Method for evaluating uniformity of cigarette formula cut tobacco structure

Abstract

This invention belongs to the field of cigarette processing technology, specifically cigarette tobacco shreds. It discloses a method for evaluating the structural uniformity of cigarette blend tobacco shreds, comprising: linearly combining the original structural indicators of the blend tobacco shreds into several uncorrelated principal components using principal component analysis; determining the number of principal components to be extracted based on their eigenvalues ​​and variance contribution rates; ensuring that the extracted principal components reflect as much information as possible about the original variables; establishing a principal component score calculation model based on eigenvectors; and standardizing the measured sample's blend tobacco shred structure data before inputting it into the model for calculation. This invention evaluates the structural uniformity of tobacco shreds based on the process capability control level of the principal component scores. It utilizes the principle of principal component analysis to reduce the dimensionality of several indicators reflecting the structure of the blend tobacco shreds and uses a comprehensive score value to characterize it. This method overcomes the shortcomings of single-indicator evaluation in existing technologies, facilitating the detection and evaluation of the structure of blend tobacco shreds during cigarette product maintenance and processing.

Description

Technical Field

[0001] This invention belongs to the field of cigarette processing technology, and relates to cigarette tobacco, specifically a method for evaluating the structural stability of cigarette formulation tobacco. Background Technology

[0002] The structure of the tobacco blend in cigarettes is a crucial factor affecting the physical quality of cigarettes, significantly influencing the stability and uniformity of tobacco quality. During cigarette manufacturing, the structure of the tobacco blend greatly impacts physical indicators such as hardness and draw resistance, smoke indicators such as tar content, sensory quality, and tobacco consumption, directly affecting the quality of the finished cigarette. Traditional tobacco structure testing and evaluation mainly focuses on aspects such as whole tobacco yield, broken tobacco yield, and whole tobacco yield conversion rate. However, with the increasing demands for high-quality cigarette development, research is gradually delving into the distribution of tobacco within the cigarette itself.

[0003] Existing cigarette-making machines have a production capacity of 10,000 cigarettes / min, while the weight of tobacco in a single cigarette is only about 0.6g. When the equipment is running at high speed, the structure of the tobacco in the cigarette often changes, and there is an urgent need for an evaluation method to assess the stability of the tobacco structure in cigarette formulations. Summary of the Invention

[0004] The purpose of this invention is to provide a method for evaluating the uniformity of the structure of cigarette formulation tobacco. The method utilizes the principle of principal component analysis to reduce the dimensionality of several indicators reflecting the structure of the formulation tobacco and then uses a comprehensive score value to characterize them. Finally, the stability of the formulation tobacco structure is evaluated using a process capability index.

[0005] This invention is achieved using the following technical solution:

[0006] A method for evaluating the structural uniformity of cigarette formulation tobacco shreds includes the following steps:

[0007] S1. Cigarette sampling

[0008] Cigarette samples were collected at the exit of the cigarette machine. Samples were taken once every t time interval, with m cigarettes sampled each time, for a total of n sampling times.

[0009] S2. Sample preparation

[0010] The sample was placed in an atmospheric environment with a certain temperature and humidity to adjust the moisture content; the tobacco shreds inside the cigarette were removed and scattered to serve as the sample.

[0011] More preferably, in step S2, the temperature conditions are between 20 and 24°C, the relative humidity is between 55 and 65%, and the placement time is 48 hours.

[0012] S3. Determine the structure of the formulated tobacco shreds in the sample.

[0013] The structure of the formulated tobacco shreds in the sample was tested using a planar rotating eccentric sieve with n layers of mesh. The proportion of tobacco shreds on each layer of mesh was denoted as T. n After multiple tests, the average value of the test results of the proportion of tobacco shreds of different sizes in the formula tobacco shreds is taken, and the structure of the cigarette formula tobacco shreds is calculated according to formula (1).

[0014] ...(1)

[0015] In the formula:

[0016] T n —The proportion of tobacco shreds in n layers of sieves, accurate to 0.01%;

[0017] F n —The mass of tobacco shreds on n layers of screens, in grams (g);

[0018] n — the number of screens and non-perforated base layers, ordered from largest to smallest aperture size.

[0019] In a further preferred embodiment, in step S3, the planar rotating eccentric detection sieve has a total of 6 layers of screens, with aperture sizes from top to bottom of 7.10mm, 4.50mm, 3.35mm, 2.00mm, 1.25mm, and 1.00mm. The proportion of tobacco shreds on each layer of screens is recorded as T1, T2, T3, T4, T5, and T6, respectively, and the proportion of tobacco shreds on the non-perforated bottom plate is recorded as T7.

[0020] S4. Data Standardization Processing

[0021] Given n tobacco samples, each sample has p tobacco structure indices, the resulting numerical matrix X of tobacco structure is obtained. ij ;

[0022]

[0023] X is processed according to the following formula (2). ij To standardize the data, subtract the mean from the value of each variable and then divide by its standard deviation.

[0024] ... (2)

[0025] More preferably, in step S4, the standardization process requires a mean of 0 and a standard deviation of 1, in order to eliminate the influence of dimensions and variances between different variables.

[0026] S5. Select principal components

[0027] S5.1. As shown in formula (3), calculate the covariance matrix S; its elements S ij Let represent the covariance between the i-th variable and the j-th variable;

[0028] ... (3)

[0029] S5.2 Calculate the eigenvalues ​​and eigenvectors of the covariance matrix.

[0030] The covariance matrix S is decomposed into eigenvalues ​​λ1, λ2, λ3...λp and corresponding eigenvectors e1, e2...ep; the eigenvalue λi represents the variance of the i-th principal component and the eigenvector ei represents the direction of the i-th principal component.

[0031] S5.3 Determine the principal components

[0032] Calculate the cumulative variance contribution rate of the p original variables according to the following formula (4), and select the first k components whose cumulative variance contribution rate reaches the set threshold as principal components.

[0033] κ(k) = ... (4)

[0034] More preferably, in step S5.3, the set threshold for the cumulative variance contribution rate is greater than or equal to 80%;

[0035] S5.4 Calculate the variance explanation contribution rate of the principal components.

[0036] Based on the proportion of variance contribution rate of each principal component in the first k cumulative variance contribution rates, the variance explanation contribution rate of each principal component is calculated as shown in the following formula (5);

[0037] νi= ... (5)

[0038] S6. Calculate principal component scores.

[0039] Multiply the original data matrix X by the selected k eigenvector matrices E to obtain the principal component score matrix.

[0040] S7. Calculate the principal component composite score.

[0041] Based on the variance-explained contribution rate of each principal component, the overall principal component score expression is derived.

[0042]

[0043] Calculate the principal component composite score of the tobacco structure for each sample unit.

[0044] More preferably, the control center line and control upper and lower limits of the subject data are calculated according to the following formulas (6)-(11);

[0045] ... (6) ... (7)

[0046] ... (8)

[0047] ... (9)

[0048] ... (10)

[0049] ... (11)

[0050] in, The mean center line; This is the center line of the range; The upper limit is controlled by the mean. This is the lower limit of the mean control; This is the upper limit of the range control. X is the lower limit of range control, n is the sample size, and X is the lower limit of range control. i R is the mean of the sample. i D is the range of the sample, and D3, D4, and A2 are constants related to the sample size.

[0051] More preferably, the process capability index CP of the subject data is calculated according to the following formula (12);

[0052] ... (12)

[0053] Where USL is the upper limit of product specifications, LSL is the lower limit of product specifications, and σ is the standard deviation of the product. The process capability index CP can judge the level of process control for the structural stability of tobacco shreds; the higher the CP, the better the process control capability.

[0054] More preferably, control charts are drawn based on the subject data, including but not limited to mean-range control charts, mean-standard deviation control charts, and individual-moving range control charts.

[0055] Compared with the prior art, the present invention has the following beneficial technical effects:

[0056] The method of this invention utilizes the principle of principal component analysis to reduce the dimensionality of several indicators reflecting the structure of the formulated tobacco shreds and then uses a comprehensive score value to characterize them. The stability of the formulated tobacco shred structure is then evaluated using a process capability index. This method overcomes the drawbacks of single-indicator evaluation in existing technologies and facilitates the detection and evaluation of the formulated tobacco shred structure during the maintenance and processing of cigarette products. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0058] Figure 1 The correlation coefficients and significance matrix of the seven tobacco shred structure indicators are shown in the figure.

[0059] Figure 2 This is a single-value range control chart of the principal component composite score.

[0060] Figure 3 The process capability report chart is based on the principal component comprehensive score. Detailed Implementation

[0061] To better understand the above-mentioned objectives, features, and advantages of the present invention, the solutions of the present invention will be further described below. It should be noted that, unless otherwise specified, the embodiments of the present invention and the features thereof can be combined with each other.

[0062] Many specific details are set forth in the following description in order to provide a full understanding of the invention, but the invention may also be practiced in other ways different from those described herein; obviously, the embodiments in the specification are only some embodiments of the invention, and not all embodiments.

[0063] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0064] A method for evaluating the structural uniformity of cigarette formulation tobacco shreds includes the following steps:

[0065] 1. Cigarette sampling

[0066] After the cigarette rolling machine is running stably, cigarette samples are taken at a fixed position at the exit of the cigarette rolling machine. The samples are taken 30 times per batch of tobacco produced, with an interval of 90 seconds between each sampling and a sample quantity of 100 cigarettes. The samples are placed in a sealed bag as samples for determining the structure of the blended tobacco.

[0067] 2. Sample preparation

[0068] The sample was placed in an atmospheric environment with a temperature between 20 and 24°C and a relative humidity between 55 and 65% for 48 hours to adjust the moisture content. The adjusted cigarette was then longitudinally split open with a blade, and the tobacco shreds were taken out and gently scattered with tweezers to serve as the sample.

[0069] 3. Determine the structure of the formulated tobacco shreds in the sample.

[0070] S3.1 Weigh 30.0 (±0.5) g of tobacco sample, accurate to 0.01 g; place it in the center of the top screen of the planar rotating eccentric test sieve, fix the screen, start the test sieve, and begin the test.

[0071] S3.2 After the measurement is completed, remove the sieve and the non-porous base plate, weigh the tobacco on each layer of sieve and non-porous base plate, and record them as F1, F2, F3, F4, F5, F6 and non-porous base plate F0 in descending order of sieve aperture size, accurate to 0.01g, and clean the sieve.

[0072] S3.3 Repeat steps 3.1 to 3.2, and test five times. Record the proportion of tobacco on each layer of the screen as T. n ; Calculate the structure of the tobacco shreds in the cigarette formula according to the following formula (1);

[0073] ... (1)

[0074] In the formula:

[0075] T n —The proportion of tobacco shreds in n layers of sieves, accurate to 0.01%;

[0076] F n —The mass of tobacco shreds on the n-layer sieve, in grams (g);

[0077] n – the number of screens and non-perforated base layers arranged in descending order of aperture size, n=1,…,7.

[0078] S3.4 Based on the calculation results of S3.3, the average value of five test results of the proportion of tobacco of different sizes in the blended tobacco is obtained, with the data accurate to 0.01%.

[0079] Specifically, the parameter settings for the planar rotating eccentric testing sieve are as follows:

[0080] a. Rotation angle 800°, rotation speed 210r / min, interval time 2s, number of rotations 4.

[0081] b. The screen mesh sizes from top to bottom are 7.10mm, 4.50mm, 3.35mm, 2.00mm, 1.25mm, and 1.00mm; the proportion of tobacco shreds on each layer of screen mesh is recorded as T1, T2, T3, T4, T5, and T6 respectively, and the proportion of tobacco shreds on the non-perforated bottom plate is recorded as T7.

[0082] c. The testing sieve is first rotated clockwise 4 times, and then counterclockwise 4 times, with a 2-second pause after each 800° rotation.

[0083] S3.5 Test 30 samples to obtain the proportion of different size tobacco segments in each sample's tobacco blend. Perform data preprocessing to create a dataset, as shown in Table 1 below.

[0084] Table 1

[0085]

[0086] 4. Data standardization processing

[0087] S4.1 Given n tobacco samples, each sample has p tobacco structure indices, obtain the tobacco structure numerical matrix X. ij .

[0088]

[0089] like Figure 1 The figure shows the correlation coefficients and significance matrix among the seven tobacco structure indicators of the test samples. The results indicate that there is a certain correlation among all indicators. Significance analysis shows that, except for T3, the correlation between T3 and the other indicators is not significant; the correlation between T4 and T6 / T7 is not significant; the correlation between T1 and T7 is not significant; and the correlations among the other indicators are significant. This indicates that the correlations among the indicators are strong, making principal component analysis a suitable method.

[0090] S4.2. According to formula (2), the original data matrix of the tobacco shred structure is standardized with a mean of 0 and a standard deviation of 1 to eliminate the influence of dimensions and variance between different variables.

[0091] ... (2)

[0092] 5. Select principal components

[0093] S5.1 Calculate the covariance matrix

[0094] As shown in formula (3), the covariance matrix S is a P×P matrix, and its elements S ij The covariance between the i-th variable and the j-th variable is expressed by the following formula;

[0095] ... (3)

[0096] S5.2 Calculate the eigenvalues ​​and eigenvectors of the covariance matrix.

[0097] The covariance matrix S is decomposed into eigenvalues ​​λ1, λ2, λ3...λp and corresponding eigenvectors e1, e2...ep; the eigenvalue λi represents the variance of the i-th principal component and the eigenvector ei represents the direction of the i-th principal component.

[0098] S5.3 Determine the principal components

[0099] Based on the magnitude of the eigenvalues, select the first k principal components from the p original variables; calculate the cumulative variance contribution rate according to the following formula (4) so ​​that the cumulative variance contribution rate reaches the set threshold (the set threshold is greater than 80%).

[0100] κ(k) = ... (4)

[0101] Note: k is the number of principal components finally selected, and k is less than or equal to the number of original variables p.

[0102] When k=1, the cumulative variance contribution rate of F1 is 47%, which is less than 80%; when k=2, the cumulative variance contribution rate of F2 is 81%, which is greater than 80%. Therefore, the first two principal components are selected for subsequent analysis.

[0103] S5.4 Calculate the variance explanation contribution rate of the principal components.

[0104] Based on the proportion of variance contribution rate of each principal component explained in the first k cumulative variance contribution rates, the variance explained contribution rate of each principal component is calculated as shown in the following formula (5):

[0105] νi= ... (5)

[0106] The summary data is shown in Table 2 below.

[0107] Table 2

[0108]

[0109] It can be seen that the eigenvalues ​​of the two principal components are both greater than 1, and the cumulative variance contribution rate reaches 81%, which can reflect most of the information of the original variables.

[0110] 6. Calculate the principal component scores.

[0111] S6.1 Multiply the original data matrix X with the selected k eigenvector matrices E to obtain the principal component score matrix; the principal component score matrix is ​​an n×k matrix, where each column represents the score of a principal component.

[0112] Calculate the eigenvectors of the principal components and the original variable T, which are the score coefficients of each variable in the principal component scores. The data is shown in Table 3 below.

[0113] Table 3

[0114]

[0115] S6.2. Based on the eigenvectors of the principal components and the original variables, list the principal component score expressions.

[0116] First principal component:

[0117] F1=-0.26×T1-0.31×T2+0.32×T3+0.29×T4+0.11×T5+0.02×T6-0.02×T7

[0118] Second principal component:

[0119] F2=0.01×T1+0.06×T2-0.45×T3-0.05×T4+0.21×T5+0.33×T6+0.35×T7

[0120] 7. Calculate the principal component composite score.

[0121] Based on the variance explanation contribution rate of the principal components (see Table 2), a comprehensive score model for the structure of formulated tobacco shreds was established.

[0122] F = 0.58 × F1 + 0.42 × F2

[0123] The principal component composite score expression is:

[0124]

[0125] Based on the comprehensive scoring model of the tobacco shred structure, the principal component score of each sample unit was calculated, and the data is shown in Table 4 below.

[0126] Table 4

[0127]

[0128] 1. Based on the principal component composite scores of the 30 sample units, draw a single-value-range control chart.

[0129] Calculate the control center line and control upper and lower limits of the subject data as shown in the following formulas (6)-(11); see the individual value-range control chart. Figure 2 .

[0130] ... (6) ... (7)

[0131] ... (8)

[0132] ... (9)

[0133] ... (10)

[0134] ... (11)

[0135] in, The mean center line; This is the center line of the range; The upper limit is controlled by the mean. This is the lower limit of the mean control; This is the upper limit of the range control. X is the lower limit of range control, n is the sample size, and X is the lower limit of range control. i R is the mean of the sample. i D is the range of the sample, and D3, D4, and A2 are constants related to the sample size.

[0136] II. Drawing a process capability report graph of principal component composite scores

[0137] The process capability index CP of the topic data is calculated according to the following formula (12); the process capability report chart of the principal component comprehensive score is shown in [Figure number missing]. Figure 3 .

[0138] ... (12)

[0139] Where USL is the upper limit of product specifications, LSL is the lower limit of product specifications, and σ is the standard deviation of the product. The data is shown in Table 5 below.

[0140] Table 5

[0141]

[0142] The process capability index (CP) can determine the level of process control for the structural stability of tobacco shreds. A higher CP indicates better process control capability. As shown in Table 5, the process capability index (CP) is positively correlated with the product qualification rate (%); the higher the process capability index (CP), the higher the product qualification rate (%).

[0143] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the present invention. Although detailed descriptions have been provided with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments, and they should all be covered within the protection scope of the claims.

Claims

1. A method for evaluating the structural uniformity of tobacco shreds in cigarette formulations, characterized in that: Includes the following steps: S1. Cigarette sampling Cigarette samples were collected at the exit of the cigarette machine. Samples were taken once every time interval t, with m cigarettes taken each time, for a total of n times. S2. Sample preparation The sample was placed in an atmospheric environment with a certain temperature and humidity to adjust the moisture content; the tobacco shreds inside the cigarette were removed and scattered to serve as the sample. S3. Determine the structure of the formulated tobacco shreds in the sample. The planar rotary eccentric detection screen with n layers of screens is used to test the cut tobacco structure of the sample, and the proportion of the cut tobacco on each layer of screens is recorded as T n After multiple tests, the average value of the proportion of the cut tobacco of different sizes in the cut tobacco is taken, and the cut tobacco structure of the cigarette is calculated according to formula (1). ……（1） In the formula: T n - the proportion of tobacco in the n layer of the screen, to the nearest 0.01%; F n - the mass of tobacco on the n layer screen, in grams g; n — the number of screens and non-perforated base layers in descending order of aperture size; S4. Data Standardization Processing There are n tobacco samples, each sample has p tobacco structure indicators, and a tobacco structure value matrix X is obtained ij ; According to the following formula (2), X ij is standardized, that is, the value of each variable is subtracted by its mean and then divided by its standard deviation. …… （2） S5. Select principal components S5.1, Calculate the covariance matrix S as shown in equation (3); its elements Sijrepresent the covariance between the ith and jth variables. ij covariance between the ith and jth variables; …… （3） S5.2 Calculate the eigenvalues ​​and eigenvectors of the covariance matrix. The covariance matrix S is decomposed into eigenvalues ​​λ1, λ2, λ3...λp and corresponding eigenvectors e1, e2...ep; the eigenvalue λi represents the variance of the i-th principal component and the eigenvector ei represents the direction of the i-th principal component. S5.3 Determine the principal components Calculate the cumulative variance contribution rate of the p original variables according to the following formula (4), and select the first k components whose cumulative variance contribution rate reaches the set threshold as principal components. κ(k) = 1 - k …… (4) S5.4 Calculate the variance explanation contribution rate of the principal components. Based on the proportion of variance contribution rate of each principal component explained in the first k cumulative variance contribution rates, the variance explanation contribution rate of each principal component is calculated as shown in the following formula (5); vi= …… (5) S6. Calculate principal component scores. Multiply the original data matrix X by the selected k eigenvector matrices E to obtain the principal component score matrix; S7. Calculate the principal component composite score. Based on the variance-explained contribution rate of each principal component, the overall principal component score expression is derived. Calculate the principal component composite score of the tobacco structure for each sample unit.

2. The method for evaluating the uniformity of cigarette formula cut tobacco structure according to claim 1, characterized in that: In step S2, the temperature conditions are between 20 and 24°C, the relative humidity is between 55 and 65%, and the placement time is 48 hours.

3. The method for evaluating the uniformity of cigarette formula cut tobacco structure according to claim 2, characterized in that: In step S3, the planar rotating eccentric detection sieve has a total of 6 layers of screens, with aperture sizes from top to bottom of 7.10mm, 4.50mm, 3.35mm, 2.00mm, 1.25mm, and 1.00mm. The proportion of tobacco shreds on each layer of screens is recorded as T1, T2, T3, T4, T5, and T6, respectively, and the proportion of tobacco shreds on the non-perforated bottom plate is recorded as T7.

4. The method for evaluating the uniformity of cigarette formula cut tobacco structure according to claim 3, characterized in that: In step S4, the standardization process requires a mean of 0 and a standard deviation of 1.

5. The method for evaluating the uniformity of tobacco structure in cigarette formulations according to claim 4, characterized in that: In step S5.3, the threshold value of the cumulative variance contribution rate is greater than or equal to 80%.

6. The method for evaluating the uniformity of tobacco structure in cigarette formulations according to any one of claims 1-5, characterized in that: Calculate the control center line and control upper and lower limits of the subject data according to the following formulas (6)-(11); …… （6） …… （7） …… （8） …… （9） …… （10） …… （11） in, The mean center line; This is the center line of the range; The upper limit is controlled by the mean. This is the lower limit of the mean control; This is the upper limit of the range control. X is the lower limit of range control, n is the sample size, and X is the lower limit of range control. i R is the mean of the sample. i D is the range of the sample, and D3, D4, and A2 are constants related to the sample size.

7. The method for evaluating the uniformity of tobacco structure in cigarette formulations according to any one of claims 1-5, characterized in that: The process capability index CP of the subject data is calculated according to the following formula (12); …… （12） Where USL is the upper limit of product specifications, LSL is the lower limit of product specifications, and σ is the standard deviation of the product.

8. The method for evaluating the uniformity of tobacco structure in cigarette formulations according to any one of claims 1-5, characterized in that: Control charts are drawn based on the subject data, including mean-range control charts, mean-standard deviation control charts, and individual-moving range control charts.

Images