A method for cigarette formulation design based on numerical control of tobacco calorific value.

Abstract

This invention discloses a cigarette formulation design method based on calorific value adjustment. The method includes the following steps: S1, collecting calorific value data; S2, constructing a tobacco leaf calorific value prediction model based on near-infrared spectroscopy; S3, extending the model in step S2 to stockpiled tobacco leaves; S4, designing cigarette formulations based on calorific value screening; S5, improving the permeability and aroma of the formulation by selecting alternative formulations with significantly higher calorific values ​​than the original formulation; and improving the smoothness and reducing the irritation of the formulation by selecting alternative formulations with significantly lower calorific values ​​than the original formulation. This invention, without changing the basic components of the existing formulation, including total sugar, reducing sugar, nicotine, chlorine, potassium, total nitrogen, pH, etc., designs cigarette formulations of different styles and qualities by adjusting the calorific value, and achieves rapid detection of tobacco leaf calorific value.

Description

Technical Field

[0001] This invention belongs to the field of cigarette formulation research technology, specifically relating to a cigarette formulation design method based on numerical control of calorific value. Background Technology

[0002] The smoking process of cigarettes mainly involves two types of chemical reactions: pyrolysis of tobacco leaves and combustion. Combustion primarily refers to the interaction between the smoke components and solid coke produced by pyrolysis and oxygen, generating a series of small molecules such as CO2 and CO, and releasing a large amount of heat. The heat released by tobacco combustion is the main heat source in the cigarette smoking process, and it directly determines the temperature field distribution inside the cigarette combustion cone. This temperature field distribution is closely related to the formation of key smoke components. Therefore, the heat released during tobacco combustion has a significant impact on smoke generation and the sensory quality of cigarettes.

[0003] Previously, researchers used combustion heat release analysis techniques based on the oxygen consumption principle, such as cone calorimeters and micro-combustion calorimeters, to characterize the combustion performance of raw materials. However, the correlation between combustion performance indicators and tobacco quality, as well as their application value in digital formulation design, have not yet been explored.

[0004] Calorific value, as one of the most basic indicators for measuring the thermal conversion efficiency of raw materials, has been widely used in the fields of coal and biomass, but its application in tobacco, a special biomass material system, has not been reported. Summary of the Invention

[0005] To address the aforementioned technical problems in existing technologies, this invention provides a cigarette formulation design method based on numerical control of tobacco leaf calorific value. This invention fills the technological gap in utilizing calorific value—a physical indicator closely related to cigarette combustion state, smoke generation, and cigarette quality—for cigarette formulation design, providing new ideas and methods for digital formulation design. The technical solution adopted by this invention is as follows:

[0006] A cigarette formulation design method based on numerical control of calorific value, characterized by the following steps:

[0007] S1. Collect calorific value data;

[0008] S2. Construct a prediction model for the calorific value of tobacco leaves based on near-infrared spectroscopy;

[0009] S3. Extend the model from step S2 to the stock of tobacco leaves;

[0010] S4. Design cigarette formulas based on calorific value screening;

[0011] S5. By selecting alternative formulas with significantly higher calorific value than the original formula, the permeability and aroma of the formula are improved; by selecting alternative formulas with significantly lower calorific value than the original formula, the smoothness of the formula is improved and the irritation is reduced.

[0012] Furthermore, in step S1, the specific process for collecting calorific value data is as follows:

[0013] The calorific value (J / g) of tobacco leaf samples was determined using a calorimeter in accordance with the national standard GB / T 213-2008. The calorific value refers to the bomb calorific value, which represents the heat released when a unit mass of sample is burned in an oxygen bomb filled with excess oxygen. The combustion products consist of oxygen, nitrogen, carbon dioxide, nitric acid and sulfuric acid, liquid water and solid ash. The bomb calorific value is the constant-volume higher heating value of the sample.

[0014] Furthermore, in step S2, the near-infrared spectrum of the tobacco leaf sample from step S1 is acquired using a Fourier transform near-infrared spectrometer, with the acquisition range being 10000–3800 cm⁻¹. -1 The near-infrared spectra of tobacco leaf samples were correlated with their corresponding calorific values, and a partial least squares algorithm was used to construct a tobacco leaf calorific value prediction model based on near-infrared spectra.

[0015] Furthermore, in step S3, the tobacco leaf calorific value prediction model obtained in step S2 is extended to tobacco leaves in storage, and the calorific value of the stored tobacco leaves is detected and analyzed using the near-infrared spectrum of the stored tobacco leaves.

[0016] Furthermore, to maintain the existing formula's composition unchanged (with relative changes of less than 5% for each component), including total sugar, reducing sugar, nicotine, chlorine, potassium, total nitrogen, pH, etc., a new formula is formed by listing the same number of tobacco leaves as the existing formula from the stock tobacco leaf samples, and then sorting them according to the calorific value of each new formula.

[0017] Compared with the prior art, the beneficial effects of the present invention are reflected in:

[0018] 1. This invention, without changing the basic components of the existing formula, including total sugar, reducing sugar, nicotine, chlorine, potassium, total nitrogen, pH, etc., designs cigarette formulas of different styles and qualities by adjusting the calorific value, and realizes rapid detection of the calorific value of tobacco leaves;

[0019] 2. Under the condition that conventional tobacco chemical indicators such as total sugar, reducing sugar, nicotine, chlorine, potassium, total nitrogen, and pH are set to fixed values, this invention designs cigarette formulas with different quality gradients by screening the calorific value.

[0020] 3. This invention achieves formulation design in two quality directions by adjusting the calorific value while keeping the original chemical composition of the tobacco unchanged. Attached Figure Description

[0021] Figure 1 These are near-infrared spectra of different tobacco powder samples in this invention;

[0022] Figure 2 These are the regression coefficients of the calorific value prediction model based on near-infrared spectroscopy in this invention;

[0023] Figure 3 This is a scatter plot of the predicted and actual results of calorific value in this invention, where O represents training set samples and * represents test set samples.

[0024] Figure 4 This is a flowchart of the design method of the present invention. Detailed Implementation

[0025] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0026] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0027] The following will refer to the appendix. Figures 1-4 The present invention will be described in detail with reference to exemplary embodiments.

[0028] (1) Detecting calorific value:

[0029] After drying 199 samples of re-dried tobacco flakes at (40±1)℃ for 4 hours, they were ground into powder using a Foss Cyclotec CT410 cyclone mill (Foss AG, Denmark). The powder samples were then passed through a 250μm (60 mesh) sieve and placed in sealed bottles for later use.

[0030] After air-drying the tobacco powder samples, the calorific value (J / g) of the samples was determined using a 5E-KC5410 calorimeter from Changsha Kaiyuan Hongsheng Technology Co., Ltd., according to the national standard GB / T 213-2008. The calorific value measured by this method refers to the bomb calorific value, which represents the heat released when a unit mass of the sample is burned in an oxygen bomb filled with excess oxygen, producing oxygen, nitrogen, carbon dioxide, nitric acid and sulfuric acid, liquid water, and solid ash. The bomb calorific value is the constant-volume higher heating value of the analyzed sample.

[0031] (2) Detection of near-infrared spectrum:

[0032] Take an appropriate amount of tobacco powder (1) (the height of which is about 1 / 3 of the total height of the sample cup), put it into the sample cup, place a fixed mass weight on top of the sample, let it compact naturally, and then perform near-infrared scanning.

[0033] Near-infrared spectra were acquired using an Antaris II Fourier transform near-infrared spectrometer manufactured by Thermo Fisher Scientific, USA. The acquisition range was 10,000–3,800 cm⁻¹, with a spectral resolution of 8 cm⁻¹; 64 scans were performed. To avoid the influence of scattering, the near-infrared spectra were smoothed using Savitzky-Golay (SG) and corrected for standard normality.

[0034] (3) Construct a prediction model for the calorific value of tobacco leaves based on near-infrared spectroscopy:

[0035] 199 types of tobacco leaves were numbered from 1 to 199 and randomly divided into 149 training samples and 50 test samples, with a training set to test set ratio of 3:1. The 149 training samples were used for model building, and the 50 test samples were used for model accuracy verification and evaluation. A partial least squares algorithm written in Matlab was used to establish a tobacco leaf calorific value prediction model based on near-infrared spectroscopy through internal cross-validation. The model regression coefficients are shown below. Figure 2 As shown. The calorific value and near-infrared spectrum at 4400 cm⁻¹ -1 5000cm -1 The left-right correlation is the strongest, at 4400cm. -1 The peak at 5000 cm⁻¹ mainly corresponds to the combination frequency of the stretching vibration and stretching vibration of methyl CH₄. -1 The peaks around the left and right may be caused by the combination of the stretching vibration and bending vibration of the OH group in the sugar.

[0036] The calorific value mainly depends on the elemental composition of the sample. During combustion, C, H and O in oxygen combine to release heat. Therefore, C and H contribute significantly to the calorific value, which is consistent with the regression coefficient results of the model.

[0037] By inputting the near-infrared spectra of 50 test set samples into the established model, the calorific value of the test set samples can be calculated and compared with the measured values. The accuracy of the model is evaluated and validated by calculating the root mean square error of the test set (RMSEP). Table 1 lists the values ​​of the model's training set root mean square error (RMSEC), cross-validation set root mean square error (RMSECV), and prediction set root mean square error (RMSEP). Figure 3Scatter plots showing the predicted and actual values ​​of 199 tobacco samples from the test and training sets are presented. The validation results from the test set and the scatter plots demonstrate that a linear model between near-infrared spectra and calorific value can be established using the partial least squares method, with an RMSEP of only 151.64, achieving the desired model performance.

[0038] Table 1. Analysis results of the tobacco calorific value prediction model based on near-infrared spectroscopy

[0039] Calorific value (J / g) LV RMSEC RMSECV RMSEP Mean(Y) 7 195.47 263.46 151.64 17003.69

[0040] *LV represents the number of latent variables in the model, and Mean(Y) represents the average calorific value of the training set samples. (4) Extend the model in (2) to the stock tobacco:

[0041] The calorific value prediction model was extended to tobacco samples from multiple production areas from 2018 to 2021. Based on the original 199 samples, the calorific value of 378 newly added tobacco samples was predicted using the model in (2). The calorific value of 577 tobacco samples was statistically analyzed by production area and part of the tobacco plant, and the results are shown in Table 2.

[0042] Table 2. Average calorific value of tobacco leaves from different parts of the plant in different provinces

[0043]

[0044]

[0045] As shown in Table 2, the calorific value of tobacco leaves varies significantly across different parts of the plant. The upper part has the highest calorific value, while the lower part has the lowest, indicating that the calorific value of the tobacco leaf gradually decreases with decreasing part size. This same pattern of calorific value variation exists across the upper, middle, and lower parts of the tobacco leaves in various production areas.

[0046] Analysis of tobacco leaves from the central parts of various producing regions revealed that those from Shandong, Sichuan, and Henan had lower calorific values; those from Guangxi, Yunnan, Anhui, Guizhou, Hunan, Canada, and Fujian had medium calorific values; and those from Brazil, Hubei, Malawi, Tanzania, the United States, Zambia, and Zimbabwe had higher calorific values. Comparing sensory evaluations of the three types of producing regions, those with high calorific values ​​exhibited a fuller aroma but lacked smoothness and a slightly less comfortable taste; those with low calorific values ​​produced a smooth and delicate smoke with a comfortable taste but lower aroma content; and those with medium calorific values ​​had a relatively balanced aroma, smoke, and taste. Blending different types of producing regions can better bring out the characteristics of the tobacco leaves from each region, resulting in a more harmonious blend.

[0047] In summary, calorific value is significantly correlated with quality dimensions such as tobacco leaf part and sensory properties, making it a potential key indicator in formulation design and evaluation.

[0048] (5) Design of cigarette formulations based on calorific value adjustment:

[0049] Existing formulas A and B were used as control samples, respectively.

[0050] Table 3 Composition, main chemical components and calorific value of Formula A

[0051] Total sugar Nicotine reducing sugars chlorine Potassium Total nitrogen pH Calorific value (J / g) Formula A 31.32 2.53 26.70 0.42 2.06 2.03 5.31 16966.06

[0052] Table 4 Composition, main chemical components and calorific value of Formula B

[0053] Total sugar Nicotine reducing sugars chlorine Potassium Total nitrogen pH Calorific value (J / g) Formula B 30.05 2.48 25.1 0.39 2.04 2.1 5.23 16968.45

[0054] For formula A, two formulas are designed, each consisting of the same number of tobacco leaves as formula A, with each tobacco leaf used in a 25% proportion. By constraining the variations in various chemical values, different blending schemes are selected from the stock of tobacco leaves. Through design, under the constraint that the chemical components are basically the same as those in formula A, low-calorific-value formula 1 and high-calorific-value formula 2 are found.

[0055] Table 5. Alternative formulations for formulation A: Low-calorific-value formulation 1 and high-calorific-value formulation 2

[0056]

[0057] Table 6 provides alternative formulations for formulation B, including low-calorific-value formulation 3 and high-calorific-value formulation 4.

[0058]

[0059] A comparative smoking evaluation method was adopted, with an evaluation team composed of seven smokers (all male, aged 26-38, with over 5 years of experience in cigarette smoking) specializing in tobacco raw materials and formulation. Cigarettes were manually rolled and balanced in a balancing chamber for 24 hours before evaluation. Using the company's tobacco quality evaluation methods, blind samples were numbered, and the evaluation team used a combination of quality comparison and written description to evaluate aspects such as aroma quality, aroma intensity, irritation, aftertaste, clumping, off-flavors, and smoothness. The results are shown in Table 7.

[0060] Table 7. Sensory evaluation results of the formulation

[0061] Evaluation results Key differences from Formula A Formula 1 Five people believe that Formula 1 differs from Formula A. Formula 1 has a clear and mellow tobacco aroma, with no other significant differences. Formula 2 Six people believed that Formula 2 differed from Formula A. Formula 2 has a strong tobacco aroma and noticeable stimulating effect; otherwise, there are no significant differences. Evaluation results Key differences from Formula B Formula 3 Six people believe that Formula 3 differs from Formula B. Formula 3 has a delicate and mellow aroma, with no other significant differences. Formula 4 Five people believe that formula 4 differs from formula B. Formula 4 has a strong, full aroma, with no other significant differences.

[0062] Sensory evaluation results from the two formulations show that even with similar levels of key chemical indicators such as total sugar, reducing sugar, nicotine, chlorine, potassium, total nitrogen, and pH, differences in calorific value result in different sensory qualities. The formulation with lower calorific value exhibits a clearer and smoother aroma, while the formulation with higher calorific value has a more pronounced aroma. These formulation experiments demonstrate that calorific value, as a parameter measuring the heat release characteristics of raw materials, can add another dimension to the quality characterization of the formulation, building upon the conventional chemical indicators of tobacco leaves.

[0063] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A cigarette formulation design method based on numerical control of calorific value, characterized in that, The method includes the following steps: S1. Collect calorific value data; S2. Construct a prediction model for the calorific value of tobacco leaves based on near-infrared spectroscopy; S3. Extend the model from step S2 to the stock of tobacco leaves; S4. Design cigarette formulas based on calorific value screening; S5. By selecting alternative formulas with significantly higher calorific value than the original formula, the permeability and aroma of the formula are improved. By selecting alternative formulas with significantly lower caloric value than the original formula, the smoothness of the formula is improved and its irritation is reduced. In step S1, the specific process for collecting calorific value data is as follows: The calorific value of tobacco leaf samples was determined using a calorimeter in accordance with the national standard GB / T 213-2008. The calorific value refers to the bomb calorific value, which represents the heat released when a unit mass of sample is burned in an oxygen bomb filled with excess oxygen. The combustion products consist of oxygen, nitrogen, carbon dioxide, nitric acid and sulfuric acid, liquid water and solid ash. The bomb calorific value is the constant volume higher heating value of the sample. In step S2, the near-infrared spectra of the tobacco leaf samples from step S1 are acquired using a Fourier transform near-infrared spectrometer, with the acquisition range being 10000–3800 cm⁻¹. -1 The near-infrared spectra of tobacco leaf samples were correlated with their corresponding calorific values, and a partial least squares algorithm was used to construct a calorific value prediction model for tobacco leaves based on near-infrared spectra. In step S3, the tobacco leaf calorific value prediction model obtained in step S2 is extended to the stock tobacco leaves, and the calorific value of the stock tobacco leaves is detected and analyzed using the near-infrared spectrum of the stock tobacco leaves. To maintain the existing formula's composition unchanged, including total sugar, reducing sugar, nicotine, chlorine, potassium, total nitrogen, and pH, a new formula is formed by listing the same number of tobacco leaves as the existing formula from the stock tobacco leaf samples, and then sorting them according to the calorific value of each new formula.

Images