A method for representing the comprehensive quality of threshed and redried finished tobacco strips

Abstract

The application discloses a kind of to beat leaf redrying finished product sheet tobacco comprehensive quality characterization method, comprising:1 collection the leaf redrying quality index of the nth processing batch,2 to leaf redrying quality index sequence is pretreated,3 using KMO test and Bartlett sphericity test respectively to the applicability of standardized data verification,4 calculate the correlation coefficient matrix of all processing batch leaf redrying quality index,5 calculate the eigenvalue and unit eigenvector of correlation coefficient matrix,6 calculate cumulative variance contribution rate, determine the number of principal components and calculate principal component score,7 calculate the nth processing batch to beat leaf redrying sheet tobacco finished product comprehensive quality score.The application can quantitatively, objectively, accurately characterize the comprehensive quality of beat leaf redrying finished product sheet tobacco, so as to solve the deficiency that traditional evaluation index is single and relatively rough.

Description

Technical Field

[0001] This invention relates to the field of tobacco processing technology, specifically to a characterization method for multi-dimensional quantitative evaluation of the overall quality of finished tobacco leaves during the leaf threshing and re-drying process. Background Technology

[0002] Re-drying tobacco leaves is a crucial step in enhancing the value of raw tobacco materials, and its quality directly impacts the stability of subsequent processes such as tobacco processing and cigarette making, as well as the final quality of cigarette products. Currently, the evaluation of the quality of re-dried tobacco leaves relies heavily on the detection and analysis of multiple independent indicators, including packing density deviation rate, impurity content, moisture content coefficient of variation, nicotine coefficient of variation, core temperature qualification rate, packing weight qualification rate, leaf structure (such as the proportion of medium-sized leaves, large and medium-sized leaves, and fragmentation rate), and stem content in the leaves.

[0003] However, existing technologies typically employ a simple comparison of indicators, where "meeting the standard means passing," lacking a method that can integrate multiple interrelated or even contradictory indicators, scientifically allocate weights, and ultimately output a comprehensive quality score that can be compared horizontally and tracked vertically. This single-indicator approach makes it difficult to scientifically, comprehensively, and quantitatively assess the impact of leaf re-drying process parameters on overall quality, hindering the optimization of overall production parameters and lean control of process quality.

[0004] Although similar characterization methods have been disclosed in the prior art, such as a characterization method for airflow drying processing intensity (publication number CN111358031B); a characterization method for thin plate drying processing intensity (publication number CN111238994B); a quantitative characterization method for tobacco processing line processing intensity (publication number CN110301664B); a characterization method for tobacco leaf and shred breakage during cigarette processing (publication number CN114609006B); a method for characterizing shred length characteristic values ​​by coupling shred skeleton length and shred distribution (publication number CN118518823A); and a method for characterizing the dryness sensation of cigarette smoking (publication number CN118032854A), none of these are used for finished sheet tobacco after leaf re-drying.

[0005] Therefore, there is an urgent need to establish a multi-dimensional and quantitative comprehensive evaluation method to achieve accurate characterization of the overall quality of finished tobacco products. Summary of the Invention

[0006] To overcome the shortcomings of the existing technologies, this invention provides a method for characterizing the overall quality of finished tobacco leaves after threshing and re-drying. This method is scientific, provides intuitive results, and is easy to operate. By integrating multiple key quality indicators and employing data preprocessing, correlation testing, and principal component analysis, a multi-dimensional comprehensive evaluation model is constructed. Ultimately, the model outputs a quantitative score that accurately characterizes the overall quality of finished tobacco leaves after threshing and re-drying under different batches and parameter conditions, providing clear data support for production optimization.

[0007] The technical solution adopted by this invention to solve the technical problem is as follows: The method for characterizing the overall quality of finished tobacco leaves after pounding and re-drying, as described in this invention, is characterized by the following steps: Step 1: Collect the quality index sequence Z of the re-drying of the leaves in the nth batch of processing. n ={A n,i |i=1,2,…,11},A n,i This represents the re-drying quality index of the i-th blade in the nth processing batch. When i=1, A n,i A represents the packing density deviation rate of the nth processing batch. When i=2, A n,i This represents the impurity content of the nth processing batch. When i=3, A n,i A represents the coefficient of variation of moisture content in the nth processing batch. When i=4, A n,i A represents the coefficient of variation of nicotine in the nth processing batch. When i=5, A n,i This represents the core temperature pass rate of the nth processing batch. When i=6, A n,i This represents the packing weight pass rate of the nth processing batch. When i=7, A n,i This represents the uniformity of the blade structure in the nth processing batch; when i=8, A n,i This represents the medium-sized piece rate of the nth processing batch; when i=9, A n,i This represents the proportion of large and medium-sized pieces in the nth processing batch; when i=10, A n,i A represents the breakage rate of the nth processing batch. When i=11, A n,i This represents the percentage of stems in the leaves of the nth processing batch; n = 1, 2, ..., N, where N is the total number of processing batches; Step 2: For Z n Preprocessing was performed to obtain the sequence of quality indicators for the re-drying of pretreated blades in the nth batch of processing. n ={ n,i |i=1,2,…,11}, where, n,i This represents the re-drying quality index of the i-th blade after standardization in the nth processing batch; Step 3: Use the KMO test and Bartlett's test of sphericity to evaluate the standardized index set { The applicability of the dataset |n=1,2,…,N;i=1,2,…,11} is verified to obtain the KMO test value. and sphericity test value ; If v is greater than the set first threshold And p is less than the set first threshold. If the result is satisfactory, proceed to step 4; otherwise, return to step 1 and re-collect the data. ∈(0.6,1.0), ∈(0,0.05); Step 4: For { Principal component analysis was performed on |n=1,2,…,N;i=1,2,…,11} to obtain the scores of the leaf re-drying quality index for all processing batches; Step 5: Calculate the comprehensive quality score D of the leaf re-drying quality index under the nth processing batch condition according to formulas (11) and (12). n : (11) (12) In equations (11) and (12), This represents the k-th feature value after filtering. This is the kth weight.

[0008] The characteristic of the method for characterizing the overall quality of finished tobacco leaves after pounding and re-drying according to the present invention is that step 2 includes the following steps: Step 2.1: Calculate the packing density deviation rate A of the nth processing batch. n,1 Moisture content variation coefficient A n,3 nicotine coefficient of variation A n,4 7. Blade structure uniformity An, fragmentation rate A n,10 Stem content A in leaves n,11 Let A be any j-th pointer in the array. n,j ; Using equation (1) for A n,j Perform a forward conversion to obtain the j-th index after the forward conversion of the nth processing batch. : =100%-A n,j (1) Using formula (2) to determine the impurity content A n,2 Perform a forward conversion to obtain the impurity content of the nth processing batch after forward conversion. : =1 / A n,2 (2) Step 2.2: { |j=1,3,4,7,10,11}、 A n,5 A n,6 A n,8 A n,9 any one of the middle Individual marker is B n,i Thus, equations (3)-(5) are used to analyze B. n,i Standardization is performed to obtain the standardized form of the nth processing batch. Individual indicators : (3) (4) (5) In equations (3)-(5), B n,i The mean across all processing batches. B n,i Standard deviation across all processing batches.

[0009] Furthermore, step 4 includes the following steps: Step 4.1: Calculate the correlation coefficient matrix of the re-drying quality indicators of all processed batches of blades according to formula (6). : (6) In equation (6), ∈(0,1) is The standardized batch number after the nth processing batch Individual indicators The correlation coefficient; Step 4.2: Calculate the correlation coefficient matrix according to equation (7). The eigenvalue sequence λ={ ,in, The characteristic value of the i-th indicator is: =0 (7) In equation (7), I is an 11th-order identity matrix; Step 4.3: Calculate the feature vector sequence α of the quality index of leaf re-drying for all processing batches according to formula (8), and normalize α to obtain the normalized feature vector sequence. ={ ,in, The feature vector representing the i-th index of all processing batches: (λI - R)×α= 0 (8) Step 4.4: Calculate the cumulative variance contribution rate C of the first t characteristic values ​​of the re-drying quality index of all processing batches of blades according to formula (9). t Thus, 11 cumulative variance contribution rates are obtained: (9) In equation (9), This represents the m-th eigenvalue; Select the 11 cumulative variance contribution rates that satisfy λ m ≥1 and C t The feature value corresponding to ≥δ is used to determine the number of feature values ​​after screening, which is K. The score F of the kth feature value of the leaf re-drying quality index in the nth processing batch is calculated according to formula (10). n,k k=1,2….,K, where δ represents the contribution rate threshold: (10) In equation (10), For all processing batches, the first The k-th feature value of an indicator.

[0010] The present invention provides an electronic device, including a memory and a processor, characterized in that the memory is used to store a program supporting the processor in performing the method described therein, and the processor is configured to execute the program stored in the memory.

[0011] The present invention discloses a computer-readable storage medium storing a computer program, characterized in that the computer program is executed by a processor to perform the steps of the method described thereon.

[0012] Compared with existing technologies, the beneficial effects of this invention are reflected in: 1. This invention comprehensively characterizes the quality of finished tobacco leaves from multiple dimensions, including physical, chemical, control, and structural aspects, thus overcoming the shortcomings of traditional evaluation indicators that are singular and relatively crude.

[0013] 2. This invention uses principal component analysis to objectively determine the weight of each indicator in the comprehensive score, avoiding the subjectivity of human weighting and making the comprehensive evaluation results more objective and scientific.

[0014] 3. This invention can satisfy the comprehensive quality characterization of finished tobacco sheets between different processing batches, as well as the comprehensive quality characterization of finished tobacco sheets under different production conditions.

[0015] 4. This invention has the advantages of scientific evaluation, intuitive results, and simple operation. It can quantitatively, objectively, and accurately characterize the overall quality of finished tobacco leaves after leaf re-drying. Attached Figure Description

[0016] Figure 1 This is a diagram of the index system for the comprehensive quality characterization method of finished tobacco leaves after leaf re-drying according to the present invention; Figure 2 This is a schematic diagram of the workflow of the comprehensive quality characterization method for finished tobacco leaves after re-drying according to the present invention. Detailed Implementation

[0017] In this embodiment, a method for characterizing the overall quality of finished tobacco leaves after re-drying involves comprehensively characterizing 11 indicators: packing density deviation rate, impurity content, moisture content coefficient of variation, nicotine coefficient of variation, core temperature qualification rate, packing weight qualification rate, leaf structure uniformity, medium leaf rate, large and medium leaf rate, fragment rate, and stem content in the leaves. Figure 1 and Figure 2 As shown, it includes the following steps: Step 1: Systematically collect the re-drying quality index sequence Z of the nth batch of tobacco leaves according to the industry standard methods (YC / T 147-2023 "Quality Requirements for Trimmed Tobacco Leaves" and YC / T 366-2010 "Evaluation of Uniformity of Flue-cured Tobacco Leaves"). n ={A n,i |i=1,2,…,11},A n,i This represents the re-drying quality index of the i-th blade in the nth processing batch. When i=1, A n,i A represents the packing density deviation rate of the nth processing batch. When i=2, A n,i This represents the impurity content of the nth processing batch. When i=3, A n,i A represents the coefficient of variation of moisture content in the nth processing batch. When i=4, A n,i A represents the coefficient of variation of nicotine in the nth processing batch. When i=5, A n,i This represents the core temperature pass rate of the nth processing batch. When i=6, A n,i This represents the packing weight pass rate of the nth processing batch. When i=7, A n,i This represents the uniformity of the blade structure in the nth processing batch; when i=8, A n,i This represents the medium-sized piece rate of the nth processing batch; when i=9, A n,i This represents the proportion of large and medium-sized pieces in the nth processing batch; when i=10, A n,i A represents the breakage rate of the nth processing batch. When i=11, A n,i This represents the percentage of stems in the leaves of the nth processing batch; n = 1, 2, ..., N, where N is the total number of processing batches.

[0018] In this embodiment, the packing density deviation rate (A1, %) reflects the uniformity of tobacco distribution in the finished tobacco box; the smaller the value, the better. Impurity content (A2, pieces / 10 boxes): reflects the control level of non-tobacco substances in the finished tobacco products; the lower the value, the better. Moisture content variation coefficient (A3, %): reflects the uniformity of moisture content of tobacco within a batch; the smaller the value, the better. Nicotine coefficient of variation (A4, %): reflects the uniformity of nicotine content in tobacco leaves within a batch; the smaller the value, the better. Box core temperature pass rate (A5, %): reflects the stability of finished box core temperature control; Packing weight pass rate (A6, %): reflects the accuracy of finished cigarette box weight control; Blade structure uniformity (A7, %): reflects the uniformity of blade size distribution; the smaller the value, the better.

[0019] Medium sheet rate (A8, %): The percentage of tobacco sheets with a size of 12.7mm×12.7mm to 25.4mm×25.4mm; Large and medium sheet rate (A9, %): The percentage of tobacco sheets with a size greater than 12.7mm × 12.7mm; Fragmentation rate (A) 10 %, %: The percentage of fragments smaller than 2.36mm × 2.36mm; the lower the value, the better. Stem content in leaves (A) 11 %,: The percentage of tobacco stems that could not be effectively separated from the leaves; the smaller the value, the better. Step 2: To make indicators with different dimensions and directions comparable and capable of comprehensive calculations, it is necessary to perform Z... n Preprocessing is performed, including positive conversion of negative indicators and standardization of all indicators, to eliminate the influence of dimensions and orders of magnitude, thereby obtaining the sequence of quality indicators for the re-drying of leaves after preprocessing in the nth batch of processing. n ={ n,i |i=1,2,…,11}, where, n,i This represents the re-drying quality index of the i-th blade after standardization in the nth processing batch; Step 2.1: For negative indicators that "smaller values ​​represent better quality," such as packing density deviation rate A1, moisture content coefficient of variation A3, nicotine coefficient of variation A4, leaf structure uniformity A7, and fragmentation rate A... 10 Stem content A in leaves 11A positive conversion of the indicators is required; that is, the packing density deviation rate A of the nth processing batch. n,1 Moisture content variation coefficient A n,3 nicotine coefficient of variation A n,4 7. Blade structure uniformity An, fragmentation rate A n,10 Stem content A in leaves n,11 Let A be any j-th pointer in the array. n,j ; Using equation (1) for A n,j Perform a forward conversion to obtain the j-th index after the forward conversion of the nth processing batch. : =100%-A n,j (1) Using formula (2) to determine the impurity content A n,2 Perform a forward conversion to obtain the impurity content of the nth processing batch after forward conversion. : =1 / A n,2 (2) Step 2.2: To eliminate the influence of different indicator units and orders of magnitude, the data needs to be standardized. Therefore, { |j=1,3,4,7,10,11}、 A n,5 A n,6 A n,8 A n,9 any one of the middle Individual marker is B n,i Thus, equations (3)-(5) are used to analyze B. n,i Standardization is performed to obtain the standardized form of the nth processing batch. Individual indicators : (3) (4) (5) In equations (3)-(5), B n,i The mean across all processing batches. B n,i Standard deviation across all processing batches; Step 3: Applicability Verification The KMO test and Bartlett's test of sphericity were used to evaluate the standardized index set { The applicability of the dataset |n=1,2,…,N;i=1,2,…,11} is verified to obtain the KMO test value. and sphericity test value ; If v is greater than the set first threshold And p is less than the set first threshold. If the data is satisfactory, it indicates that the data is suitable for principal component analysis. Step 4 then proceeds with the calculation of the correlation coefficient matrix, eigenvalues, unit eigenvectors, and cumulative variance contribution rate. Based on preset criteria, the number of principal components and the weights of each principal component are determined. Otherwise, the process returns to step 1 for re-collection. ∈(0.6,1.0), ∈(0,0.05); Step 4: For { Principal component analysis was performed on |n=1,2,…,N;i=1,2,…,11} to obtain the scores of the leaf re-drying quality index for all processing batches; Step 4.1: Calculate the correlation coefficient matrix of the re-drying quality indicators of all processed batches of blades according to formula (6). : (6) In equation (6), ∈(0,1) is The standardized batch number after the nth processing batch Individual indicators The correlation coefficient; Step 4.2: Calculate the correlation coefficient matrix according to equation (7). The eigenvalue sequence λ={ ,in, The characteristic value of the i-th indicator is: =0 (7) In equation (7), I is an 11th-order identity matrix; Step 4.3: Calculate the feature vector sequence α of the re-drying quality index of all processed batches of blades according to formula (8), and normalize α (i.e., divide the vector by its own magnitude) to obtain the normalized feature vector sequence. ={ ,in, The feature vector representing the i-th index of all processing batches: (λI - R)×α= 0 (8) Step 4.3: Calculate the cumulative variance contribution rate C of the first t characteristic values ​​of the re-drying quality index of all processing batches of blades according to formula (9). t Thus, 11 cumulative variance contribution rates are obtained: (9) In equation (9), This represents the m-th eigenvalue; Select the 11 cumulative variance contribution rates that satisfy λ m ≥1 and C t The feature value corresponding to ≥δ is used to determine the number of feature values ​​after screening, which is K. The score F of the kth feature value of the leaf re-drying quality index in the nth processing batch is calculated according to formula (10). n,k k=1,2….,K, where δ represents the contribution rate threshold: (10) In equation (10), For all processing batches, the first The k-th characteristic value of an indicator; Step 5: Based on the principal component scores and weights, calculate the comprehensive quality score D of the leaf re-drying quality index under the nth processing batch condition according to equations (11) and (12). n .

[0020] (11) (12) In equations (11) and (12), This represents the k-th feature value after filtering. The k-th weight is used to ensure that the overall quality score reflects the importance of each principal component.

[0021] Based on the comprehensive score evaluation of each processing batch, the comprehensive quality level of different processing batches is compared. The higher the score, the better the comprehensive quality. Therefore, the comprehensive quality scores under different processing batch conditions are arranged in descending order. The higher the comprehensive score, the higher the comprehensive quality of the finished tobacco leaves after leaf re-drying under that processing batch condition, and it is regarded as the optimal processing batch.

[0022] In this embodiment, an electronic device includes a memory and a processor. The memory stores a program that supports the processor in executing the above-described method, and the processor is configured to execute the program stored in the memory.

[0023] In this embodiment, a computer-readable storage medium stores a computer program, which is executed by a processor to perform the steps of the above method.

[0024] The specific embodiments of the present invention will now be described in further detail with reference to the accompanying drawings.

[0025] A re-drying plant conducted a comprehensive quality evaluation of three batches of finished tobacco leaves after re-drying. The process followed the steps outlined above: (1) Data collection: Data was collected for 11 indicators under the conditions of 3 processing batches, including packing density deviation rate, impurity content, moisture content coefficient of variation, nicotine coefficient of variation, core temperature qualification rate, packing weight qualification rate, blade structure uniformity, medium blade rate, large and medium blade rate, fragmentation rate, and stem content in the leaf, as shown in Table 1.

[0026] Table 1 Summary of test data for each processing batch

[0027] (2) Data preprocessing: The original data of the three batches of indicators are preprocessed according to formulas (1) to (2), including the positive conversion of negative indicators (Table 2) and standardization processing (Table 3).

[0028] Table 2 Negative Indicator Conversion Data

[0029] Table 3. Data after standardization

[0030] (3) Applicability verification: The standardized data were subjected to KMO test and Bartlett's sphericity test according to formulas (3) to (5). The KMO test value v = 0.76 ∈ (0.6, 1.0), indicating that the data is suitable for principal component analysis; the Bartlett's sphericity test χ² = 332.58, p value = 0.001 (<0.05), indicating that there is a significant correlation between the indicators, which meets the premise of principal component analysis. (4) Principal component analysis: ① Using the standardized data in Table 3, construct the correlation coefficient matrix R according to formula (6), as shown in Table 4; Table 4. Correlation Coefficient Matrix of Indicators R

[0031] ② Solve for the eigenvalues ​​λ of the correlation coefficient matrix R according to equations (7) and (8). i and eigenvector α i , α i Normalized to α i The variance contribution rate and cumulative variance contribution rate are calculated according to formula (9), as shown in Tables 5 and 6. Table 5. Eigenvalues, variance contribution rate, and cumulative variance contribution rate

[0032] Note: Since the cumulative variance contribution rate of the first 6 principal components is close to 100%, the analysis results of the last 5 principal components are omitted.

[0033] Table 6 Principal Component Eigenvector Matrix

[0034] ③ Based on the dual criteria (λm≥1 and cumulative variance contribution rate≥85%), determine the number of principal components K=3 (cumulative variance contribution rate 86.72%), and calculate the principal component scores according to formula (10), see Table 7; Table 7 Principal Component Scores for Each Processing Batch

[0035] (5) Comprehensive quality characterization: ① The weights of the three principal components are calculated according to formula (11), and the results are as follows: The weight of principal component 1, w1, is calculated as follows: w1 = 4.88 / (4.88 + 2.65 + 2.01) = 4.88 / 9.54 ≈ 0.512. The weight of principal component 2, w2, is approximately 0.278 (2.65 / 9.54). The weight of principal component 3, w3, is approximately 0.210 (2.01 / 9.54). ② Calculate the overall quality score of each processing batch according to formula (12), and the results are as follows: Calculate the overall quality score (D) for each processing batch. n =w1×F1+w2×F2+w3×F3) Batch 1: D1 = 0.512 × (-1.291) + 0.278 × 0.90 + 210 × (-0.33) ≈ -0.549; Batch 2: D2 = 0.512 × 0.35 + 0.278 × 0.13 + 0.210 × 0.49 ≈ 0.318; Batch 3: D3=512×0.94+0.278×(-1.03)+0.210×(-0.16)≈0.161.

[0036] According to D n Based on the principle that the higher the overall quality, the better, the overall quality scores of the processed batches of finished tobacco leaves after pounding and re-drying are ranked as follows: Batch 2 > Batch 3 > Batch 1. Therefore, Batch 2 is determined to be the batch of finished tobacco leaves after pounding and re-drying with the best overall quality.

Claims

1. A method for characterizing the overall quality of finished tobacco leaves after threshing and re-drying, characterized in that, Includes the following steps: Step 1: Collect the quality index sequence Z of the re-drying of the leaves in the nth batch of processing. n ={A n,i |i=1,2,…,11},A n,i This represents the re-drying quality index of the i-th blade in the nth processing batch. When i=1, A n,i A represents the packing density deviation rate of the nth processing batch. When i=2, A n,i This represents the impurity content of the nth processing batch. When i=3, A n,i A represents the coefficient of variation of moisture content in the nth processing batch. When i=4, A n,i A represents the coefficient of variation of nicotine in the nth processing batch. When i=5, A n,i This represents the core temperature pass rate of the nth processing batch. When i=6, A n,i This represents the packing weight pass rate of the nth processing batch. When i=7, A n,i This represents the uniformity of the blade structure in the nth processing batch; when i=8, A n,i This represents the medium-sized piece rate of the nth processing batch; when i=9, A n,i This represents the proportion of large and medium-sized pieces in the nth processing batch; when i=10, A n,i A represents the breakage rate of the nth processing batch. When i=11, A n,i This represents the percentage of stems in the leaves of the nth processing batch; n = 1, 2, ..., N, where N is the total number of processing batches; Step 2: For Z n Preprocessing was performed to obtain the sequence of quality indicators for the re-drying of pretreated blades in the nth batch of processing. n ={ n,i |i=1,2,…,11}, where, n,i This represents the re-drying quality index of the i-th blade after standardization in the nth processing batch; Step 3: Use the KMO test and Bartlett's test of sphericity to evaluate the standardized index set { The applicability of the dataset |n=1,2,…,N;i=1,2,…,11} is verified to obtain the KMO test value. and sphericity test value ; If v is greater than the set first threshold And p is less than the set first threshold. If the result is satisfactory, proceed to step 4; otherwise, return to step 1 and re-collect the data. ∈(0.6,1.0), ∈(0,0.05); Step 4: For { Principal component analysis was performed on |n=1,2,…,N;i=1,2,…,11} to obtain the scores of the leaf re-drying quality index for all processing batches; Step 5: Calculate the comprehensive quality score D of the leaf re-drying quality index under the nth processing batch condition according to formulas (11) and (12). n : (11) (12) In equations (11) and (12), This represents the k-th feature value after filtering. This is the kth weight.

2. The method for characterizing the overall quality of finished tobacco leaves after threshing and re-drying, as described in claim 1, is characterized in that... Step 2 includes the following steps: Step 2.1: Calculate the packing density deviation rate A of the nth processing batch. n,1 Moisture content variation coefficient A n,3 nicotine coefficient of variation A n,4 7. Blade structure uniformity An, fragmentation rate A n,10 Stem content A in leaves n,11 Let A be any j-th pointer in the array. n,j ; Using equation (1) for A n,j Perform a forward conversion to obtain the j-th index after the forward conversion of the nth processing batch. : =100%-A n,j (1) Using formula (2) to determine the impurity content A n,2 Perform a forward conversion to obtain the impurity content of the nth processing batch after forward conversion. : =1 / A n,2 (2) Step 2.2: { |j=1,3,4,7,10,11}、 A n,5 A n,6 A n,8 A n,9 any one of the middle Individual marker is B n,i Thus, equations (3)-(5) are used to analyze B. n,i Standardization is performed to obtain the standardized form of the nth processing batch. Individual indicators : (3) (4) (5) In equations (3)-(5), B n,i The mean across all processing batches. B n,i Standard deviation across all processing batches.

3. The method for characterizing the overall quality of finished tobacco leaves after threshing and re-drying, as described in claim 2, is characterized in that... Step 4 includes the following steps: Step 4.1: Calculate the correlation coefficient matrix of the re-drying quality indicators of all processed batches of blades according to formula (6). : (6) In equation (6), ∈(0,1) is The standardized batch number after the nth processing batch Individual indicators The correlation coefficient; Step 4.2: Calculate the correlation coefficient matrix according to equation (7). The eigenvalue sequence λ={ ,in, The characteristic value of the i-th indicator is: =0 (7) In equation (7), I is an 11th-order identity matrix; Step 4.3: Calculate the feature vector sequence α of the quality index of leaf re-drying for all processing batches according to formula (8), and normalize α to obtain the normalized feature vector sequence. ={ ,in, The feature vector representing the i-th index of all processing batches: (λI - R)×α= 0 (8) Step 4.4: Calculate the cumulative variance contribution rate C of the first t characteristic values ​​of the re-drying quality index of all processing batches of blades according to formula (9). t Thus, 11 cumulative variance contribution rates are obtained: (9) In equation (9), This represents the m-th eigenvalue; Select the λ-satisfaction rate from the 11 cumulative variance contribution rates. m ≥1 and C t The feature value corresponding to ≥δ is used to determine the number of feature values ​​after screening, which is K. The score F of the kth feature value of the leaf re-drying quality index in the nth processing batch is calculated according to formula (10). n,k k=1,2….,K, where δ represents the contribution rate threshold: (10) In equation (10), For all processing batches, the first The k-th feature value of an indicator.

4. An electronic device, comprising a memory and a processor, characterized in that, The memory is used to store a program that supports a processor in executing the method of any one of claims 1-3, the processor being configured to execute the program stored in the memory.

5. A computer-readable storage medium storing a computer program thereon, characterized in that, The computer program is executed by the processor to perform the steps of the method according to any one of claims 1-3.

Images