A single tobacco producing area and part identification method, alternative method and system

Abstract

The application discloses a single tobacco producing area and part identification method, a substitution method and a system, wherein the producing area and part identification method comprises the following steps: dividing single tobacco into multiple categories and setting category labels; constructing a training set; constructing a near-infrared spectrum matrix based on the near-infrared spectrum in the training set and performing standardization processing, calculating the covariance matrix of the standardized near-infrared spectrum matrix X; calculating the eigenvalue and eigenvector of the covariance matrix and performing sorting, taking the eigenvectors corresponding to the first k eigenvalues to form a dimension reduction matrix W; taking X*W and the corresponding category labels as training data to perform linear discriminant analysis classification training, and obtaining a single tobacco producing area and part identification model; and predicting the producing area and tobacco leaf part grade of a single tobacco to be predicted based on the single tobacco producing area and part identification model. The category label containing the single tobacco producing area and tobacco leaf part grade information can be predicted, the single tobacco producing area and part are accurately identified, and thus the accurate and efficient substitution of single tobacco in a formula leaf group is achieved.

Description

Technical Field

[0001] This invention relates to the field of cigarette and cigarette product quality testing technology, and in particular to a method, alternative method and system for identifying the production area and part of a single-material tobacco product. Background Technology

[0002] In the research, development, production, and maintenance of cigarettes, it is necessary to adjust the amount of different single-origin tobaccos used in a particular specification of cigarette based on the availability of tobacco leaves. Sometimes, it is necessary to substitute single-origin tobaccos with low inventory. Currently, cigarette formula maintenance personnel mainly use sensory evaluation to select substitute tobacco raw materials. This not only requires formula maintenance personnel to have a high level of sensory evaluation skills but also indirectly leads to a large workload and formula maintenance costs. However, with the development of analytical testing technology and computer intelligent algorithms, people can use different analytical testing methods and algorithms to extract data with common and dissimilar characteristics from single-origin tobaccos, enabling tobacco leaf grading, cigarette formula maintenance, and prediction of smoke component indicators.

[0003] Near-infrared spectroscopy is widely used in industry due to its speed, accuracy, and non-destructive nature. Near-infrared spectroscopy primarily records the overtones and combination frequencies of vibrations in hydrogen-containing groups X-H (X = C, N, O) that transition from the ground state to higher energy levels due to the anharmonicity of molecular vibrations. Different groups (such as methyl, methylene, and benzene rings) or the same group in different chemical environments exhibit significant differences in near-infrared absorption wavelengths and intensities. Near-infrared spectroscopy provides rich structural and compositional information, making it highly suitable for measuring the composition and properties of hydrocarbon organic substances.

[0004] Patent CN109975238A discloses "A Method for Substituting Tobacco Leaves and Cigarette Leaf Combinations Based on Near-Infrared Spectroscopy." Using near-infrared spectral information of tobacco leaves as the object, it discloses a method for substituting tobacco leaves and cigarette leaf combinations based on near-infrared spectral information. Through a locally corrected spectral preprocessing method and an integrated spectral similarity algorithm, it eliminates interference factors such as baseline and scattering in the spectrum, achieving a conversion from spectral similarity to tobacco leaf similarity, thus providing a method for tobacco leaf substitution in the process of assisting in the maintenance of cigarette leaf combinations. However, this method only seeks substitute tobacco leaves based on the similarity of tobacco leaf spectra, ignoring the stylistic differences inherent in tobacco leaves from different producing areas. Therefore, this method is unlikely to meet the needs of formulation maintenance personnel in practical applications.

[0005] Therefore, there is an urgent need for a technical solution that utilizes the near-infrared spectrum of tobacco leaves to quickly and objectively determine the aroma type, production area, and part of different single-origin tobaccos, so as to achieve accurate and efficient substitution of single-origin tobaccos in the blended leaf group. Summary of the Invention

[0006] This invention provides a method, substitution method, and system for identifying the production area and part of a single-origin tobacco plant. Its purpose is to use the near-infrared spectrum of tobacco leaves to quickly and objectively determine the aroma type, production area, and part of different single-origin tobacco plants, and to achieve accurate and efficient substitution of single-origin tobacco plants in a blend of tobacco leaves.

[0007] Firstly, a method for identifying the production area and part of a single-origin tobacco crop is provided, including:

[0008] Single-origin tobacco is divided into multiple categories based on the production area and grade of the tobacco leaf part. Each category has a corresponding label that includes information on the production area and grade of the tobacco leaf part.

[0009] Obtain the near-infrared spectra and corresponding class labels of several single-material tobacco products of various categories as a training set;

[0010] Construct a near-infrared spectral matrix of dimension n×p based on the near-infrared spectra in the training set, where n is the total number of near-infrared spectra and p is the dimension of the near-infrared spectra;

[0011] The near-infrared spectral matrix is ​​normalized to obtain the normalized near-infrared spectral matrix X;

[0012] Calculate the covariance matrix of the normalized near-infrared spectral matrix:

[0013] Calculate the eigenvalues ​​and corresponding eigenvectors of the covariance matrix;

[0014] Sort the eigenvalues ​​and their corresponding eigenvectors in descending order, and take the eigenvectors corresponding to the first k eigenvalues ​​to form a dimension-reduced matrix W with dimension p × k, where k is a preset value;

[0015] Using trainX=X*W as training data, and taking the corresponding class labels as parameters, we call the Matlab ClassificationDiscriminant.fit function to perform linear discriminant analysis classification training, and obtain the single-material tobacco production area and part identification model F;

[0016] Based on the single-origin tobacco production area and part identification model F, the production area and tobacco leaf part grade of the single-origin tobacco to be predicted are predicted: the near-infrared spectrum of the single-origin tobacco to be predicted is obtained and labeled to obtain testX, and the predict function of Matlab is called to obtain the predicted class label.

[0017] Furthermore, the process of calculating the covariance matrix of the normalized near-infrared spectral matrix includes:

[0018] Calculate the mean vector of the normalized near-infrared spectral matrix:

[0019]

[0020] Where, x i This represents the vector formed by the i-th row of the normalized near-infrared spectral matrix;

[0021] The covariance matrix S of the normalized near-infrared spectral matrix is ​​calculated using the following formula:

[0022]

[0023] Furthermore, the near-infrared spectral matrix is ​​standardized to obtain a standardized near-infrared spectral matrix X. The process includes:

[0024] The near-infrared spectral matrix is ​​Z-score normalized, and each element is normalized as follows:

[0025]

[0026] in, Let j be the standard deviation of the j-th variable. Let x represent the average value of the j-th variable, where x is absorbance. ij This represents the element in the i-th row and j-th column of the near-infrared spectral matrix.

[0027] Furthermore, the category label includes an integer part and a decimal part; the integer part represents the production area of ​​the single-origin tobacco; the decimal part ranges from 1 to 3, corresponding to the tobacco leaf part grade as upper leaf, middle leaf and lower leaf respectively.

[0028] Furthermore, the near-infrared spectrum of the single-element smoke was obtained using a near-infrared spectrometer with a spectral range of [12800 cm⁻¹]. -1 3600cm -1 [780nm, 2778nm]; scan speed range: [1 round / second, 64 rounds / second], number of scans per round range: [1, 128], resolution range: [2cm] -1 64cm -1 The total number of data points collected ranges from 1 to 2592.

[0029] Furthermore, the form of single-material tobacco includes bundles of tobacco, sheet tobacco, or tobacco powder.

[0030] Secondly, a single-origin tobacco production area and part identification system is provided, including at least one processor and at least one memory, wherein the memory stores a computer program, which, when loaded by the processor, executes the single-origin tobacco production area and part identification method as described above.

[0031] Thirdly, a method for replacing single-element tobacco is provided, including:

[0032] Construct a database of single-origin tobacco data, which includes the production area, leaf part grade, category label and corresponding near-infrared spectrum of single-origin tobacco;

[0033] Using the single-origin tobacco production area and part identification method described above, the category C that is to be replaced by single-origin tobacco is predicted;

[0034] In the single-material smoke data information database, the near-infrared spectral matrix of each single-material smoke labeled C is used to calculate the Euclidean distance with the near-infrared spectral matrix of the single-material smoke to be replaced.

[0035] Sort the Euclidean distances in ascending order, and select the top Z positions as candidate alternative single-material tobacco products in the database, where Z is a preset value.

[0036] Furthermore, the Euclidean distance d i The calculation formula is as follows:

[0037]

[0038] Where testX represents the near-infrared spectral matrix to be replaced by single-element smoke, X i This represents the near-infrared spectral matrix of a single-material smoke product with class label C in the single-material smoke product database.

[0039] Fourthly, a single-material smoke substitution system is provided, comprising at least one processor and at least one memory, wherein the memory stores a computer program that, when loaded by the processor, executes the single-material smoke substitution method as described above.

[0040] Beneficial effects

[0041] This invention proposes a method, substitution method, and system for identifying the production area and part of a single-origin tobacco plant. Based on the characteristic that tobacco leaves from different production areas and parts have different chemical compositions, resulting in different near-infrared spectra, and combined with linear discriminant analysis classification technology, it can predict class labels containing information on the production area and part of the single-origin tobacco plant, thereby achieving accurate identification of the production area and part of the single-origin tobacco plant. By comprehensively considering the search for candidate substitute single-origin tobacco plants based on three dimensions—production area, part of the tobacco plant, and near-infrared spectral similarity—the method takes into account the stylistic differences inherent in tobacco leaves from different production areas, thus meeting the needs of formula maintenance personnel. It can quickly and objectively reflect the differences between different tobacco leaves, achieving accurate and efficient substitution of single-origin tobacco plants in the formula leaf group, while also saving significant labor costs and effectively improving the efficiency of cigarette formula maintenance. Attached Figure Description

[0042] 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, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a flowchart of a method for identifying the production area and part of a single-origin tobacco plant, provided by an embodiment of the present invention;

[0044] Figure 2 This is a flowchart of a single-material smoke substitution method provided by an embodiment of the present invention;

[0045] Figure 3 This is a comparison image of near-infrared radiation and content of 2017-Yunnan Dali-C3F (red) and 2017-Yunnan Kunming-C4F (black) provided in an embodiment of the present invention;

[0046] Figure 4 This is a comparison image of near-infrared radiation and content of 2017-Yunnan Dali-C3F (red) and 2017-Yunnan Pu'er-C3L (black) provided in an embodiment of the present invention;

[0047] Figure 5 This is a comparison chart of near-infrared radiation and content of 2017-Yunnan Kunming-C4F (red) and 2016-Sichuan Panzhihua-C4F (black) provided in an embodiment of the present invention. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be described in detail below. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other implementation methods obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0049] Example 1

[0050] like Figure 1 As shown, this embodiment provides a method for identifying the production area and part of a single-origin tobacco crop, including:

[0051] S1: Single-origin tobacco is divided into multiple categories based on its production area and leaf part grade. Each category has a corresponding category label containing information on the production area and leaf part grade. In this embodiment, single-origin tobacco is divided into 25 categories based on production area and leaf part grade. The raw materials and category label information for each category are shown in the table below. The category label includes an integer part and a decimal part. The integer part ranges from 1 to 9, representing different production areas of single-origin tobacco. The decimal part ranges from 1 to 3, corresponding to the leaf part grades of upper leaf, middle leaf, and lower leaf. When the integer part is 9, it represents imported tobacco leaves, and no decimal part is set.

[0052]

[0053] S2: Obtain near-infrared spectra and corresponding class labels of several single-material tobacco products of various categories as a training set. In this embodiment, 74 single-material tobacco products and their corresponding near-infrared spectra (see table below) are randomly selected as the training set for linear discriminant analysis training.

[0054]

[0055]

[0056] S3: Construct a near-infrared spectral matrix of dimension n×p based on the near-infrared spectra in the training set, where n is the total number of near-infrared spectra and p is the dimension of the near-infrared spectra. In this embodiment, n is 74 and p is 1296.

[0057] S4: Normalize the near-infrared spectral matrix to obtain the normalized near-infrared spectral matrix X. n×p Specifically, this includes:

[0058] The near-infrared spectral matrix is ​​Z-score normalized, and each element is normalized as follows:

[0059]

[0060] in, Let j be the standard deviation of the j-th variable. Let x represent the average value of the j-th variable, where x is absorbance. ij This represents the element in the i-th row and j-th column of the near-infrared spectral matrix.

[0061] S5: Calculate the covariance matrix S of the normalized near-infrared spectral matrix. Specifically, this includes:

[0062] Calculate the mean vector of the normalized near-infrared spectral matrix:

[0063]

[0064] Where, x iThis represents the vector formed by the i-th row of the normalized near-infrared spectral matrix;

[0065] The covariance matrix S of the normalized near-infrared spectral matrix is ​​obtained by the following formula.

[0066]

[0067] S6: Calculate the eigenvalues ​​λ of the covariance matrix S according to the following formula. i and the corresponding feature vector v i .

[0068]

[0069] S7: Eigenvalue λ i and the corresponding feature vector v i The eigenvectors corresponding to the first k eigenvalues ​​are sorted in descending order to form a dimension-reduced matrix W with dimension p×k; in this embodiment, p is 48, so the dimension of the dimension-reduced matrix W is 1296×48.

[0070] S8: Use trainX = X*W as the training data, and call the Matlab ClassificationDiscriminant.fit function with the corresponding class labels as parameters to perform linear discriminant analysis classification training, and obtain the single-material tobacco production area and part identification model F: F = ClassificationDiscriminant.fit(trainX,trainid,'discrimType','linear').

[0071] S9: Based on the single-origin tobacco production area and part identification model F, predict the production area and tobacco leaf part grade of the single-origin tobacco to be predicted: Obtain the near-infrared spectrum of the single-origin tobacco to be predicted and perform annotation processing to obtain testX. Call the predict function of Matlab to obtain the prediction class label C: C = predict(F, testX*W).

[0072] During implementation, the near-infrared spectrum of the single-element smoke was acquired using a near-infrared spectrometer, the spectral range of which is: [12800cm]. -1 3600cm -1 [780nm, 2778nm]; scan speed range: [1 round / second, 64 rounds / second], number of scans per round range: [1, 128], resolution range: [2cm] -1 64cm -1 The total number of data points collected ranges from [1, 2592].

[0073] It should be noted that the form of single-material smoke includes, but is not limited to, bundled smoke, sheet smoke, or smoke powder. The dimensions of the infrared spectrum and the number of training set samples can be selected according to actual needs.

[0074] According to the method provided in this embodiment, the grade of tobacco leaves from 30 different production areas and leaf parts was predicted. The accuracy of the results was 80% compared with the manually added classification mark. The details of the prediction results are shown in the table below.

[0075]

[0076]

[0077] Example 2

[0078] This embodiment provides a single-origin tobacco production area and part identification system, including at least one processor and at least one memory. The memory stores a computer program, which, when loaded by the processor, executes the single-origin tobacco production area and part identification method as described in Embodiment 1.

[0079] Example 3

[0080] like Figure 2 As shown, this embodiment provides a method for replacing single-material smoke, including:

[0081] S21: Construct a database of single-origin tobacco data, which includes the production area, leaf part grade, category label, and corresponding near-infrared spectrum of single-origin tobacco.

[0082] S22: The class C to be replaced by the single-origin tobacco production area and part identification method as described in Example 1 is predicted.

[0083] S23: In the single-material smoke data information database, calculate the Euclidean distance between the near-infrared spectral matrix of each single-material smoke labeled C and the near-infrared spectral matrix of the single-material smoke to be replaced; the Euclidean distance d i The calculation formula is as follows:

[0084]

[0085] Where testX represents the near-infrared spectral matrix to be replaced by single-element smoke, X i This represents the near-infrared spectral matrix of a single-material smoke product with class label C in the single-material smoke product database.

[0086] S24: Sort the Euclidean distances in ascending order, and remove the top Z positions as candidate alternative single-material smokes in the database, where Z is a preset value, and in this embodiment, 3 is removed.

[0087] Using "2017-Yunnan Dali-C3F" as the input for the proposed alternative single-material tobacco, and based on the method provided in this embodiment, the recommended candidate results are as follows:

[0088] 1st Place: 2017-Kunming, Yunnan-C4F;

[0089] 2nd place: 2017-Yunnan Pu'er-C3L;

[0090] 3rd place: 2016-Panzhihua, Sichuan-C4F.

[0091] To verify the accuracy of the recommendations, the near-infrared spectra of the proposed substitute single-element tobacco and the candidate single-element tobacco were compared, as well as the contents of the main chemical components (total sugar, reducing sugar, total alkali, total nitrogen, and chlorine). The results are as follows: Figure 3 , Figure 4 and Figure 5 As shown. The results indicate that the recommended single-material tobacco obtained according to the method of this embodiment is very similar in composition to the proposed alternative single-material tobacco.

[0092] Example 4

[0093] This embodiment provides a single-material smoke substitution system, including at least one processor and at least one memory. The memory stores a computer program, which, when loaded by the processor, executes the single-material smoke substitution method as described in Embodiment 3.

[0094] It is understood that the same or similar parts in the above embodiments can be referred to each other, and the contents not described in detail in some embodiments can be referred to the same or similar contents in other embodiments.

[0095] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0096] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for identifying the production area and part of a single-origin tobacco plant, characterized in that, include: Single-origin tobacco is divided into multiple categories based on the production area and grade of the tobacco leaf part. Each category has a corresponding label that includes information on the production area and grade of the tobacco leaf part. Obtain the near-infrared spectra and corresponding class labels of several single-material tobacco products of various categories as a training set; Construct a near-infrared spectral matrix of dimension n×p based on the near-infrared spectra in the training set, where n is the total number of near-infrared spectra and p is the dimension of the near-infrared spectra; The near-infrared spectral matrix is ​​normalized to obtain the normalized near-infrared spectral matrix. ; Calculate the covariance matrix of the normalized near-infrared spectral matrix: Calculate the eigenvalues ​​and corresponding eigenvectors of the covariance matrix; Sort the eigenvalues ​​and their corresponding eigenvectors in descending order, and take the eigenvectors corresponding to the first k eigenvalues ​​to form a dimension-reduced matrix W with dimension p×k; Using trainX=X*W as training data, and taking the corresponding class labels as parameters, we call the Matlab ClassificationDiscriminant.fit function to perform linear discriminant analysis classification training, and obtain the single-material tobacco production area and part identification model F; Based on the single-origin tobacco production area and part identification model F, the production area and tobacco leaf part grade of the single-origin tobacco to be predicted are predicted: the near-infrared spectrum of the single-origin tobacco to be predicted is obtained and standardized to obtain testX. The predict function of Matlab is called to obtain the prediction class label C: C=predict(F, testX*W). The category label includes an integer part and a decimal part; the integer part represents the production area of ​​single-origin tobacco; the decimal part ranges from 1 to 3, corresponding to the grade of tobacco leaf parts as upper leaf, middle leaf and lower leaf respectively.

2. The method for identifying the production area and part of a single-origin tobacco plant according to claim 1, characterized in that, The process of calculating the covariance matrix of the normalized near-infrared spectral matrix includes: Calculate the mean vector of the normalized near-infrared spectral matrix: ； in, The first element of the normalized near-infrared spectral matrix i A vector formed by rows; The covariance matrix S of the normalized near-infrared spectral matrix is ​​calculated using the following formula: 。 3. The method for identifying the production area and part of a single-origin tobacco plant according to claim 1, characterized in that, The near-infrared spectral matrix is ​​then normalized to obtain a normalized near-infrared spectral matrix. The process includes: The near-infrared spectral matrix is ​​Z-score normalized, and each element is normalized as follows: ； in, Let j be the standard deviation of the j-th variable. Let represent the average value of the j-th variable, where the variable is absorbance. This represents the element in the i-th row and j-th column of the near-infrared spectral matrix.

4. The method for identifying the production area and part of a single-origin tobacco plant according to claim 1, characterized in that, The near-infrared spectrum of the single-element smoke was obtained using a near-infrared spectrometer with a spectral range of [12800 cm⁻¹]. -1 3600 cm -1 [780 nm, 2778 nm]; scan speed range: [1 round / second, 64 rounds / second], number of scans per round range: [1, 128], resolution range: [2 cm] -1 64cm -1 The total number of data points collected ranges from 1 to 2592.

5. The method for identifying the production area and part of a single-origin tobacco plant according to claim 1, characterized in that, The forms of single-material tobacco include bundles of tobacco, sheet tobacco, or tobacco powder.

6. A system for identifying the production area and part of a single-origin tobacco crop, characterized in that, It includes at least one processor and at least one memory, wherein the memory stores a computer program that, when loaded by the processor, executes the single-origin tobacco production area and part identification method as described in any one of claims 1 to 5.

7. A method for replacing single-element smoke, characterized in that, include: Construct a database of single-origin tobacco data, which includes the production area, leaf part grade, category label and corresponding near-infrared spectrum of single-origin tobacco; The method for identifying the production area and part of single-origin tobacco as described in any one of claims 1 to 5 is used to predict the category C of the single-origin tobacco to be replaced; In the single-material smoke data information database, the near-infrared spectral matrix of each single-material smoke labeled C is used to calculate the Euclidean distance with the near-infrared spectral matrix of the single-material smoke to be replaced. Sort the Euclidean distances in ascending order, and take the top Z positions as candidate alternative single-material tobacco products in the database, where Z is a preset value.

8. The method for replacing single-material smoke according to claim 7, characterized in that, The Euclidean distance The calculation formula is as follows: ； in, This represents the near-infrared spectral matrix of the material intended to replace single-element smoke. This represents the near-infrared spectral matrix of single-material smoke with class label C in the single-material smoke data information database.

9. A single-element smoke replacement system, characterized in that, It includes at least one processor and at least one memory, the memory storing a computer program that, when loaded by the processor, executes the single-material tobacco substitution method as described in claim 7 or 8.

Images