Pattern Recognition Analysis Method for Flavor and Fragrance Systems Based on GC-IMS

Abstract

The present invention discloses a pattern recognition analysis method for flavor and fragrance systems based on GC-IMS, which includes: detecting various samples of the flavor and fragrance systems to obtain GC-IMS data of various types of samples; using a supervised pattern recognition method to perform feature pattern recognition of sensory omics data on the GC-IMS data of various types of samples, and establishing a classification model for the complex flavor and fragrance systems; and determining the sample category to which the unknown sample belongs according to the classification model. The pattern recognition analysis method for flavor and fragrance systems based on GC-IMS provided by the present invention performs feature pattern recognition based on GC-IMS sensory omics data in a supervised manner, conducts omics data analysis on the aroma compound information in combination with chemometrics, performs pattern recognition on volatile compounds and analyzes differential compounds, and determines the changes in the aroma components corresponding to each stage of tobacco flavor and fragrance and the influence law thereof on sensory stability.

Description

Technical Field

[0001] The invention relates to the technical field of tobacco product quality evaluation, and in particular to a GC-IMS-based pattern recognition analysis method for flavor and fragrance systems. Background Art

[0002] Flavor and fragrance system refers to special cigarette paper made by adding flavors, fragrances, extracts and their materials with the purpose of enhancing flavor, sweetening and coloring during the manufacturing process of cigarette paper. When the cigarette burns, the flavoring additives on the cigarette paper release the flavor components through volatilization, cracking and other methods to achieve the purpose of imparting a certain characteristic flavor. In recent years, cigarette paper flavoring technology has been widely used in the production of high-end cigarettes to improve the smoking quality of cigarettes. It has the advantages of effectively masking the impurities of cigarettes, giving the smoke a sweet and smooth feeling, reducing the irritation of cigarettes, and increasing the softness and delicacy of smoke. Due to the large number of volatile components, complex components and low content of aroma components contained in the flavor and fragrance system, it is difficult to trace the raw materials of the flavor and fragrance system, and there is a lack of effective stability monitoring methods. Therefore, there is no mature and reliable quality control system for the quality of the flavor and fragrance system, which affects the sensory quality stability of the cigarette brand.

[0003] At present, the main method for quality control of flavors and fragrances in my country is still physical evaluation indicators such as acidity, miscibility, refractive index, and density. In addition, the existing GC / MS method in the national standard has the disadvantages of insufficient specificity and sensitivity for the detection of trace aroma compounds, and mainly relies on qualitative and quantitative analysis methods of targeted compounds, lacking the overall quality evaluation method for the complex system of flavors and fragrances. The above problems constitute the shortcomings of the current quality control of flavors and fragrances.

[0004] How to effectively analyze the pattern recognition of flavor and fragrance systems to achieve accurate evaluation of the overall product quality has become a key technical bottleneck that needs to be urgently resolved in the tobacco industry.

[0005] Therefore, there is an urgent need for a pattern recognition analysis method for flavor and fragrance systems based on GC-IMS. Summary of the invention

[0006] The purpose of the present invention is to provide a pattern recognition analysis method for flavor and fragrance systems based on GC-IMS to solve the problems in the above-mentioned prior art. It can use a supervised pattern recognition method to perform feature pattern recognition based on sample GC-IMS sensory omics data on flavored cigarette paper, finished cigarette paper and blank paper samples, determine the sample category to which the unknown sample belongs, and help to achieve accurate evaluation of the overall product quality.

[0007] The present invention provides a pattern recognition analysis method for flavor and fragrance system based on GC-IMS, which comprises:

[0008] Detect the flavored cigarette paper samples, finished cigarette paper samples, finished cigarette paper samples after combustion, and blank paper samples to obtain the gas chromatography-tandem ion mobility spectrometry data of various types of samples;

[0009] Adopt a supervised pattern recognition method to perform characteristic pattern recognition of sensory omics data on the gas chromatography-tandem ion mobility spectrometry data of various types of samples, and establish a classification model for the complex system of flavors and fragrances;

[0010] According to the classification model, determine the sample category to which the unknown sample belongs.

[0011] The pattern recognition analysis method for the flavors and fragrances system based on GC-IMS as described above, wherein, preferably, the supervised pattern recognition method is used to perform characteristic pattern recognition of sensory omics data on the gas chromatography-tandem ion mobility spectrometry data of various types of samples, and establish a classification model for the complex system of flavors and fragrances, specifically including:

[0012] Through the ModelLab Matman general chemometrics solution software, perform characteristic analysis of sensory omics data on the gas chromatography-tandem ion mobility spectrometry data of various types of samples;

[0013] In the ModelLab Matman general chemometrics solution software, perform characteristic pattern recognition of sensory omics data on the gas chromatography-tandem ion mobility spectrometry data of various types of samples through at least one of the random forest pattern recognition modeling algorithm and the partial least squares discriminant modeling algorithm.

[0014] The pattern recognition analysis method for the flavors and fragrances system based on GC-IMS as described above, wherein, preferably, the number of decision trees adopted by the random forest pattern recognition modeling algorithm is 100; the eigenvalue algorithm adopted is Sqrt; the maximum tree depth is 30; the minimum impurity reduction is 0.01; the preprocessing algorithm adopted is UV scaling.

[0015] The pattern recognition analysis method for the flavors and fragrances system based on GC-IMS as described above, wherein, preferably, the number of latent variables retained by the partial least squares discriminant modeling algorithm is 5; k = 7-fold cross-validation is adopted in cross-validation; the number of random simulations is 10 times; the preprocessing algorithm adopted is UV scaling.

[0016] The pattern recognition analysis method for the flavors and fragrances system based on GC-IMS as described above, wherein, preferably, the sample categories to which the unknown sample belongs include flavored cigarette paper samples, finished cigarette paper samples, finished cigarette paper samples after combustion, and blank paper samples.

[0017] The pattern recognition analysis method of the flavor and fragrance system based on GC-IMS as described above, wherein, preferably, determining the sample category to which the unknown sample belongs according to the classification model specifically includes:

[0018] Detecting the unknown sample to obtain the gas chromatography tandem ion mobility spectrometry data of the unknown sample;

[0019] Inputting the gas chromatography tandem ion mobility spectrometry data of the unknown sample into the classification model to obtain the sample category to which the unknown sample belongs.

[0020] The pattern recognition analysis method of the flavor and fragrance system based on GC-IMS as described above, wherein, preferably, detecting the flavored cigarette paper sample, the finished cigarette paper sample, the burned finished cigarette paper sample and the blank paper sample to obtain the gas chromatography tandem ion mobility spectrometry data of each category of sample specifically includes:

[0021] Using a GC-IMS flavor analyzer to detect the flavored cigarette paper sample, the finished cigarette paper sample, the burned finished cigarette paper sample and the blank paper sample to obtain the gas chromatography tandem ion mobility spectrometry data of each category of sample.

[0022] The pattern recognition analysis method of the flavor and fragrance system based on GC-IMS as described above, wherein, preferably, using a GC-IMS flavor analyzer to detect the flavored cigarette paper sample, the finished cigarette paper sample, the burned finished cigarette paper sample and the blank paper sample to obtain the gas chromatography tandem ion mobility spectrometry data of each category of sample specifically includes:

[0023] Using a GC-IMS flavor analyzer to detect the straight ribbed wood pulp cigarette paper samples, wood pulp cigarette paper samples, and imported cross ribbed cigarette paper samples of different batches, and injecting samples repeatedly 3 times for each sample to obtain the gas chromatography tandem ion mobility spectrometry of the aroma volatile compounds of each flavored cigarette paper sample and the identification results of typical compounds;

[0024] Using a GC-IMS flavor analyzer to detect the finished cigarette paper and the burned finished cigarette paper from different origins and different batches, and injecting samples repeatedly 3 times for each sample to obtain the gas chromatography tandem ion mobility spectrometry of the aroma volatile compounds of each finished cigarette paper sample and the identification results of typical compounds;

[0025] Using a GC-IMS flavor analyzer to detect the plain cigarette paper base paper, and injecting samples repeatedly 3 times to obtain the gas chromatography tandem ion mobility spectrometry of the aroma volatile compounds of the blank paper sample and the identification results of typical compounds.

[0026] The pattern recognition analysis method of the flavor and fragrance system based on GC-IMS as described above, wherein, preferably, before detecting with a GC-IMS flavor analyzer, the method further includes:

[0027] Pretreating the sample, specifically including:

[0028] Take 0.5 g of cigarette paper and place it in a 20 mL headspace vial, incubate at 90 °C for 20 min and then inject the sample.

[0029] The pattern recognition analysis method of the flavor and fragrance system based on GC-IMS as described above, wherein, preferably, the headspace injection conditions when detecting with a GC-IMS flavor analyzer include:

[0030] The injection volume is 200 μL; the incubation time is 20 min; the incubation temperature is 90 °C; the injection needle temperature is 95 °C; the incubation rotation speed is 500 rpm;

[0031] The chromatographic conditions when detecting with a GC-IMS flavor analyzer include:

[0032] The chromatographic conditions of gas-phase ion mobility spectrometry are: the analysis time is 20 min; the chromatographic column type is WAX; the column length is 30 m; the inner diameter is ID-0.53 mm; the film thickness is FT 1 μm; the column temperature is 60 °C; the carrier gas / drift gas is N2; the IMS temperature is 45 °C;

[0033] GC chromatographic conditions: at the injection time of 0, the drift gas flow rate is 150 mL / min, the carrier gas flow rate is 2 mL / min, and the acquisition state is rec; at the injection time of 2 min, the drift gas flow rate is 150 mL / min, the carrier gas flow rate is 10 mL / min; at the injection time of 20 min, the drift gas flow rate is 150 mL / min, the carrier gas flow rate is 100 mL / min; at the injection time of 30 min, the drift gas flow rate is 150 mL / min, the carrier gas flow rate is 100 mL / min, and the acquisition state is stop.

[0034] The present invention provides a pattern recognition analysis method for flavor and fragrance systems based on GC-IMS. A supervised pattern recognition method is used to perform characteristic pattern recognition on flavor-impregnated cigarette papers, finished cigarette papers, and blank paper samples based on the sensory omics data of the samples in GC-IMS. Omics data analysis is carried out on the aroma compound information contained in the obtained GC-IMS spectra in combination with chemometrics, and pattern recognition and differential compound analysis are performed on volatile compounds, so as to determine the changes in the corresponding flavor components of tobacco flavor and fragrance from flavor-impregnated cigarette papers, finished cigarettes to those after combustion and their influence laws on sensory stability; it can preferably distinguish among four types of samples, namely flavor-impregnated cigarette papers, finished cigarette papers, the products after combustion of finished cigarette papers, and blank papers, indicating that there are significant differences in the composition of volatile compounds among various samples. Description of the Drawings

[0035] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described below in conjunction with the drawings, where:

[0036] Figure 1 It is a flowchart of an embodiment of the pattern recognition analysis method for flavor and fragrance systems based on GC-IMS provided by the present invention;

[0037] Figure 2 It is a schematic diagram of the prediction confusion matrix of the random forest model for the GC-IMS data of each sample;

[0038] Figure 3 It is a tree distribution diagram of the random forest model for the GC-IMS data of each sample;

[0039] Figure 4 It is a schematic flowchart of the topology of the random forest decision tree for the GC-IMS data of each sample;

[0040] Figure 5 It is a bar chart of the contribution scores of independent variables of the random forest for the GC-IMS data of each sample;

[0041] Figure 6 It is a score diagram of PLS-DA pattern recognition for the GC-IMS data of each sample;

[0042] Figure 7 It is a probability distribution diagram of PLS-DA pattern recognition prediction classification for the GC-IMS data of each sample;

[0043] Figure 8 It is the result of the recognition rate of cross-validation of PLS-DA pattern recognition for the GC-IMS data of each sample;

[0044] Figure 9 It is an S-plot diagram of the PLS-DA pattern recognition model for the GC-IMS data of each sample. Detailed Implementation Modes

[0045] Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. The description of the exemplary embodiments is merely illustrative and in no way limits the present disclosure or its application or use. The present disclosure can be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are provided to make the present disclosure thorough and complete, and to fully convey the scope of the present disclosure to those skilled in the art. It should be noted that: unless otherwise specifically stated, the relative arrangements of components and steps, the components of materials, numerical expressions, and numerical values set forth in these embodiments should be construed as merely exemplary and not as limitations.

[0046] The terms "first", "second", and the like used in the present disclosure do not denote any order, quantity, or importance, but are only used to distinguish different parts. Words such as "including" or "comprising" mean that the elements before this word cover the elements listed after this word, and do not exclude the possibility of also covering other elements. Terms such as "upper" and "lower" are only used to indicate relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0047] In the present disclosure, when it is described that a specific component is located between a first component and a second component, there may or may not be an intermediate component between the specific component and the first component or the second component. When it is described that a specific component is connected to other components, the specific component may be directly connected to the other components without an intermediate component, or may not be directly connected to the other components but have an intermediate component.

[0048] All terms used in the present disclosure (including technical terms or scientific terms) have the same meaning as understood by those of ordinary skill in the art to which the present disclosure pertains, unless otherwise specifically defined. It should also be understood that terms defined in a general dictionary, such as those, should be construed to have a meaning consistent with their meaning in the context of the relevant art and should not be interpreted in an idealized or overly formal sense, unless specifically defined as such herein.

[0049] Technologies, methods, and devices known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and devices should be regarded as part of the specification.

[0050] As Figure 1 shown, in the actual execution process of the pattern recognition analysis method for the flavor and fragrance system based on GC-IMS provided in this embodiment, it specifically includes:

[0051] Step S1: Detect the flavored cigarette paper samples, finished cigarette paper samples, burned finished cigarette paper samples, and blank paper samples to obtain the gas chromatography - ion mobility spectrometry (GC - IMS) data of various types of samples.

[0052] Specifically, use a GC - IMS flavor analyzer for detection to obtain the gas chromatography - ion mobility spectrometry data of various types of samples. Exemplarily, a German G.A.S. FlavourSpec GC - IMS flavor analyzer can be used for detection. The present invention does not specifically limit the manufacturer and model of the GC - IMS flavor analyzer.

[0053] Among them, GC - IMS combines the advantages of high separation of gas chromatography and high sensitivity of ion mobility spectrometry. Without any special sample pretreatment, it can quickly detect trace volatile organic compounds in samples and is used to measure the volatile headspace components in solid or liquid samples.

[0054] In an embodiment of the pattern recognition analysis method of the flavor and fragrance system based on GC - IMS of the present invention, the step S1 may specifically include:

[0055] Step S11: Use a GC - IMS flavor analyzer to detect the straight ribbed wood pulp cigarette paper samples, wood pulp cigarette paper samples, and imported cross - ribbed cigarette paper samples of different batches (see Table 1). Inject each sample for determination 3 times repeatedly to obtain the gas chromatography - ion mobility spectrometry of the aroma - forming volatile compounds of each flavored cigarette paper sample and the identification results of typical compounds.

[0056] Step S12: Use a GC - IMS flavor analyzer to detect the finished cigarette paper and burned finished cigarette paper of different origins and batches (see Table 1). Inject each sample for determination 3 times repeatedly to obtain the gas chromatography - ion mobility spectrometry of the aroma - forming volatile compounds of each finished cigarette paper sample and the identification results of typical compounds.

[0057] Step S13: Use a GC - IMS flavor analyzer to detect the plain cigarette paper base paper (see Table 1). Inject for determination 3 times repeatedly to obtain the gas chromatography - ion mobility spectrometry of the aroma - forming volatile compounds of the blank paper sample and the identification results of typical compounds.

[0058] It should be noted that the present invention does not specifically limit the sources, origins, batches, etc. of each sample.

[0059] Table 1 Summary of data of flavored cigarette paper samples for modeling analysis

[0060]

[0061]

[0062] Specifically, the headspace injection conditions when using a GC-IMS flavor analyzer for detection include:

[0063] The injection volume is 200 μL; the incubation time is 20 min; the incubation temperature is 90 °C; the injection needle temperature is 95 °C; the incubation rotation speed is 500 rpm;

[0064] The chromatographic conditions when using a GC-IMS flavor analyzer for detection include:

[0065] The chromatographic conditions of gas-phase ion mobility spectrometry are: the analysis time is 20 min; the type of chromatographic column is WAX; the column length is 30 m; the inner diameter is ID - 0.53 mm; the film thickness is FT 1 μm; the column temperature is 60 °C; the carrier gas / drift gas is N2; the IMS temperature is 45 °C;

[0066] The GC chromatographic conditions: at an injection time of 0, the drift gas flow rate is 150 mL / min, the carrier gas flow rate is 2 mL / min, and the acquisition status is rec; at an injection time of 2 min, the drift gas flow rate is 150 mL / min, the carrier gas flow rate is 10 mL / min; at an injection time of 20 min, the drift gas flow rate is 150 mL / min, the carrier gas flow rate is 100 mL / min; at an injection time of 30 min, the drift gas flow rate is 150 mL / min, the carrier gas flow rate is 100 mL / min, and the acquisition status is stop.

[0067] Before using a GC-IMS flavor analyzer for detection, the method further includes:

[0068] Performing pretreatment on the sample, specifically including:

[0069] Taking 0.5 g of cigarette paper and placing it in a 20 mL headspace vial, incubating at 90 °C for 20 min and then injecting the sample.

[0070] Step S2: Using a supervised pattern recognition method, perform feature pattern recognition on the gas chromatography tandem ion mobility spectrometry data of various types of samples for sensory omics data, and establish a classification model for the complex system of flavors and fragrances.

[0071] Specifically, in the present invention, through the ModelLab Matman general chemometric solution software (Chemmind Technologies, Beijing, China), the gas chromatography-tandem ion mobility spectrometry data of various types of samples are subjected to feature analysis of sensory omics data, and in the ModelLab Matman general chemometric solution software, through at least one of the random forest pattern recognition modeling algorithm and the partial least squares discriminant (PLS-DA) modeling algorithm, the gas chromatography-tandem ion mobility spectrometry data of various types of samples are subjected to pattern recognition of sensory omics data.

[0072] Among them, the number of decision trees adopted by the random forest pattern recognition modeling algorithm is 100; the eigenvalue algorithm adopted is Sqrt; the maximum tree depth is 30; the minimum impurity reduction is 0.01; the preprocessing algorithm adopted is UV scaling.

[0073] The prediction confusion matrix of the random forest model for each sample is shown in Figure 2 , Figure 2 showing the accuracy of the prediction results when the classification model established based on the random forest pattern recognition modeling algorithm groups each type of sample. When all predictions are correct, all samples will be located on the diagonal of the matrix in the figure. If there is a discrepancy between the prediction and the actual grouping, or if there is no attribution, the number of misjudged samples corresponding to it will deviate from the diagonal and be shown.

[0074] The tree distribution results of the random forest model for each sample are shown in Figure 3 , Figure 3 showing the distribution of 100 random decision trees for random forest pattern recognition modeling, Figure 3 in which it shows that most of the discriminant results are good. Although there are individual trees with poor prediction results, through the voting and scoring mode, the high accuracy of the overall prediction results is ensured. The effectiveness of the clear water sample is 77.33%, the effectiveness of the cigarette paper is 99.68%, the effectiveness of the formed cigarette is 99.68%, and the effectiveness of the combustion is 99.35%.

[0075] Figure 4 is a schematic flow chart of the random forest decision tree topology for each sample, Figure 4 showing the decision topology structure of one of the random forest trees, which is used as an example decision process. Through the binary tree method, each key variable is divided in turn (≤ comparison), and finally the correct prediction for each type of sample is obtained (classified with different symbols).

[0076] Figure 5 is a bar chart of the contribution scores of the independent variables of the random forest for each sample, Figure 5Show the statistical results of all independent variables (compounds) arranged in descending order of their contribution degrees to the random forest prediction grouping (the increase in binary tree purity and the decrease in entropy).

[0077] Furthermore, the number of latent variables retained by the partial least squares discriminant (PLS-DA) modeling algorithm is 5; k = 7-fold cross-validation is adopted in cross-validation; the number of random simulations is 10 times; the preprocessing algorithm adopted is UV scaling.

[0078] Among them, Figure 6 is the PLS-DA pattern recognition score plot, Figure 6 showing that all 4 types of samples are well distinguished, while the blank sample is closer to the original cigarette paper, and most samples do not exceed the 95% confidence interval of Totelling’s T2 (shown by the ellipse).

[0079] Figure 7 is the PLS-DA pattern recognition prediction classification probability distribution plot, Figure 7 showing the normal distribution probability density function of the predicted values of the PLS-DA pattern recognition model for each sample category. The sharper the peak shape and the less the overlap between the peaks, the higher the prediction performance of the classification model for the samples of this category.

[0080] Figure 8 is the PLS-DA pattern recognition cross-validation recognition rate result, Figure 8 showing the trend change relationship between the number of latent variables selected by the PLS-DA pattern recognition model for modeling and the model recognition rate (%). The optimal number of latent variables used for modeling should be at the highest point of the curve in the figure. The statistical table of the PLS-DA pattern recognition model recognition rate of the flavoring cigarette paper and its products is shown in Table 2.

[0081] Table 2 Statistical table of the PLS-DA pattern recognition model recognition rate of the GC-IMS data of the flavored cigarette paper and its products

[0082]

[0083] The S-plot of the PLS-DA pattern recognition model of each sample is shown in Figure 9 , Figure 9The figure shows a scatter plot of S-type independent variables (compounds) drawn with the covariance of independent variables (Pcorr contribution) and the correlation of independent variables (reliability) as the coordinate axes in the PLS-DA pattern recognition, illustrating the importance of each independent variable for regression classification prediction. Among them, the independent variables located in the first quadrant of the coordinate axis (plus sign in the upper right corner) play a positive correlation role in classification, while the independent variables located in the third quadrant of the coordinate axis (plus sign in the lower left corner) play an exactly opposite negative correlation role in classification. The statistical results of the significance of independent variables in the PLS-DA pattern recognition model for the GC-IMS data of each sample are shown in Table 3, and Table 3 only lists the top 30 independent variables (compounds) with the highest VIP scores.

[0084] Table 3 Statistical table of the significance of independent variables in the PLS-DA pattern recognition model for the GC-IMS data of flavored cigarette papers and their products

[0085]

[0086]

[0087] Step S3: Determine the sample category to which the unknown sample belongs according to the classification model.

[0088] Among them, the sample categories to which the unknown sample belongs include flavored cigarette paper samples, finished cigarette paper samples, burned finished cigarette paper samples, and blank paper samples.

[0089] In an implementation manner of the pattern recognition analysis method of the flavor and fragrance system based on GC-IMS of the present invention, step S3 may specifically include:

[0090] Step S31: Detect the unknown sample to obtain the gas chromatography tandem ion mobility spectrometry data of the unknown sample.

[0091] Step S32: Input the gas chromatography tandem ion mobility spectrometry data of the unknown sample into the classification model to obtain the sample category to which the unknown sample belongs.

[0092] Figures 2 - 9 The results show that both the random forest pattern recognition modeling algorithm and the partial least squares discriminant modeling algorithm adopted in the present invention can preferably distinguish four categories of samples, namely flavored cigarette papers, finished cigarette papers, burned finished cigarette papers, and blank papers. The recognition rates of each model algorithm for the four categories of samples are between 99% and 100%. It is indicated that there are significant differences in the composition of volatile compounds among different types of samples.

[0093] The pattern recognition analysis method of the flavor and fragrance system based on GC-IMS provided by the embodiments of the present invention uses a supervised pattern recognition method to perform feature pattern recognition on flavor-added cigarette paper, finished cigarette paper, and blank paper samples based on the sensory omics data of the samples in GC-IMS. It conducts omics data analysis on the aroma compound information contained in the obtained GC-IMS spectra in combination with chemometrics, performs pattern recognition on volatile compounds and analyzes differential compounds, so as to determine the changes in the corresponding flavor components of tobacco flavor and fragrance from flavor-added cigarette paper, finished cigarettes to after combustion and their influence laws on sensory stability; it can better distinguish the four categories of samples, namely flavor-added cigarette paper, finished cigarette paper, after burning of finished cigarette paper, and blank paper, indicating that there are significant differences in the composition of volatile compounds among different samples.

[0094] So far, the embodiments of the present disclosure have been described in detail. To avoid obscuring the concept of the present disclosure, some details known in the art have not been described. Those skilled in the art can fully understand how to implement the technical solutions disclosed herein based on the above description.

[0095] Although some specific embodiments of the present disclosure have been described in detail by way of examples, those skilled in the art should understand that the above examples are for illustrative purposes only and not for limiting the scope of the present disclosure. Those skilled in the art should understand that the above embodiments can be modified or some technical features can be equivalently replaced without departing from the scope and spirit of the present disclosure. The scope of the present disclosure is defined by the appended claims.

Claims

1. A pattern recognition analysis method for flavor and fragrance systems based on GC-IMS, characterized in that, Including: Testing the flavored cigarette paper samples, finished cigarette paper samples, burned finished cigarette paper samples and blank paper samples to obtain the gas chromatography-tandem ion mobility spectrometry data of various types of samples; Using a supervised pattern recognition method to perform feature pattern recognition on the gas chromatography-tandem ion mobility spectrometry data of various types of samples for sensory omics data, and establishing a classification model for the complex system of flavors and fragrances; Determining the sample category to which the unknown sample belongs according to the classification model, The testing of the flavored cigarette paper samples, finished cigarette paper samples, burned finished cigarette paper samples and blank paper samples to obtain the gas chromatography-tandem ion mobility spectrometry data of various types of samples specifically includes: Using a GC-IMS flavor analyzer to test the flavored cigarette paper samples, finished cigarette paper samples, burned finished cigarette paper samples and blank paper samples to obtain the gas chromatography-tandem ion mobility spectrometry data of various types of samples, specifically including: Using a GC-IMS flavor analyzer to test the straight ribbed wood pulp cigarette paper samples, wood pulp cigarette paper samples, and imported cross ribbed cigarette paper samples of different batches, and repeating the injection measurement 3 times for each sample to obtain the gas chromatography-tandem ion mobility spectrometry of the aroma volatile compounds of each flavored cigarette paper sample and the identification results of typical compounds; Using a GC-IMS flavor analyzer to test the finished cigarette paper and the burned finished cigarette paper from different origins and different batches, and repeating the injection measurement 3 times for each sample to obtain the gas chromatography-tandem ion mobility spectrometry of the aroma volatile compounds of each finished cigarette paper sample and the identification results of typical compounds; Using a GC-IMS flavor analyzer to test the plain cigarette paper base paper, and repeating the injection measurement 3 times to obtain the gas chromatography-tandem ion mobility spectrometry of the aroma volatile compounds of the blank paper sample and the identification results of typical compounds.

2. The pattern recognition analysis method for the flavor and fragrance system based on GC-IMS according to claim 1, wherein The using of a supervised pattern recognition method to perform feature pattern recognition on the gas chromatography-tandem ion mobility spectrometry data of various types of samples for sensory omics data, and establishing a classification model for the complex system of flavors and fragrances specifically includes: Performing feature analysis on the gas chromatography-tandem ion mobility spectrometry data of various types of samples for sensory omics data through the ModelLab Matman general chemometrics solution software; In the ModelLab Matman general chemometrics solution software, performing feature pattern recognition on the gas chromatography-tandem ion mobility spectrometry data of various types of samples for sensory omics data through at least one of the random forest pattern recognition modeling algorithm and the partial least squares discriminant modeling algorithm.

3. The pattern recognition analysis method for the flavor and fragrance system based on GC-IMS according to claim 2, wherein The number of decision trees used in the random forest pattern recognition modeling algorithm is 100; the eigenvalue algorithm used is Sqrt; the maximum tree depth is 30; the minimum impurity reduction is 0.01; the preprocessing algorithm used is UV scaling.

4. The pattern recognition analysis method for the flavor and fragrance system based on GC-IMS according to claim 2, characterized in that, The number of latent variables retained in the partial least squares discriminant modeling algorithm is 5; k = 7-fold cross-validation is used in cross-validation; the number of random simulations is 10 times; the preprocessing algorithm used is UV scaling.

5. The pattern recognition analysis method for the flavor and fragrance system based on GC-IMS according to claim 1, wherein The sample categories to which the unknown sample belongs include flavored cigarette paper samples, finished cigarette paper samples, finished cigarette paper samples after combustion, and blank paper samples.

6. The pattern recognition analysis method for the flavor and fragrance system based on GC-IMS according to claim 5, wherein Determining the sample category to which the unknown sample belongs according to the classification model specifically includes: Detecting the unknown sample to obtain the gas chromatography-tandem ion mobility spectrometry data of the unknown sample; Inputting the gas chromatography-tandem ion mobility spectrometry data of the unknown sample into the classification model to obtain the sample category to which the unknown sample belongs.

7. The pattern recognition analysis method of the flavor and fragrance system based on GC-IMS according to claim 1, wherein Before detecting with a GC-IMS flavor analyzer, the method further includes: Performing pretreatment on the sample, specifically including: Taking 0.5 g of cigarette paper and placing it in a 20 mL headspace vial, incubating at 90 °C for 20 min and then injecting the sample.

8. The pattern recognition analysis method for the flavor and fragrance system based on GC-IMS according to claim 1, characterized in that, The headspace injection conditions when detecting with a GC-IMS flavor analyzer include: The injection volume is 200 μL; the incubation time is 20 min; the incubation temperature is 90 °C; the injection needle temperature is 95 °C; the incubation rotation speed is 500 rpm; The chromatographic conditions when detecting with a GC-IMS flavor analyzer include: The chromatographic conditions of gas-phase ion mobility spectrometry are: the analysis time is 20 min; the chromatographic column type is WAX; the column length is 30 m; the inner diameter is ID - 0.53 mm; the film thickness is FT 1 μm; the column temperature is 60 °C; the carrier gas / drift gas is N2; the IMS temperature is 45 °C; GC chromatographic conditions: at an injection time of 0, the drift gas flow rate is 150 mL / min, the carrier gas flow rate is 2 mL / min, and the acquisition state is rec; at an injection time of 2 min, the drift gas flow rate is 150 mL / min, the carrier gas flow rate is 10 mL / min; at an injection time of 20 min, the drift gas flow rate is 150 mL / min, the carrier gas flow rate is 100 mL / min; at an injection time of 30 min, the drift gas flow rate is 150 mL / min, the carrier gas flow rate is 100 mL / min, and the acquisition state is stop.

Images