Method for identifying quality of alcoholized tobacco leaves

Abstract

The invention relates to the technical field of quality identification of alcoholized tobacco leaves, and discloses a quality identification method of alcoholized tobacco leaves. The method comprises the following steps: firstly, correcting an electronic eye, identifying the color of the alcoholized tobacco lamina by adopting the electronic eye, and carrying out color spectrum analysis to form a color distribution result; then detecting the alcoholized tobacco strip sample by using an electronic nose to obtain a volatile odor fingerprint spectrum of two chromatographic columns with different polarities, determining a difference chromatographic peak of the fingerprint spectrum, integrating data of the electronic eye and the electronic nose together, carrying out data normalization processing, extracting characteristic data, and carrying out PCA and DFA discrimination model analysis. According to the method, an electronic eye and a rapid gas-phase electronic nose are utilized to quantitatively describe and analyze the color and the difference smell of the alcoholized tobacco lamina, and the color and the difference smell are graphically represented; and after the data of the electronic eye and the electronic nose are independent or fitted, the redried tobacco strips in different packaging and alcoholization periods are identified through simple principal component analysis and discriminant factor model analysis.

Description

Technical Field

[0001] The invention relates to the technical field of quality identification of aged tobacco leaves, and in particular to a method for identifying the quality of aged tobacco leaves. Background Art

[0002] At present, the judgment of the aging process of tobacco leaves is mainly based on the traditional appearance of tobacco leaves (eye observation) and sensory evaluation (smoking evaluation), which are highly subjective, vague in description, and have a certain lag. For this reason, researchers continue to explore the quantitative methods of tobacco leaf appearance and sensory evaluation. At present, the colorimeter method is mostly used for instrumental quantification of tobacco color, and a model is established on this basis to quantify the browning of aging tobacco leaves. However, the colorimeter has the defect that it can only measure flat, regular, and non-colorful samples. The electronic eye can analyze the visual information of complex and heterogeneous samples; the electronic nose can quickly identify various odors by analyzing the odor components, and has the characteristics of high sensitivity, fast response, and accurate recognition. Therefore, in recent years, intelligent sensory systems such as electronic eyes and electronic noses have been widely used in the field of taste analysis, and have an incomparable role in rapid quality analysis. Some researchers have integrated a variety of intelligent sensory data and combined them with pattern recognition algorithms to identify red wines of different years, or medicinal materials such as Rehmannia glutinosa and American ginseng from different origins, as well as black tea, coffee, tobacco and other flavor styles. Although these methods can be used well for sample identification, most of them require complex mathematical modeling, which is labor-intensive and time-consuming.

[0003] Therefore, it is necessary to develop faster and simpler intelligent instrument methods for the quality identification of aged tobacco leaves. Summary of the invention

[0004] 1. Technical issues to be solved

[0005] In view of the deficiencies in the prior art, the present invention provides a method for identifying the quality of aged tobacco leaves, which has the advantages of quickly identifying the quality of aged tobacco leaves and solves the above-mentioned technical problems.

[0006] (II) Technical solution

[0007] To achieve the above object, the present invention provides the following technical solution: a method for identifying the quality of aged tobacco leaves, comprising the following steps:

[0008] S1. After calibrating the electronic eye, use the electronic eye to identify the color of the aged tobacco sample, perform color spectrum analysis, and form a color distribution result;

[0009] S2. After calibrating the electronic nose, the alcoholized tobacco slice samples are processed and detected by the electronic nose to obtain the volatile odor fingerprints of two chromatographic columns with different polarities, determine the differential chromatographic peaks of the fingerprints, and qualitatively identify the differential compounds;

[0010] S3. Integrate the electronic eye and electronic nose data, extract feature data after data normalization, and perform PCA and DFA discriminant model analysis.

[0011] As a preferred technical solution of the present invention, the specific steps of identifying the color of the aged tobacco leaf sample by the electronic eye in step S1 are as follows:

[0012] S1.1. After the electronic eye is turned on and stabilized, use a 24-color color calibration plate or 150 standard color cards and an average white card and a 5mm aperture for calibration;

[0013] S1.2. Place the aged tobacco leaf sample on a sample tray and select at least 5 different parts to collect images, perform color spectrum analysis, and determine the proportion information of each visible color, color distribution, and surface change.

[0014] As a preferred technical solution of the present invention, the specific steps of step S2 are as follows:

[0015] S2.1. Use normal alkane standard solution to calibrate the electronic nose;

[0016] S2.2, weigh 1.0g alcoholized tobacco sample into a 20mL headspace bottle for electronic nose, seal it with a polytetrafluoroethylene septum, heat and shake it in an incubator at 70℃ for 20min, and then separate the components to be tested into the dual chromatographic columns of the instrument for separation and detection through headspace injection and embedded pre-concentration trap, and obtain the volatile odor fingerprints of low-polarity and medium-polarity chromatographic columns;

[0017] S2.3. Use PCA statistics to find out the differences in volatile odor fingerprints between alcoholized tobacco samples.

[0018] As a preferred technical solution of the present invention, the specific process in step S3 is as follows:

[0019] S3.1. Integrate the electronic eye and electronic nose data of different aged tobacco strip samples together as a data matrix of aged tobacco strip samples;

[0020] S3.2, normalizing the data matrix of the sample in step S3.1, extracting characteristic data, and performing PCA analysis;

[0021] S3.3. Verify and evaluate the results of the PCA analysis in step S3.2 through the DFA model.

[0022] Compared with the prior art, the present invention provides a method for identifying the quality of aged tobacco leaves, which has the following characteristics:

[0023] Beneficial effects:

[0024] 1. Using electronic eyes and rapid gas phase electronic noses, the color and different odors of aged tobacco leaves are quantitatively described and analyzed, and represented graphically, which is more scientific and objective than conventional manual color grading and sensory evaluation methods.

[0025] 2. The electronic eye can analyze the color of a tobacco leaf sample very quickly, usually within a few seconds to a few minutes, while the rapid gas phase electronic nose can analyze the smell of a tobacco leaf sample in just 3 minutes, and can also obtain a sensory description of the volatile components. And after the electronic eye and electronic nose data are analyzed separately or after fitting, tobacco leaves of different packaging and aging periods can be distinguished through simple principal component analysis and discriminant factor model analysis; that is, the electronic eye, electronic nose and the fitting of the two can well distinguish tobacco leaves of different packaging and aging periods.

[0026] 3. Electronic eyes and rapid gas phase electronic noses have the advantages of being fast, convenient, objective and green in identifying aged tobacco strips, and can provide a scientific basis for the quality control of aged tobacco strips and the selection of aged cycles. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a schematic diagram of the color distribution of alcoholized tobacco sheets in different packages of the present invention;

[0028] Figure 2 Schematic diagram of fingerprints of volatile components of tobacco slices in different packages of the present invention;

[0029] Figure 3 Schematic diagram of PCA analysis of electronic nose signals of tobacco leaves in different packages according to the present invention;

[0030] Figure 4 A schematic diagram (2D) of PCA analysis of the electronic eye and electronic nose signal fitting of different tobacco leaves of the present invention;

[0031] Figure 5 The PCA analysis schematic diagram (3D) of the electronic eye and electronic nose signal fitting of different tobacco leaves of the present invention;

[0032] Fig. 6A The color spectrum data of the original image of the CG-12 sample of the present invention;

[0033] Figure 6B The original image color spectrum data of the CG-14 sample of the present invention;

[0034] Figure 6C The color spectrum data of the original image of the CG-16 sample of the present invention;

[0035] Fig.6D The original image color spectrum data of the CD-12 sample of the present invention;

[0036] Fig. 6EThe original image color spectrum data of the CD-14 sample of the present invention;

[0037] Fig. 6F The original image color spectrum data of the CD-16 sample of the present invention;

[0038] Figure 7 It is a schematic diagram of the Euclidean distance clustering diagram of the colors of the alcoholized tobacco sheets with different packages according to the present invention;

[0039] Figure 8 This is a schematic diagram of fitting and analyzing data of the electronic eye and electronic nose of the present invention;

[0040] Fig. 9 Schematic diagram of DFA analysis for fitting electronic eye and electronic nose signals of tobacco leaves in different packages according to the present invention. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] See also Figure 1-9 A method for identifying the quality of aged tobacco leaves comprises the following steps:

[0043] S1. After calibrating the electronic eye, use the electronic eye to identify the color of the aged tobacco sample, perform color spectrum analysis, and form a color distribution result;

[0044] The specific steps of identifying the color of the aged tobacco leaf sample by the electronic eye in step S1 are as follows:

[0045] S1.1. After the electronic eye is turned on and stabilized, use a 24-color color calibration plate or 150 standard color cards and an average white card, 5mm aperture for calibration, and turn on the top light source to eliminate the background;

[0046] S1.2. Place the aged tobacco leaf sample on a sample tray (420×560 mm) and select at least 5 different parts to collect images (image area 600,000 pixels). Use the RGB color processing system (colors are divided into 4096 colors) to determine the proportion information of each visible color, color distribution, and surface changes of the tobacco leaf sample. The results are as follows: Figure 1 As shown, the colors of alcoholized tobacco leaves are mainly 48 kinds, ranging from yellow to brown, and there are differences between samples treated with different methods.

[0047] S2. After calibrating the electronic nose, the alcoholized tobacco slice samples are processed and detected by the electronic nose to obtain the volatile odor fingerprints of two chromatographic columns with different polarities, determine the differential chromatographic peaks of the fingerprints, and qualitatively identify the differential compounds;

[0048] The specific steps of step S2 are as follows:

[0049] S2.1. Before the sample is tested, a standard solution of normal alkanes (nC 6 ~nC 16 ) Calibrate the dual fast gas phase electronic nose;

[0050] S2.2, weigh 1.0g of tobacco sample into a 20mL headspace bottle for electronic nose, seal it with PTFE (polytetrafluoroethylene) septum, heat and shake it in an incubator at 70℃ for 20min, and then separate the components to be tested into the dual chromatographic columns (low polarity and medium polarity columns) of the instrument for separation and detection after headspace injection and embedded pre-concentration trap capture. The volatile odor fingerprints of two different polarity chromatographic columns MXT-5 (low polarity) and MXT-1701 (medium polarity) were obtained. The results are as follows: Figure 2 As shown in the figure, it can be seen that the chromatographic peaks of different packaging cigarettes are between 52 and 68, and their peak areas and retention times are different, indicating that there are certain differences in the types and contents of volatile components in the samples;

[0051] S2.3. Find the differences between sample groups by comparing and amplifying the chromatograms. This process is rather tedious and complicated. Therefore, the method of statistics first and then chemistry is adopted. PCA statistics is used to find the differences in the gas chromatographic fingerprints between sample groups and determine the differential chromatographic peaks (discriminating ability>0.900, peak area>1000, and chromatographic peaks with good separation effect). The PCA results are as follows: Figure 3 As shown in the figure, it can be seen that the sum of the two principal components reaches 90.05%, representing most of the sample information. The recognition index is 93, indicating that the established electronic nose PCA model can distinguish alcoholized tobacco sheets under different packaging conditions.

[0052] The Discrimination Index (DI) is an evaluation of the discrimination quality of the selected target. It is obtained by calculating the surface area between groups and the surface area of ​​each group.

[0053] The discriminant index is positive: when the groups are independent of each other, the index is positive. The maximum value is 100, and the closer it is to 100, the better the effect.

[0054] The discrimination index is negative: There is overlap between the groups and the index is negative.

[0055] DI=100*{1-[Surface(A)+Surface(B)+surface(C)+…+surface(i)] / (Totalsurface)}=-(IntersectionSurface / Totalsurface)×100;

[0056] Then, based on the PCA analysis, the retention time of each component was converted into a retention index, and then the substance was qualitatively identified through the Arochembase database (including the retention index values ​​of 150,000 compounds), possible differential volatile substances were found, and sensory descriptions were performed;

[0057] S3. Integrate the electronic eye and electronic nose data, extract feature data after data normalization, and perform PCA and DFA discriminant model analysis.

[0058] The specific process in step S3 is as follows:

[0059] S3.1. Integrate the electronic eye and electronic nose data of different aged tobacco strip samples together as a data matrix of aged tobacco strip samples;

[0060] S3.2, normalize the data matrix of the sample in step S3.1, extract characteristic data, and perform PCA analysis, including normalization processing X'=(X-Xmin) / (Xmax-Xmin), calculate the covariance matrix, solve the eigenvalues ​​corresponding to the eigenvectors specified by the covariance matrix, determine the conversion matrix, solve the dimensionality reduction data, and obtain a 2D or 3D PCA chart with excellent visibility. Figure 4 and Figure 5 As shown. Figure 4 As can be seen from the figure, samples from different groups are clustered together and are far apart (no overlap), with a recognition index of 92. This shows that after the electronic eye and electronic nose data are fitted together, samples with different alcoholization treatments can be distinguished. It also shows that there are overall differences in color and smell between samples. The farther the distance between samples, the greater the overall sensory difference.

[0061] The distance between samples is the Euclidean distance;

[0062] S3.3. Verify and evaluate the results of the PCA analysis in step S3.2 through the DFA model.

[0063] Embodiment 1:

[0064] This embodiment uses an electronic eye to analyze the quality of aged tobacco leaves, specifically including the following steps:

[0065] 1. After the electronic eye is turned on and stabilized, use a 24-color color calibration plate or 150 standard color cards and an average white card, 5mm aperture for calibration, and turn on the top light source to eliminate the background.

[0066] 2. Place the aged tobacco leaf sample on a sample tray (420×560mm) and select at least 5 different parts to collect images (image area 600,000 pixels). Use the RGB color processing system (colors are divided into 4096 colors) to determine the proportion information of each visible color, color distribution, and surface changes of the tobacco leaf sample. The results are as follows Figure 5 In the figure, the horizontal axis is different color codes, CD and CG represent nitrogen filling and oxygen reduction and conventional sealed packaging, and the numbers 12, 14, and 16 represent aging for 12, 14, and 16 months.

[0067] Experimental results analysis: 6A to 6F as well as Figure 1 It can be seen that the colors of alcoholized tobacco leaves are mainly 48 kinds, ranging from yellow to brown, and there are differences between samples with different treatments. Figure 7 It can be seen that CD-12, CD-14, CD-16, CG12 and CG14 treatments are grouped together, indicating that their colors are somewhat similar, while CG-16 treatment is the farthest from other treatments, indicating that its color is the most different from other samples. Therefore, from the perspective of tobacco leaf color, nitrogen-filled packaging treatment delays the browning and deepening of tobacco leaf color during the aging process to a certain extent.

[0068] From the above analysis, it can be seen that the electronic eye is very fast to analyze the quality of aged tobacco leaves, but the quality of aged tobacco leaves is a comprehensive concept, including appearance, physical properties, chemical composition and sensory quality. Tobacco leaves of similar colors may have different chemical compositions and odors. Therefore, it is necessary to combine other indicators to evaluate the quality of tobacco leaves in a more comprehensive and reliable way.

[0069] Embodiment 2:

[0070] This embodiment uses a rapid gas phase electronic nose to analyze the quality of aged tobacco leaves, specifically including the following steps:

[0071] 1. Before the sample is tested, use a standard solution of normal alkanes (nC 6 ~nC 16 ) to calibrate the dual fast gas phase electronic nose.

[0072] 2. Weigh 1.0g of tobacco sample into a 20mL headspace bottle for electronic nose, seal it with a polytetrafluoroethylene (PTFE) septum, heat and shake it in an incubator at 70℃ for 20min, and then separate the components to be tested into the dual chromatographic columns (low polarity and medium polarity columns) of the instrument for separation and detection after headspace injection and embedded pre-concentration trap capture. The volatile odor fingerprints of two different polarity chromatographic columns, MXT-5 (low polarity) and MXT-1701 (medium polarity), were obtained. The results are as follows: Figure 2 As shown in the figure, the number of chromatographic peaks of different packaging cigarettes ranges from 52 to 68, and their peak areas are different, indicating that there are certain differences in the types and contents of volatile components in the samples.

[0073] 3. The differences between sample groups were found by comparing and amplifying the chromatograms. This process was rather tedious and complicated, so the method of statistics followed by chemistry was adopted. PCA statistics were used to find the differences in the gas chromatographic fingerprints between sample groups and determine the differential chromatographic peaks (discriminating ability > 0.900, peak area > 1000, and chromatographic peaks with good separation effect).

[0074] 4. Based on PCA analysis ( Figure 3 ), the retention time of each component was converted into a retention index, and then the substance was qualitatively identified through the Arochembase database. A total of 18 possible differential volatile substances were found and sensory descriptions were made. The results are shown in Table 1.

[0075] Table 1 Odor components and sensory description information of alcoholized tobacco slices

[0076]

[0077]

[0078] “—” means the sample does not contain the ingredient

[0079] Analysis of experimental results: From Table 1, it can be seen that the main volatile compounds with different packaging and aging cycles include alcohols (7 kinds), aldehydes (2 kinds), esters (2 kinds), heterocyclic compounds (2 kinds) and other compounds (5 kinds). Figure 2 The compounds with higher contents in each treatment were acetic acid, isopropanol, methyl formate, damascone and valeraldehyde.

[0080] Both packaging method and aging time affect the chemical quality of tobacco leaves, and the effect of time is more significant. Conventional packaging tobacco strips age quickly, and the content of damascena and myristic acid ether is the highest at 14 months of ageing; but as time goes by, the aroma components such as n-propanol and isopentenol in the tobacco strips decrease rapidly, while unpleasant odor substances such as acetic acid and pyrazine increase significantly. Nitrogen-filled packaging helps the accumulation of aroma components in tobacco leaves and can improve the quality of tobacco strips to a certain extent, but it is not conducive to the escape of compounds such as pyridine and meta-xylene.

[0081] From the above analysis, it can be seen that the rapid gas phase electronic nose is very fast to analyze the quality of aged tobacco leaves, but this method only obtains some volatile components with significant differences in each sample, and the contribution of these components to the smell of tobacco leaves is not very clear. Therefore, the electronic nose alone cannot comprehensively judge the quality of aged tobacco leaves.

[0082] Embodiment three:

[0083] The method of the present invention, the experimental steps of the two fittings are as follows:

[0084] 1. In order to investigate the identification effect of electronic eye and electronic nose combined analysis on aged tobacco leaves, the electronic eye and electronic nose data of tobacco leaf samples with different packaging methods and aged periods were integrated together as the data matrix or data set of the samples, such as Figure 8 As shown;

[0085] 2. Normalize the matrix data, extract the feature data, and perform PCA analysis. The results are as follows: Figure 4 As shown in the figure, it can be seen that the tobacco leaves of different packaging and aging cycles are grouped together, indicating that the electronic nose and electronic eye can distinguish the tobacco leaves of different aging cycles very well;

[0086] 3. To confirm the reliability of the electronic eye and electronic nose fitting models, DFA discriminant model analysis can also be performed. DFA is a classification technique that optimizes discrimination by recombining sensor data. Its purpose is to maximize the distance between groups while minimizing the difference within the group. The results are as follows Figure 7 As shown;

[0087] Experimental results analysis: Fig. 9 It can be seen that the sum of factors 1 and 2 of DFA is 82.53%, and each group of samples is well separated. The recognition rate is calculated by calculating the samples with good recognition effect, and its verification score is 94 (DFA models with scores greater than 90 are effective), indicating that the color and odor analysis can distinguish each sample well after fitting. The recognition rate calculation of DFA analysis usually involves comparing the number of correctly classified samples with the total number of samples. Specifically, the recognition rate can be calculated by the following formula:

[0088] Recognition rate = (number of correctly classified samples / total number of samples) * 100%

[0089] In the figure, tobacco leaves of different packaging are distributed in different areas of the figure, indicating that the packaging method and aging time cause the overall sensory differences of aging tobacco leaves. The DFA results of the fitted data are consistent with the PCA analysis results, indicating that the combination of electronic eyes and electronic noses can effectively distinguish the differences in volatile components of samples of different packaging. The longer the time, the greater the difference, and the farther the relative distance between samples at different times.

[0090] By comparison, it can be seen that the fitting method is fast, convenient, objective and green to identify tobacco strips with different packaging and different aging periods. Compared with the electronic eye and rapid gas phase electronic nose alone, the fitting method combines the appearance and odor components (sensory) of tobacco leaves, so it is more comprehensive in evaluating the quality of tobacco strips and the identification will be more accurate. This method can provide a scientific basis for the quality control of tobacco strips and the selection of aging cycles, so it also has broad application prospects.

[0091] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for identifying the quality of aged tobacco leaves, characterized in that: The following steps are involved: S1. After calibrating the electronic eye, use the electronic eye to identify the color of the aged tobacco sample, perform color spectrum analysis, and form a color distribution result; S2. After calibrating the electronic nose, the alcoholized tobacco leaf samples are processed and detected by the electronic nose to obtain volatile odor fingerprints of two chromatographic columns with different polarities, and determine the differential chromatographic peaks of the fingerprints; S3. Integrate the electronic eye and electronic nose data, extract feature data after data normalization, and perform PCA and DFA discriminant model analysis.

2. The method for identifying the quality of aged tobacco leaves according to claim 1, characterized in that: The specific steps of identifying the color of the aged tobacco leaf sample by the electronic eye in step S1 are as follows: S1.

1. After the electronic eye is turned on and stabilized, use a 24-color color calibration plate or 150 standard color cards and an average white card and a 5mm aperture for calibration; S1.

2. Place the aged tobacco leaf sample on a sample tray and select at least 5 different parts to collect images, perform color spectrum analysis, and determine the proportion information of each visible color, color distribution, and surface change.

3. The method for identifying the quality of aged tobacco leaves according to claim 1, characterized in that: The specific steps of step S2 are as follows: S2.

1. Use normal alkane standard solution to calibrate the electronic nose; S2.2, weigh 1.0g alcoholized tobacco sample into a 20mL headspace bottle for electronic nose, seal it with a polytetrafluoroethylene septum, heat and shake it in an incubator at 70℃ for 20min, and then separate the components to be tested into the dual chromatographic columns of the instrument for separation and detection after headspace injection and embedded pre-concentration trap collection to obtain low-polarity and medium-polarity volatile odor fingerprints; S2.3, using PCA statistics to find out the differences in volatile odor fingerprints between alcoholized tobacco samples; S2.

4. Use retention indices and databases to qualitatively identify different volatile compounds.

4. The method for identifying the quality of aged tobacco leaves according to claim 1, characterized in that: The specific process in step S3 is as follows: S3.

1. Integrate the electronic eye and electronic nose data of different aged tobacco strip samples together to serve as a data matrix for the aged tobacco strip samples; S3.2, normalizing the data matrix of the sample in step S3.1, extracting characteristic data, and performing PCA analysis; S3.

3. Verify and evaluate the results of the PCA analysis in step S3.2 through the DFA model.

Images