consumers is obtained.
[0080] Specifically, the feedback adjustment module includes:
[0081] After cigar production is completed, flavor testing and consumer sampling evaluations are conducted. The test results are then compared with the expected values in the recipe generation module to calculate the flavor deviation. :
[0082] ;
[0083] At the same time, feedback from consumers who tried the product was collected, including satisfaction ratings for the flavor. (Scoring range: 1-10 points), improvement suggestions, etc.; adjust and optimize the parameters of the recipe generation module based on flavor deviation and consumer feedback.
[0084] For example, if the flavor deviation exceeds the set threshold δ and consumers have low satisfaction with a certain flavor, the proportion of relevant raw materials or processing parameters in the formula will be adjusted, and production and testing will be carried out again until the personalized needs of consumers are met.
[0085] Example 2:
[0086] A dynamic flavor control system for cigars is provided in this embodiment. Based on Embodiment 1, the feedback adjustment module, during feedback adjustment, further includes establishing a mathematical model of flavor score and entry depth using data analysis methods through trial smoking experiments, and adjusting the strategy accordingly. The details are as follows:
[0087] Through consumer inhalation tests, samples were collected at different inhalation depths. Below is consumer rating data for various flavor profiles of cigars (such as intensity, aroma, and sweetness); let the consumer rating for the k-th flavor at different depths of the palate be denoted as . (k=1,2,…,s, where s is the number of flavor types);
[0088] Using data analysis methods, a mathematical model of flavor score and entry depth was established. Multiple regression analysis was then employed to obtain the following results:
[0089] ;
[0090] in, Equal to the regression coefficients, The random error term is represented by these models. Through these models, we can gain a deeper understanding of the influence of different inlet depths on the perception of various flavors.
[0091] Adjust the formula based on the depth of the consumer's entry point; combine personal information and preferences from the consumer habit database, as well as the correlation model between entry point depth and flavor, to formulate personalized formula adjustment strategies; if consumers have a deep entry point habit and prefer a strong flavor, appropriately increase the proportion of tobacco raw materials or the amount of additives with strong flavor release characteristics in the formula generation module; let the proportion of a certain strong flavor tobacco raw material in the original formula be... The ratio adjusted according to the inlet depth is: ,in, This is the adjustment amount calculated based on the model, which is related to the inlet depth. It is related to the degree of consumer preference for that flavor, for example ( Rate consumers' preferences for rich flavors. , (For adjustment coefficients).
[0092] Example 3:
[0093] A method for dynamically controlling cigar flavor includes the following steps:
[0094] S1: Data acquisition and analysis;
[0095] S11: Collect flavor data of cigars from different brands and origins, analyze the data using GC-MS, electronic nose and electronic tongue devices in the flavor analysis module, and establish a cigar flavor database.
[0096] S12: Collect consumers' personal information and cigar consumption habits data, clean, organize and analyze the data, and build a consumer habit database;
[0097] S2: Formulation design and production;
[0098] S21: The demand collection module collects consumers' personalized customization needs and inputs relevant consumer information and preference data into the system;
[0099] S22: The recipe generation module generates personalized cigar recipes based on the association model in the consumer habit database and the optimization objective function;
[0100] S23: The production management module produces corresponding products based on the generated personalized cigar recipes;
[0101] S3: Quality Inspection and Feedback;
[0102] S31: After production is completed, flavor testing is conducted on the customized cigars, including GC-MS analysis, electronic nose and electronic tongue evaluation, and consumers are organized to try them out.
[0103] S32: The feedback adjustment module calculates the flavor deviation and satisfaction score based on the test results and consumer feedback. If the product does not meet the requirements, the relevant information is fed back to the formula generation module, the parameters are adjusted and production is restarted until the flavor effect that satisfies consumers is achieved.
[0104] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.
Claims
1. A dynamic flavor control system for cigars, characterized in that, include: The flavor analysis module analyzes the volatile components in cigar tobacco leaves, performs sensory evaluation of the overall flavor of the cigar, and establishes a cigar flavor fingerprint. The demand collection module collects current consumer demand, gathers consumers' personal information and cigar consumption habit data. The consumer habit database stores consumers' personal information and cigar consumption habit data, and establishes a correlation model between consumer habits and cigar flavor characteristics through data mining and analysis techniques. The recipe generation module generates personalized cigar recipes based on the correlation model in the consumer habit database and the specific information and preferences of current consumers. The production management module manages production based on the generated personalized cigar recipes. The feedback adjustment module performs flavor testing and consumer sampling evaluations after cigar production, and then makes feedback adjustments. The flavor analysis module performs sensory evaluation using an electronic nose and electronic tongue. The electronic nose's sensor array responds differently to different odor molecules. Let sensor j be the number of sensors (j=1,2,…,m), where m is the number of sensors. The response value to the cigar sample is… Principal component analysis was used to process the response values to obtain the principal component scores characterizing the aroma features of cigars. ; Principal component analysis was used to obtain the principal component scores of the electronic tongue representing taste characteristics. Where k = 1, 2, ..., p, and p is the number of electronic tongue sensors; a cigar flavor fingerprint spectrum is established by integrating data from GC-MS, electronic nose, and electronic tongue; the demand acquisition module collects consumers' personal information and cigar consumption habit data as follows: consumer information includes age, gender, smoking years, and basic regional information, denoted as Y, gender coefficient G, smoking years X, and regional coefficient D; cigar consumption habit data includes preferred cigar brands, commonly used sizes, daily smoking volume, and preference ratings for different flavor descriptive terms; the recipe generation module generates personalized cigar recipes as follows: the ratio vector of tobacco leaf raw materials in the generated recipe is denoted as... The additive type ratio vector is ; Recipe generation is based on the following optimization objective function: Where q represents the number of types of tobacco raw materials, and r represents the number of types of additives. This represents the actual content of the i-th flavor compound in a cigar produced according to the formula. The expected content is determined based on consumer preferences; and Principal component scores for the actual and expected electronic nose sensor j, respectively; and The principal component scores of the electronic tongue sensor k are the actual and expected scores, respectively; λ and μ are weighting coefficients used to balance the importance of chemical component content and sensory evaluation indicators; by solving this optimization problem, the optimal formula that meets the personalized needs of consumers is obtained; when the feedback adjustment module performs feedback adjustment, it also includes establishing a mathematical model of flavor score and entry depth through a trial sucking experiment using data analysis methods, and adjusting the strategy accordingly, as follows: through consumer trial sucking experiments, data are collected at different entry depths... Below is consumer rating data for various cigar flavors; let the consumer's rating for the k-th flavor at different depths of entry be... Where k = 1, 2, ..., s, and s is the number of flavor types; using data analysis methods, a mathematical model of flavor score and entry depth is established, and multiple regression analysis is used to obtain: ,in, , , , , For regression coefficients, The random error term is used to understand the influence of different entry depths on the perception of various flavors through the model. When the feedback adjustment module makes feedback adjustments, it adjusts the formula according to the consumer's entry depth. It combines personal information and preferences in the consumer habit database with the correlation model between entry depth and flavor to formulate personalized formula adjustment strategies.
2. The cigar flavor dynamic control system according to claim 1, characterized in that, The flavor analysis module analyzes volatile components using the following method: GC-MS is used to analyze the volatile components in cigar tobacco leaves; let the content of a certain flavor substance in the cigar tobacco leaf sample be... Where i = 1, 2, ..., n, and n is the number of flavor compounds, the peak area of the compound is obtained by GC-MS detection. The content was calculated using the standard curve method; the equation of the standard curve is: ,in and It is a coefficient obtained by measuring a standard with a known concentration.
3. The cigar flavor dynamic control system according to claim 2, characterized in that, When establishing a correlation model between consumer habits and cigar flavor characteristics using the aforementioned consumer habit database, the specific steps are as follows: Let the preference rating for intensity be... The preference for light flavor is rated as follows: The sweetness preference rating is The spiciness preference rating is Through data mining and analysis techniques, a correlation model between consumer habits and cigar flavor characteristics is established; a multiple linear regression model is used to establish the degree of consumer preference for a certain flavor. Relationship with consumer information and consumption habit data: ,in: , , , , , , , , For regression coefficients, This is the random error term.
4. The cigar flavor dynamic control system according to claim 3, characterized in that, Specifically, the feedback adjustment module performs the following steps after cigar production: flavor testing and consumer feedback are conducted; the test results are compared with the expected values in the formula generation module to calculate the flavor deviation. : At the same time, feedback from consumers who tried the product was collected, including satisfaction ratings for the flavor. Suggestions for improvement: Adjust and optimize the parameters of the recipe generation module based on flavor deviation and consumer feedback.
5. The cigar flavor dynamic control system according to claim 4, characterized in that, The control method of the system includes the following steps: S1: Data acquisition and analysis; S11: Collect flavor data of cigars from different brands and origins, analyze the data using GC-MS, electronic nose and electronic tongue devices in the flavor analysis module, and establish a cigar flavor database. S12: Collect consumers' personal information and cigar consumption habits data, clean, organize and analyze the data, and build a consumer habit database; S2: Formulation design and production; S21: The demand collection module collects consumers' personalized customization needs and inputs relevant consumer information and preference data into the system; S22: The recipe generation module generates personalized cigar recipes based on the association model in the consumer habit database and the optimization objective function; S23: The production management module produces corresponding products based on the generated personalized cigar recipes; S3: Quality Inspection and Feedback; S31: After production is completed, flavor testing is conducted on the customized cigars, including GC-MS analysis, electronic nose and electronic tongue evaluation, and consumers are organized to try them out. S32: The feedback adjustment module calculates the flavor deviation and satisfaction score based on the test results and consumer feedback. If the product does not meet the requirements, the relevant information is fed back to the formula generation module, the parameters are adjusted and production is restarted until the flavor effect that satisfies consumers is achieved.