A method for analyzing a fragrance formulation

Abstract

The present application relates to a method for analyzing a fragrance formula, a method for simulating a fragrance, a system, a device and a computer readable storage medium for analyzing a fragrance formula. The method for analyzing a fragrance formula does not need to analyze a spectrum, is simple to operate, and can quickly provide a fragrance base formula for a perfumer.

Description

Technical Field

[0001] This invention belongs to the field of cigarette flavoring, specifically relating to a method for analyzing flavor formulas, a method for imitating flavor formulas, and a system, apparatus, and computer-readable storage medium for analyzing flavor formulas. Background Technology

[0002] When replicating cigarette flavorings, perfumers usually rely on their own experience and need to repeatedly blend the flavors to achieve a certain degree of realism. This not only requires perfumers to have a high level of perfumery skills and experience, but also requires a lot of effort and time.

[0003] The purpose of fragrance analysis is to identify the types of fragrance ingredients used and their proportions, providing perfumers with a basic formula for replicating fragrances and assisting them in the process. Due to the complexity of the fragrance system, a complete analysis of fragrances is currently impossible. Even if the composition and content of each substance in a fragrance were determined, factors such as the rarity of the ingredients and cost would prevent the actual replication of fragrances through single-component combinations (Computer-Aided Perfume: Research and Application, Zhong Kejun et al., Chemical Industry Press, 2015). Summary of the Invention

[0004] This invention provides a method for analyzing fragrance formulations, comprising:

[0005] Collect n kinds of fragrances, wherein the fragrances are selected from synthetic monomeric fragrances, fragrance extracts (including essential oils, extracts, fragrance resins, balms) and any combination thereof;

[0006] Collect chromatograms of n spices;

[0007] Extract the electrical signal data of the chromatogram over time and save it in digital format. Each spice chromatogram has m electrical signal data points.

[0008] The m electrical signal data points are arranged vertically and the n spices are arranged horizontally to form an m×n matrix, thus obtaining a chromatographic library.

[0009] Collect the chromatogram of the fragrance to be analyzed, extract the electrical signal data of the chromatogram changes over time, save it in digital format, and arrange it in the chromatogram library;

[0010] The electrical signal data of the fragrance were set as independent variables, and the electrical signal data of the fragrance to be analyzed were set as dependent variables. Stepwise regression analysis was used to screen out the fragrances that were strongly correlated with the fragrance to be analyzed.

[0011] The electrical signal data of the highly correlated fragrances obtained from the screening were set as independent variables, and the electrical signal data of the fragrance to be analyzed were set as dependent variables. A first regression equation was established, and the ratio of the coefficients of each fragrance in the first regression equation was the mass ratio of each fragrance in the fragrance to be analyzed, thus obtaining the formula of the fragrance to be analyzed.

[0012] In some implementation schemes, stepwise regression analysis is used to screen for fragrances that are strongly correlated with the fragrance to be analyzed, including:

[0013] Stepwise regression analysis was used to remove independent variables that were not significant after testing, and the independent variables that were significant after testing were selected. The second regression equation was established with the electrical signal data of the fragrance to be analyzed as the dependent variable.

[0014] Calculate the regression coefficients of the independent variables in the second regression equation, and further screen out the independent variables with regression coefficients greater than 0, which are the fragrances that are strongly correlated with the fragrance to be analyzed.

[0015] In some implementations, the first or second regression equation is established using the least squares method.

[0016] In some implementations, chromatographic analysis (e.g., dynamic headspace-gas chromatography-mass spectrometry) is used to collect chromatograms of n fragrances or flavorings to be analyzed.

[0017] In some implementations, n is greater than or equal to 20, such as 26, 50, 80, 100, 120, 150, 180, 200, 250, 280, 300, 500, etc.

[0018] In some implementations, after obtaining the chromatogram, the process further includes preprocessing the chromatogram and then extracting the electrical signal from the chromatogram as it changes over time.

[0019] In some implementations, preprocessing the chromatogram includes at least one of the following methods:

[0020] Baseline noise is removed;

[0021] The Savitzky-Golay convolution smoothing algorithm was used to process the spectral image.

[0022] The present invention also provides a method for imitating fragrances, comprising:

[0023] 1) The formula of the imitated fragrance is analyzed using the method described in this invention;

[0024] 2) Prepare the fragrance according to the formula obtained in 1).

[0025] The present invention also provides a system for analyzing fragrance formulations, comprising:

[0026] Detection instruments are used to collect chromatograms of the fragrance to be analyzed and the fragrance used for library construction;

[0027] The data processing module is used to extract the electrical signal data of the chromatogram over time and save it in digital format, where the chromatogram data for each spice has m data points.

[0028] The library building module is used to construct a chromatographic library, in which m data points are arranged vertically and n spices are arranged horizontally to form an m×n matrix to obtain the chromatographic library.

[0029] The analysis module is used to screen fragrance components that are strongly correlated with the fragrance to be analyzed. Fragrance component data are set as independent variables, and the component data of the fragrance to be analyzed are set as dependent variables. Stepwise regression analysis is used to screen for fragrance components that are strongly correlated with the fragrance to be analyzed.

[0030] The regression equation module is used to construct the first regression equation to determine the mass ratio of various flavorings in the flavoring to be analyzed. The identified flavoring components with strong correlation are set as independent variables, and the components of the flavoring to be analyzed are set as dependent variables. The first regression equation is established, and the ratio of the coefficients of each flavoring in the first regression equation is the mass ratio of various flavorings in the flavoring to be analyzed.

[0031] In some embodiments, the system for analyzing fragrance formulations further includes a preprocessing module for preprocessing the chromatogram, the preprocessing including at least one of the following preprocessing methods:

[0032] Baseline noise is removed;

[0033] The Savitzky-Golay convolution smoothing algorithm was used to process the spectral image.

[0034] The present invention also provides an apparatus for analyzing fragrance formulations, comprising:

[0035] Memory, and

[0036] A processor coupled to the memory is configured to execute the method for analyzing fragrance formulations according to the present invention based on instructions stored in the memory.

[0037] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for analyzing fragrance formulations described in the present invention.

[0038] Terminology Definition

[0039] In this invention, unless otherwise stated, the scientific and technical terms used herein have the meanings commonly understood by those skilled in the art. Furthermore, to better understand this invention, definitions and explanations of relevant terms are provided below.

[0040] In this invention, the term "fragrance" refers to a mixture artificially formulated containing two or more, or even dozens, of fragrances (sometimes also containing suitable solvents or carriers), which has a certain aroma.

[0041] In this invention, the term "fragrance" refers to a substance that can be smelled or tasted, including natural and synthetic fragrances. Natural fragrances are aromatic compounds extracted from the organs or secretions of fragrant plants and animals through processing, such as various fragrance extracts, including essential oils, extracts, resins, and balms. Synthetic fragrances are synthetic monomeric fragrances prepared using different raw materials through chemical (or biological) synthesis.

[0042] In this invention, the term "essential oil" refers to a product containing aromatic substances extracted from aromatic plants by steam distillation, pressing, cold grinding, or dry distillation.

[0043] In this invention, the term "extract" refers to a fragrance product obtained by extracting fragrance plant tissues (such as flowers, leaves, stems, bark, roots, fruits, etc.) that do not contain exudates using an organic solvent, and the finished product does not contain the original solvent and water.

[0044] In this invention, the term "fragrant balm," also known as fragrant liquid, refers to a resinous substance containing aromatic components that is exuded by aromatic plants due to physiological or pathological reasons.

[0045] In this invention, the term "fragrant resin" refers to a resinous substance containing fragrant components that is extracted from fragrant plants using an organic solvent, and the final product is prepared by removing the solvent and water used.

[0046] Beneficial technical effects of the present invention

[0047] The method for analyzing flavor formulations provided by this invention does not require the analysis of spectra. Instead, it uses a stepwise regression method from chemometrics to process the electrical signals and analyze the types and proportions of flavorings used in cigarette flavorings.

[0048] The method for analyzing fragrance formulas provided by this invention is simple to operate, requires minimal personnel, and can quickly provide perfumers with basic fragrance formulas, reducing their workload. Detailed Implementation

[0049] The embodiments of the present invention will be described in detail below with reference to examples. However, those skilled in the art will understand that the following examples are for illustrative purposes only and should not be considered as limiting the scope of the invention. Unless otherwise specified in the examples, conventional conditions or conditions recommended by the manufacturer are followed. Raw materials, equipment, or instruments whose manufacturers are not specified are all commercially available conventional products.

[0050] In this embodiment of the invention, the least squares method is used to establish the regression equation.

[0051] Example 1: Analysis of the formulation of fragrance A, composed of fully synthetic monomeric flavorings

[0052] 1. Sample collection

[0053] We have collected 300 commonly used fragrances, including synthetic monomeric fragrances and different types of fragrance extracts such as essential oils, extracts, resins, and balms.

[0054] 2. DHS-GC / MS (Dynamic Headspace-Gas Chromatography-Mass Spectrometry) Analysis

[0055] Weigh 0.1–1.0 g of fragrance or flavoring agent and place it in a 20 mL headspace vial. Add 0.2–0.4 mL of saturated sodium chloride solution, then seal the vial for analysis. Analyze using DHS-GCMS to obtain the chromatogram for each fragrance / flavoring agent. Instrument conditions are as follows:

[0056] Dynamic headspace analyzer: 20 mL headspace vial, sample equilibration temperature 80℃, injection needle temperature 90℃, transfer line temperature 100℃, trap high temperature 250℃, trap low temperature 40℃; dry blow 3.0 min, desorption 0.5 min, equilibration time 30.0 min, trap hold 5.0 min; sample vial pressurization time 2.0 min, injection time 0.20 min, needle withdrawal time 0.50 min. Injection mode: trap, column pressure: 35 psi, vial pressure: 40 psi, desorption pressure: 35 psi, outlet separation 15 mL / min.

[0057] Gas chromatograph: HP-5MS column (60m×0.25mm×0.25m); injection port temperature 250℃; column oven temperature program: initial temperature 60℃, increase to 170℃ at 2℃ / min and hold for 1min, then increase to 250℃ at 50℃ / min and hold for 5.4min; carrier gas: helium (purity 99.999%); carrier gas flow rate 1.0mL / min; split ratio 5:1.

[0058] Mass spectrometer detector: transfer line temperature 200℃, ion source temperature 200℃, tuning voltage 1420V, solvent delay: 5min; scan mass range: 35~455u, detection mode: full scan mode, spectral library search using NIST library.

[0059] 3. Establishment of a chromatographic library

[0060] The obtained chromatograms were preprocessed using baseline noise subtraction and Savitzky-Golay smoothing filtering (11 points, polynomial degree 3). Then, the electrical signal data of the chromatograms over time were extracted and saved in digital format. Each sample's chromatogram data contained 13845 electrical signal data points. A 13845×n matrix was constructed using the 13845 electrical signal data points from n samples to obtain the chromatogram library.

[0061] 4. Analysis of the formula for fragrance A

[0062] The formula for fragrance A is: ethyl acetate: propyl acetate: ethyl valerate: ethyl butyrate: isoamyl isovalerate = 2:2:1:2:1.5.

[0063] Fragrance A was analyzed and detected using the DHS-GC / MS (Dynamic Headspace-Gas Chromatography-Mass Spectrometry) method described in section 2, and chromatograms were obtained. The obtained chromatograms were preprocessed by subtracting baseline noise and using Savitzky-Golay smoothing filtering (11 points, polynomial degree 3). Then, the electrical signal data of the chromatograms changing over time were extracted and arranged into a chromatogram library.

[0064] The electrical signal data of the fragrance sample A to be analyzed were set as the dependent variable, and the electrical signal data in the chromatographic library were set as the independent variables. Stepwise regression was used for analysis. Variables with an F-distribution probability less than 0.05 were included in the regression analysis, while those greater than 0.1 were removed from the regression analysis.

[0065] Through stepwise regression calculations, five fragrances—ethyl butyrate, propyl acetate, ethyl valerate, ethyl acetate, and isoamyl isovalerate—were accurately screened from 300 raw materials. Regression equations were established, resulting in five models, as shown in Table 1.1. Table 1.1 shows that as the number of variables increases, the correlation coefficient, coefficient of determination, and adjusted coefficient of determination of the models gradually increase, indicating that the model performance gradually improves.

[0066] Table 1.1 Analytical Model for HS-GC / MS Chemical Composition Data

[0067]

[0068] 1: Constant, ethyl butyrate

[0069] 2: Constants, ethyl butyrate, propyl acetate

[0070] 3: Constants, ethyl butyrate, propyl acetate, ethyl valerate

[0071] 4: Constants, ethyl butyrate, propyl acetate, ethyl valerate, ethyl acetate

[0072] 5: Constants, ethyl butyrate, propyl acetate, ethyl valerate, ethyl acetate, isoamyl isovalerate

[0073] The results of the ANOVA for the five models are shown in Table 1.2, and the significance test results for the regression coefficients are shown in Table 1.3. The results show that the p-values ​​of all models are less than 0.05, and all established models are statistically significant. The regression coefficients in the models are statistically significant, and the ratio between the regression coefficients represents the proportion of spices used in the formula. The analyzed formula proportions are quite close to the actual proportions.

[0074] Table 1.2 Results of Analysis of Variance for the Data Analysis Model of Fragrance A

[0075]

[0076] Table 1.3 Significance test results of regression coefficients in the data analysis model for flavor A

[0077]

[0078]

[0079] Example 2: Analysis of Fragrance Formula B, composed of all-natural flavorings

[0080] The formula for flavoring B is sweet orange oil: white lemon oil: grapefruit oil: bergamot oil = 2:1:0.8:0.2.

[0081] The DHS-GC / MS analysis method in Example 1 was used to detect and analyze flavor B, and a chromatogram was obtained. The obtained chromatogram was preprocessed by removing baseline noise and Savitzky-Golay smoothing filter (11 points, polynomial degree 3). Then, the electrical signal data of the chromatogram changing over time was extracted and arranged into the chromatogram library obtained in Example 1.

[0082] The electrical signal data of the fragrance B sample to be analyzed were set as the dependent variable, and the electrical signal data from the chromatographic library were set as the independent variables. Stepwise regression was used for analysis. Variables with an F-distribution probability less than 0.05 were included in the regression analysis, while those greater than 0.1 were removed from the regression analysis.

[0083] Through stepwise regression calculations, four flavorings—sweet orange oil, white lemon oil, bitter orange oil, and grapefruit oil—were accurately selected from 300 raw materials. Regression equations were established, resulting in four models, as shown in Table 2.1. Table 2.1 shows that as the number of variables increases, the correlation coefficient, coefficient of determination, and adjusted coefficient of determination of the models gradually increase, indicating that the model performance gradually improves.

[0084] Table 2.1 Analytical Model for Flavor B Data

[0085]

[0086]

[0087] 1: Constant, sweet orange oil

[0088] 2: Constant, sweet orange oil, white lemon oil

[0089] 3: Constant, sweet orange oil, white lemon oil, bitter orange oil

[0090] 4: Constant, sweet orange oil, white lemon oil, bitter orange oil, grapefruit oil

[0091] Table 2.2 Correlation Analysis Among Spices

[0092]

[0093] * The correlation was significant at the 0.05 level (two-sided).

[0094] **. The correlation was significant at the 0.01 level (two-tailed).

[0095] Correlation analysis was performed on sweet orange oil, white lemon oil, grapefruit oil, bergamot oil, and bitter orange oil to calculate the Pearson correlation coefficients among the flavorings. The results are shown in Table 2.2. The correlation analysis shows that bergamot oil has a weak correlation with flavoring B, a significant correlation with sweet orange oil, and a highly significant correlation with bitter orange oil. Since flavoring B contains relatively little bergamot oil, it was not included in the calculation model during variable selection. Bitter orange oil showed significant correlations with flavoring B, sweet orange oil, grapefruit oil, and white lemon oil, and was therefore included in the calculation model.

[0096] The significance test of the regression coefficients of the established regression equations is shown in Table 2.3. The results show that the coefficients of bitter orange oil in Model 3 and Model 4 are both negative, indicating that it is unlikely to be present in the formula. Therefore, bitter orange oil was removed from the formula.

[0097] After excluding bitter orange oil, regression analysis was conducted using the electrical signal data of flavoring B as the dependent variable and the electrical signal data of sweet orange oil, white lemon oil, and grapefruit oil as independent variables. Regression equations were established, resulting in three models. The significance of the regression coefficients was tested, and the results are shown in Table 2.4. The results show that the regression coefficients are statistically significant, and the ratio between the regression coefficients represents the proportion of natural flavorings used in the formula. The obtained formula proportions are close to the actual formula proportions, indicating good results.

[0098] Table 2.3 Significance Test of Regression Coefficients in the Analytical Model of Flavor B

[0099]

[0100]

[0101] Table 2.4 Significance Test of Regression Coefficients in the Analytical Model of Flavor B

[0102]

[0103] Analysis of the flavor formula C composed of mixed fragrances in Example 3

[0104] Fragrance C is formulated as follows: Zimbabwean tobacco essential oil: 2,3,5-trimethylpyrazine: sweet orange oil: 2-acetylpyridine: 2-methylbutyric acid: Peruvian extract: ethyl valerate: cinnamyl acetate = 1.7:1:1:0.5:0.01:0.8:0.3:0.2.

[0105] The DHS-GC / MS analysis method in Example 1 was used to detect and analyze flavor C, and a chromatogram was obtained. The obtained chromatogram was preprocessed by removing baseline noise and Savitzky-Golay smoothing filter (11 points, polynomial degree 3). Then, the electrical signal data of the chromatogram changing over time was extracted and arranged into the chromatogram library obtained in Example 1.

[0106] The electrical signal data of the fragrance C sample to be analyzed were set as the dependent variable, and the electrical signal data in the chromatographic library were set as the independent variables. Stepwise regression was used for analysis. Variables with an F-distribution probability less than 0.05 were included in the regression analysis, while those greater than 0.1 were removed from the regression analysis.

[0107] Through stepwise regression calculations, 13 flavorings were selected from 300 raw materials: flue-cured tobacco essential oil, 2,3,5-trimethylpyrazine, bitter orange oil, Zimbabwean tobacco essential oil, cinnamic acid, 2-acetylpyridine, benzoic acid, sweet orange oil, 2-methylbutyric acid, cinnamyl alcohol, coriander seed oil, Peruvian balsam, and ethyl valerate. Regression equations were established, resulting in 13 models, as shown in Table 3.1. Table 3.1 shows that as the number of variables increases, the correlation coefficient, coefficient of determination, and adjusted coefficient of determination of the models gradually increase, indicating that the model performance gradually improves.

[0108] Table 3.1 Analytical Model for Fragrance C Data

[0109]

[0110] 1: Flue-cured tobacco essential oil

[0111] 2: Flue-cured tobacco essential oil, 2,3,5-trimethylpyrazine,

[0112] 3: Flue-cured tobacco essential oil, 2,3,5-trimethylpyrazine, bitter orange oil,

[0113] 4: Flue-cured tobacco essential oil, 2,3,5-trimethylpyrazine, bitter orange oil, Zimbabwean tobacco essential oil.

[0114] 5: Flue-cured tobacco essential oil, 2,3,5-trimethylpyrazine, bitter orange oil, Zimbabwean tobacco essential oil, cinnamic acid,

[0115] 6: Flue-cured tobacco essential oil, 2,3,5-trimethylpyrazine, bitter orange oil, Zimbabwean tobacco essential oil, cinnamic acid, 2-acetylpyridine

[0116] 7: Flue-cured tobacco essential oil, 2,3,5-trimethylpyrazine, bitter orange oil, Zimbabwean tobacco essential oil, cinnamic acid, 2-acetylpyridine, benzoic acid

[0117] 8: Flue-cured tobacco essential oil, 2,3,5-trimethylpyrazine, bitter orange oil, Zimbabwean tobacco essential oil, cinnamic acid, 2-acetylpyridine, benzoic acid, sweet orange oil

[0118] 9: Flue-cured tobacco essential oil, 2,3,5-trimethylpyrazine, bitter orange oil, Zimbabwean tobacco essential oil, cinnamic acid, 2-acetylpyridine, benzoic acid, sweet orange oil, 2-methylbutyric acid

[0119] 10: Flue-cured tobacco essential oil, 2,3,5-trimethylpyrazine, bitter orange oil, Zimbabwean tobacco essential oil, cinnamic acid, 2-acetylpyridine, benzoic acid, sweet orange oil, 2-methylbutyric acid, cinnamyl alcohol

[0120] 11: Flue-cured tobacco essential oil, 2,3,5-trimethylpyrazine, bitter orange oil, Zimbabwean tobacco essential oil, cinnamic acid, 2-acetylpyridine, benzoic acid, sweet orange oil, 2-methylbutyric acid, cinnamyl alcohol, coriander seed oil

[0121] 12: Flue-cured tobacco essential oil, 2,3,5-trimethylpyrazine, bitter orange oil, Zimbabwean tobacco essential oil, cinnamic acid, 2-acetylpyridine, benzoic acid, sweet orange oil, 2-methylbutyric acid, cinnamyl alcohol, coriander seed oil, Peruvian balm

[0122] 13: Flue-cured tobacco essential oil, 2,3,5-trimethylpyrazine, bitter orange oil, Zimbabwean tobacco essential oil, cinnamic acid, 2-acetylpyridine, benzoic acid, sweet orange oil, 2-methylbutyric acid, cinnamyl alcohol, coriander seed oil, Peruvian balsam, ethyl valerate.

[0123] Of the 13 selected flavorings, 7 were found in Flavor C formulation. Cinnamyl acetate was not selected, possibly because the selected Peruvian balm contained cinnamyl acetate. Six flavorings—flue-cured tobacco oil, bitter orange oil, cinnamic acid, benzoic acid, cinnamyl alcohol, and coriander seed oil—were not found in Flavor C formulation. This is likely because flue-cured tobacco oil is related to Zimbabwean tobacco oil, bitter orange oil and coriander seed oil are related to sweet orange oil, and cinnamic acid, benzoic acid, and cinnamyl alcohol are abundant in Peruvian balm.

[0124] Analysis of variance and significance tests of regression coefficients were performed on the established regression equations. The results are shown in Tables 3.2 and 3.3. The results show that the p-values ​​of all 13 models were less than 0.05, indicating that all established models were statistically significant. The regression coefficients of flavorings in the models were statistically significant, and the ratios between the regression coefficients represent the proportion of natural flavorings used in the formula. Except for 2-methylbutyric acid, the analyzed formula proportions were close to the actual proportions, indicating ideal analytical results.

[0125] The fragrance C obtained from Model 13 was formulated and then blended. The blended fragrance and the imitated fragrance C were subjected to sensory evaluation. The results showed that the aroma of the imitated fragrance was basically consistent with the aroma category of the target imitated fragrance.

[0126] Table 3.2 Analysis of Variance for Fragrance C Data Model

[0127]

[0128]

[0129] Table 3.3 Significance Test of Regression Coefficients in the Flavor C Analytical Model

[0130]

[0131]

[0132]

[0133] Adjustment: Adjust the selected ingredients to be similar to those in the actual formula, such as adjusting the content of flue-cured tobacco essential oil to include the content of Zimbabwean tobacco essential oil.

[0134] Zimbabwean tobacco essential oil; pyrazine represents 2,3,5-trimethylpyrazine; sweet represents sweet orange oil; pyridine represents 2-acetylpyridine; butyl represents 2-methylbutyric acid; sec represents Peruvian extract; pent represents ethyl valerate.

[0135] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them; although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications can still be made to the specific implementation of the present invention or equivalent substitutions can be made to some technical features without departing from the spirit of the technical solutions of the present invention, and all such modifications and substitutions should be covered within the scope of the technical solutions claimed in the present invention.

Claims

1. A method for analyzing a fragrance formulation, comprising: collecting n kinds of fragrances selected from the group consisting of synthetic monomeric fragrances, fragrance extracts, and any combination thereof; collecting chromatograms of the n kinds of fragrances; extracting time-varying electrical signal data of the chromatograms and saving the data in digital format, wherein each kind of fragrance has m electrical signal data points; arranging the m electrical signal data points longitudinally and the n kinds of fragrances transversely to form an m x n matrix, thereby obtaining a chromatogram library; collecting a chromatogram of a fragrance to be analyzed, extracting time-varying electrical signal data of the chromatogram and saving the data in digital format, and arranging the data into the chromatogram library; setting the electrical signal data of the fragrances as independent variables and the electrical signal data of the fragrance to be analyzed as dependent variables, and using stepwise regression analysis to screen out fragrances having strong correlation with the fragrance to be analyzed; setting the electrical signal data of the screened fragrances having strong correlation as independent variables and the electrical signal data of the fragrance to be analyzed as dependent variables, and establishing a first regression equation, wherein the coefficient ratio of each fragrance in the first regression equation is the mass ratio of each fragrance in the fragrance to be analyzed, thereby obtaining the formulation of the fragrance to be analyzed.

2. The method of claim 1, wherein the fragrance extracts comprise essential oils, extracts, balsams, or resins.

3. The method of claim 1, wherein using stepwise regression analysis to screen out fragrances having strong correlation with the fragrance to be analyzed comprises: using stepwise regression analysis to eliminate independent variables that are not significant in the test, screening out independent variables that are significant in the test, and establishing a second regression equation with the electrical signal data of the fragrance to be analyzed as the dependent variable; calculating the regression coefficients of the independent variables in the second regression equation, and further screening out independent variables having regression coefficients greater than 0, which are the fragrances having strong correlation with the fragrance to be analyzed.

4. The method of claim 1, wherein the first regression equation or the second regression equation is established by least squares method.

5. The method of claim 1, wherein the chromatograms of the n kinds of fragrances or the fragrance to be analyzed are collected by chromatographic analysis.

6. The method of claim 5, wherein the chromatographic analysis is dynamic headspace-gas chromatography-mass spectrometry.

7. The method of any one of claims 1-6, wherein n is greater than or equal to 20.

8. The method of claim 7, wherein n is 26, 50, 80, 100, 120, 150, 180, 200, 250, 280, 300, or 500.

9. The method of any one of claims 1-6, wherein after obtaining the chromatograms, the method further comprises pre-processing the chromatograms, and then extracting time-varying electrical signals of the chromatograms.

10. The method of claim 9, wherein the pre-processing of the chromatograms comprises at least one of the following processing methods: subtracting baseline noise; processing the chromatograms using Savitzky-Golay convolution smoothing algorithm.

11. A method for imitating a fragrance, comprising: 1) analyzing the formulation of a fragrance to be imitated using the method of any one of claims 1-10; 2) preparing a fragrance according to the formulation obtained in 1).

12. A system for analyzing a fragrance formulation, comprising: a detection instrument for collecting chromatograms of fragrances to be analyzed and fragrances used for library building. a data processing module configured to extract the time-varying electrical signal data of the chromatogram and save the data in a digital format, wherein the chromatogram data of each flavor has m data points, a database building module configured to build a chromatogram database, wherein the m data points are arranged longitudinally and the n flavors are arranged laterally to form an m×n matrix, thereby obtaining the chromatogram database; an analysis module configured to screen the flavor having a strong correlation with the essence to be analyzed, wherein the flavor component data is set as the independent variable and the component data of the essence to be analyzed is set as the dependent variable, and a stepwise regression analysis is used to screen the flavor having a strong correlation with the essence to be analyzed; a regression equation module configured to build a first regression equation and determine the mass ratio of each flavor in the essence to be analyzed, wherein the component data of the flavor having a strong correlation obtained by discrimination is set as the independent variable and the component of the essence to be analyzed is set as the dependent variable, thereby establishing the first regression equation, and the coefficient ratio of each flavor in the first regression equation is the mass ratio of each flavor in the essence to be analyzed.

13. The system of claim 12, further comprising a preprocessing module configured to preprocess the chromatogram, wherein the preprocessing comprises at least one of the following preprocessing methods: subtracting the baseline noise; processing the spectrum using a Savitzky-Golay convolution smoothing algorithm.

14. An apparatus for analyzing the formula of an essence, comprising: a memory, and a processor coupled to the memory, wherein the processor is configured to execute the method for analyzing the formula of an essence according to any one of claims 1-10 based on the instructions stored in the memory.

15. A computer readable storage medium having stored thereon a computer program, wherein the program is executed by a processor to implement the method for analyzing the formula of an essence according to any one of claims 1-10. ​

Images