Formula design method and device based on key chemical indexes

Abstract

The invention relates to the technical field of tobacco leaf processing, in particular to a formula design method based on key chemical indexes, which comprises the following steps of: calculating various derivative chemical indexes of different types of new tobacco leaf raw materials based on conventional component data of different types of new tobacco leaf raw materials, and selecting an iterative algorithm by adopting information gain and Fisher criteria to obtain a formula of the new tobacco leaf raw materials; key chemical indexes of the new tobacco leaf raw materials are screened out from conventional chemical components of the new tobacco leaf raw materials of different grades and derivative chemical indexes of the new tobacco leaf raw materials of different grades, the key chemical indexes of the tobacco leaf raw materials of different grades over the years are counted, and sensory evaluation scores of the new tobacco leaf raw materials of different grades are combined. Constructing a tobacco leaf quality evaluation model based on a chi-square function, inputting the obtained key chemical index data of the new tobacco leaf raw material sample into the tobacco leaf quality evaluation model based on the chi-square function to obtain a quality score of the new tobacco leaf raw material sample before matching, and designing a tobacco leaf formula in the redrying module based on the quality score of the new tobacco leaf raw material sample before matching.

Description

Technical Field

[0001] The present invention relates to the technical field of tobacco leaf processing, and in particular to a formula design method and device based on key chemical indicators. Background Art

[0002] Tobacco leaf formulation design is a very important link in the tobacco industry, which directly affects the quality, flavor and market competitiveness of the final product. Traditional tobacco leaf formulation design methods mainly rely on empirical judgment and sensory evaluation, lack a systematic scientific basis, and have limitations such as strong subjectivity, low efficiency, and waste of resources. However, with the development of near-infrared spectroscopy analysis technology, chemometric methods and machine learning algorithms, tobacco leaf quality evaluation and formulation design have gradually shifted from traditional sensory evaluation to more scientific and quantitative means.

[0003] However, key issues remain to be addressed in tobacco leaf quality evaluation and formulation design. For example, among the numerous chemical indicators in tobacco leaves, which ones should be used as the basis for efficient grading and formulation design? To date, there is no unified and effective industry norm or standard for intrinsic quality evaluation. While tobacco leaf quality evaluation methods based on chemical indicators are beginning to attract attention, they have not yet been closely integrated with tobacco leaf formulation design to achieve systematic optimization of tobacco leaf intrinsic quality and achieve optimal tobacco leaf quality.

[0004] Therefore, it is necessary to design a method that closely combines the tobacco leaf quality evaluation results with the tobacco leaf formula design, so as to solve the problem that the tobacco leaf quality evaluation results are not closely combined with the tobacco leaf formula design when designing the tobacco leaf formula, the intrinsic quality of the tobacco leaves cannot be optimized, and the industrial availability of low-quality tobacco leaves cannot be significantly improved. Summary of the Invention

[0005] The purpose of the present invention is to propose a formula design method and device based on key chemical indicators to solve the problem that when designing tobacco formula, tobacco quality evaluation results are not closely integrated with tobacco formula design, the intrinsic quality of tobacco cannot be optimized, and the industrial availability of low-quality tobacco cannot be significantly improved.

[0006] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present application proposes a formulation design method based on key chemical indicators, comprising the following steps:

[0008] Step 1. Obtain conventional chemical composition data of different types of new tobacco leaf raw materials, and based on the conventional composition data of different types of new tobacco leaf raw materials, calculate multiple derived chemical indicators of different types of new tobacco leaf raw materials; wherein the conventional composition data include total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine; the derived chemical indicators include total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine and total nitrogen / nicotine;

[0009] Step 2: Using information gain and Fisher criterion and then selecting an iterative algorithm, the key chemical indicators of new tobacco leaf raw materials are screened from the conventional chemical components of different grades of new tobacco leaf raw materials and the derived chemical indicators of different grades of new tobacco leaf raw materials. Among them, different grades of new tobacco leaf raw materials are classified based on the appearance and color of the new tobacco leaf raw materials.

[0010] Step 3. Using the key chemical indicators of the selected new tobacco leaf raw materials as a reference, the key chemical indicators of tobacco leaf raw materials of different grades over the years were counted, and combined with the sensory evaluation scores of new tobacco leaf raw materials of different grades, a tobacco leaf quality evaluation model based on the chi-square function was constructed; wherein, different grades of new tobacco leaf raw materials are classified based on the different appearance and sensory quality grades of new tobacco leaves;

[0011] Step 4. Obtain key chemical index data of the new tobacco leaf raw material sample, and input the key chemical index data of the new tobacco leaf raw material sample into the tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the new tobacco leaf raw material sample, that is, the quality score of the new tobacco leaf raw material sample before proportioning;

[0012] Step 5: Based on the quality scores of the new tobacco leaf raw material samples before blending, the tobacco leaf recipe in the redrying module is designed, and different grades of tobacco leaves are mixed in specific proportions. The key chemical indicators of the mixed tobacco leaf raw materials of different grades are then calculated. The different grades of tobacco leaves are graded based on the appearance and color of the tobacco leaf raw materials.

[0013] Step 6. Input the recalculated key chemical index data of different grades of tobacco leaf raw materials into the tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the tobacco leaf raw materials after proportioning, that is, the tobacco leaf raw material quality score after proportioning;

[0014] Step 7. Based on the comparison results of the quality score of the tobacco raw materials after proportioning in Step 6 and the quality score of the tobacco raw material samples before proportioning in Step 4, the process of Step 5 to Step 6 is executed sequentially multiple times to optimize the tobacco leaf formula list in the redrying module, and then obtain the optimized tobacco leaf formula list.

[0015] Preferably, the Step 1, obtaining conventional chemical composition data of different categories of new tobacco leaf raw materials, and calculating a plurality of derived chemical indicators of different categories of new tobacco leaf raw materials based on the conventional composition data of different categories of new tobacco leaf raw materials, comprises the following steps:

[0016] Step 1.1. Collect multiple varieties of freshly-cured tobacco leaves from different tobacco producing areas. Select freshly-cured tobacco leaves from different growing parts among the freshly-cured tobacco leaves of different varieties to obtain freshly-cured tobacco leaves of different categories.

[0017] Step 1.2, using an online near-infrared conventional chemical composition quantitative model, the conventional chemical component content of different types of new tobacco raw materials is measured to obtain conventional chemical composition data such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine of different types of new tobacco raw materials;

[0018] Step 1.3. Based on the conventional chemical composition data of different categories of new tobacco leaf raw materials, such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine, the derived chemical indicators such as total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, total nitrogen / nicotine of different categories of new tobacco leaf raw materials are calculated.

[0019] Preferably, the Step 2, using information gain and Fisher criterion to select an iterative algorithm, to screen out key chemical indicators of new tobacco leaf raw materials from conventional chemical components of different grades of new tobacco leaf raw materials and derived chemical indicators of different grades of new tobacco leaf raw materials, comprises the following steps:

[0020] Step 2.1. Based on the appearance and color of the new tobacco leaf raw materials, different types of new tobacco leaf raw materials are graded to obtain new tobacco leaf raw materials of different grades;

[0021] Step 2.2, determine the conventional chemical composition of different grades of new tobacco leaf raw materials and the derived chemical indicators of different grades of new tobacco leaf raw materials; wherein, conventional component data include total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine; derived chemical indicators include total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine and total nitrogen / nicotine;

[0022] Step 2.3. Use information gain and Fisher criterion to select iterative algorithm to screen out key chemical indicators of new tobacco leaf raw materials from the conventional chemical components of different grades of new tobacco leaf raw materials and the derived chemical indicators of different grades of new tobacco leaf raw materials; among them, conventional components include total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine; derived chemical indicators include total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine and total nitrogen / nicotine.

[0023] Preferably, the Step 3, using the key chemical indicators of the screened new tobacco raw materials as a reference, statistically analyzing the key chemical indicators of tobacco raw materials of different grades over the years, and combining the sensory evaluation scores of new tobacco raw materials of different grades to construct a tobacco quality evaluation model based on the chi-square function, includes the following steps:

[0024] Step 3.1. Using the selected key chemical indicators of the new tobacco leaf raw materials as a reference, select the key chemical indicators of different categories of tobacco leaf raw materials over the years and perform data statistics to obtain the key chemical indicator data statistics of different categories of tobacco leaf raw materials over the years; wherein the key chemical indicator data statistics of different categories of tobacco leaf raw materials over the years include the data mean and the data standard deviation;

[0025] Step 3.2: Based on the appearance and color of different categories of tobacco leaves over the years, classify the different categories of tobacco leaf raw materials over the years to obtain tobacco leaf raw materials of different grades over the years, and determine the statistical data of key chemical indicators of tobacco leaf raw materials of different grades over the years;

[0026] Step 3.3, based on the different appearance and sensory quality grades of different categories of new tobacco leaf raw materials, different categories of new tobacco leaf raw materials are graded to obtain new tobacco leaf raw materials of different grades;

[0027] Step 3.4, sensory evaluation and smoking scores are conducted on the cut tobacco made from the graded new tobacco leaves of different grades, so as to obtain sensory evaluation scores of the new tobacco leaves of different grades;

[0028] Step 3.5: Based on the sensory evaluation scores of different grades of new tobacco raw materials obtained in Step 3.3 and the statistical data of key chemical indicators of tobacco raw materials of different grades in previous years obtained in Step 3.2, design an evaluation curve of the key chemical indicator content of tobacco raw materials and the mass fraction of tobacco raw materials;

[0029] Step 3.6: Use genetic algorithm to find the optimal solution and fit the chi-square function to the evaluation curve of key chemical index content and tobacco leaf raw material mass fraction to obtain the key chemical index evaluation function. The key chemical index evaluation function expression is as follows:

[0030]

[0031] Among them, y is the evaluation value of key chemical indicators; M=A×(xC) 2 +B×(xC)+E, where A, B, C, D, and E are the parameters that constitute the evaluation function of the key chemical indicators of tobacco leaves. Adjusting these parameters A, B, C, D, and E can change the chemical indicator evaluation function to evaluate different chemical indicators;

[0032] Step 3.7, evaluate the function y for each key chemical index respectively i The weight ratio is set, and the evaluation function of each key chemical index after the weight ratio is summed to construct a tobacco leaf quality evaluation model based on the chi-square function. The tobacco leaf quality evaluation function expression is as follows:

[0033]

[0034] Among them, Y is the comprehensive quality evaluation score of tobacco leaves; N is the number of chemical indicators involved in the evaluation; w i is the weight of the i-th chemical index evaluation function in tobacco leaf quality evaluation, y i is the evaluation function of the i-th chemical index.

[0035] Preferably, Step 4, obtaining key chemical index data of the new tobacco leaf raw material sample, and inputting the key chemical index data of the new tobacco leaf raw material sample into a tobacco leaf quality evaluation model based on a chi-square function to obtain a tobacco leaf quality score of the new tobacco leaf raw material sample, that is, a score of the new tobacco leaf raw material sample before proportioning, comprises the following steps:

[0036] Step 4.1. Use an online near-infrared conventional chemical composition quantitative model to measure the conventional chemical composition content of the new tobacco leaf raw material sample to obtain the conventional chemical composition data of the new tobacco leaf raw material sample, such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine;

[0037] Step 4.2. Based on the conventional chemical composition data of the new tobacco leaf raw material sample, such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium, and chlorine, calculate the derived chemical indicators of the new tobacco leaf raw material sample, such as total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, and total nitrogen / nicotine;

[0038] Step 4.3, using information gain and Fisher criterion to select iterative algorithm, from the conventional chemical components of the new tobacco leaf raw material samples obtained in Step 4.1, such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine, and the derived chemical indicators of the new tobacco leaf raw material samples obtained in Step 4.2, such as total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, and total nitrogen / nicotine, to screen out the key chemical indicator data of the new tobacco leaf raw material samples;

[0039] Step 4.4: Input the key chemical index data of the new tobacco leaf raw material sample in Step 4.3 into the tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the new tobacco leaf raw material sample, that is, the quality score of the new tobacco leaf raw material sample before proportioning.

[0040] Preferably, the Step 5, designing the tobacco recipe in the redrying module based on the quality score of the new tobacco raw material sample before proportioning, mixing different grades of tobacco raw materials in a specific proportion, and then calculating the key chemical indicators of the mixed tobacco raw materials of different grades, comprises the following steps:

[0041] Step 5.1. Based on the quality scores of the new tobacco leaf raw material samples before blending, design the tobacco leaf formula in the redrying module to obtain a preliminary tobacco leaf formula;

[0042] Step 5.2: Based on the quality scores of the new tobacco leaf raw material samples before blending, select tobacco leaf raw materials with lower quality or less obvious style characteristics to form different grades of tobacco leaf raw materials;

[0043] Step 5.3: Based on the quality score of the new tobacco leaf raw material sample before proportioning, the different grades of tobacco leaf raw materials in Step 5.2 are mixed in a specific proportion, and the key chemical index data of the different grades of tobacco leaf raw materials after mixing are calculated to obtain the key chemical index data of the different grades of tobacco leaf raw materials.

[0044] Preferably, in Step 5.41, the conventional chemical component contents of different grades of tobacco leaf raw materials are measured using an online near-infrared conventional chemical component quantification model to obtain conventional chemical component data such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine of different grades of tobacco leaf raw materials;

[0045] Step 5.42. Based on the conventional chemical composition data of different grades of tobacco leaf raw materials, such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium, and chlorine, calculate the derived chemical indicators such as total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, and total nitrogen / nicotine of different grades of tobacco leaf raw materials;

[0046] Step 5.43. Use information gain and Fisher criterion to select iterative algorithm again to screen out key chemical indicator data of different grades of tobacco leaf raw materials from the conventional chemical components such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine of different grades of tobacco leaf raw materials obtained in Step 5.41 and the derived chemical indicators such as total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, total nitrogen / nicotine of different grades of tobacco leaf raw materials obtained in Step 5.42.

[0047] Preferably, Step 7, based on the comparison result of the quality score of the tobacco raw material after proportioning in Step 6 and the quality score of the tobacco raw material sample before proportioning in Step 4, repeatedly executing the process of part of Step 5 to Step 6 to optimize the tobacco leaf formula list in the redrying module, thereby obtaining the optimized tobacco leaf formula list, includes the following steps:

[0048] Step 7.1. Based on the comparison of the quality scores of the tobacco raw materials after blending in Step 6 and before blending in Step 4, design the tobacco recipe for the redrying module, mix different grades of new tobacco raw materials in specific proportions, and then calculate the key chemical indicators of the different grades of mixed new tobacco raw materials; the different grades of tobacco raw materials are classified based on their appearance and color;

[0049] Step 7.2, inputting the recalculated key chemical index data of different grades of tobacco leaf raw materials into a tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the tobacco leaf raw materials after blending, i.e., the tobacco leaf raw material quality score after blending;

[0050] Step 7.3: Repeat Step 7.1-Step 7.2 multiple times to optimize the tobacco leaf formula in the redrying module, and then obtain the optimized tobacco leaf formula.

[0051] In a second aspect, the present application proposes a formulation design device based on key chemical indicators, comprising:

[0052] The data acquisition and calculation module is used to obtain conventional chemical composition data of different types of new tobacco leaf raw materials, and based on the conventional composition data of different types of new tobacco leaf raw materials, calculate a variety of derived chemical indicators of different types of new tobacco leaf raw materials; wherein, conventional composition data include total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine; derived chemical indicators include total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine and total nitrogen / nicotine;

[0053] The screening and grading module and the data acquisition and calculation module are used to select the key chemical indicators of new tobacco leaf raw materials from the conventional chemical components of different grades of new tobacco leaf raw materials and the derived chemical indicators of different grades of new tobacco leaf raw materials by using information gain and Fisher criterion and then selecting iterative algorithms; among them, different grades of new tobacco leaf raw materials are classified based on the appearance and color of the new tobacco leaf raw materials;

[0054] The quality evaluation module is connected to the screening and grading module and is used to use the key chemical indicators of the screened new tobacco leaf raw materials as a reference, to compile statistics on the key chemical indicators of tobacco leaf raw materials of different grades over the years, and to build a tobacco leaf quality evaluation model based on the chi-square function in combination with the distribution patterns of the key chemical indicators of new tobacco leaf raw materials of different grades. The key chemical indicator data of the new tobacco leaf raw material samples are then input into the tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the new tobacco leaf raw material samples, that is, the quality score of the new tobacco leaf raw material samples before proportioning;

[0055] The tobacco recipe design module, connected to the quality evaluation module, is used to design the tobacco recipe in the redrying module based on the quality score of the new tobacco raw material samples before proportioning, mix different grades of tobacco raw materials in specific proportions, and then calculate the key chemical indicators of the mixed tobacco raw materials of different grades. Among them, different grades of tobacco raw materials are classified based on their appearance and color.

[0056] The tobacco leaf formula optimization module is connected to the tobacco leaf formula design module and is used to input the recalculated key chemical index data of different grades of tobacco leaf raw materials into the tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the tobacco leaf raw materials after proportioning, that is, the quality score of the tobacco leaf raw materials after proportioning, and based on the comparison results of the quality score of the tobacco leaf raw materials after proportioning and the quality score of the tobacco leaf raw material samples before proportioning, the processing process from the tobacco leaf formula design module to the tobacco leaf formula optimization module is executed in a loop multiple times to optimize the tobacco leaf formula list in the redrying module, and then obtain the optimized tobacco leaf formula list.

[0057] Preferably, a recipe design device based on key chemical indicators further includes a ratio execution module and a label printing module;

[0058] The proportion execution module is connected to the tobacco leaf formula optimization module and is used to find different grades of tobacco leaves based on the optimized tobacco leaf formula and mix them evenly multiple times in a specific proportion;

[0059] The label printing module is connected to the ratio execution module and is used to print the tobacco leaf formula sheet corresponding to the tobacco leaves after mixing to obtain a tobacco leaf formula information label.

[0060] Compared with the prior art, the present invention has at least one of the following beneficial effects:

[0061] 1. A formula design method based on key chemical indicators in the present invention first obtains conventional chemical composition data of new tobacco leaf raw materials of different categories, and based on the conventional component data of new tobacco leaf raw materials of different categories, calculates multiple derived chemical indicators of new tobacco leaf raw materials of different categories, and then uses information gain and Fisher criterion to select an iterative algorithm to screen out key chemical indicators of new tobacco leaf raw materials from the conventional chemical components of new tobacco leaf raw materials of different grades and the derived chemical indicators of new tobacco leaf raw materials of different grades. Then, with the key chemical indicators of the screened new tobacco leaf raw materials as a reference, the key chemical indicators of tobacco leaf raw materials of different grades over the years are statistically analyzed, and the key chemical indicators of new tobacco leaf raw materials of different grades are combined with the distribution law of key chemical indicators of new tobacco leaf raw materials of different grades to construct a basic formula design method. A tobacco leaf quality evaluation model based on the chi-square function is developed, and then the key chemical index data of the obtained new tobacco leaf raw material samples are input into the tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the new tobacco leaf raw material sample, that is, the quality score of the new tobacco leaf raw material sample before proportioning. Finally, based on the quality score of the new tobacco leaf raw material sample before proportioning, the tobacco leaf formula in the redrying module is designed. In this way, not only can the intrinsic quality of tobacco leaves be optimized and the industrial availability of low-quality tobacco leaves be improved, but also a scientific and reasonable formula design basis can be provided for the tobacco industry to solve the problem that when designing tobacco leaf formulas, the tobacco leaf quality evaluation results are not closely integrated with the tobacco leaf formula design, the intrinsic quality of tobacco leaves cannot be optimized, and the industrial availability of low-quality tobacco leaves cannot be significantly improved.

[0062] 2. The present invention screens out key chemical indicators that affect tobacco leaf quality (such as the reducing sugar-nitrogen ratio, sugar-alkali ratio, nitrogen-alkali ratio, etc.) and establishes a tobacco leaf quality evaluation model based on the chi-square function. This not only realizes data-driven formula design, but also significantly improves scientificity and objectivity. At the same time, by closely combining the tobacco leaf quality evaluation results with formula design, it provides a scientific basis for tobacco leaf quality optimization.

[0063] 3. The present invention utilizes online near-infrared spectroscopy technology in combination with a tobacco leaf quality evaluation model based on the chi-square function, which not only enables rapid evaluation of the intrinsic quality of tobacco leaves, but also significantly improves the efficiency of formula design.

[0064] 4. The present invention optimizes the formula ratio through multiple iterations, which can not only dynamically adjust the formula design according to the specific conditions of the tobacco raw materials, but also is applicable to tobacco raw materials of different origins and varieties.

[0065] 5. By optimizing the formula design and improving the industrial availability of low-quality tobacco leaves, the present invention not only significantly reduces production costs, but also improves product quality and creates higher economic benefits for enterprises. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 This is a flow chart of a tobacco leaf formula design method based on key chemical indicators according to an embodiment of the present invention.

[0067] Figure 2 This is an evaluation curve of the content and mass fraction of key chemical indicators in a specific embodiment of the present invention.

[0068] Figure 3 This is a schematic structural diagram of a tobacco leaf formula design device based on key chemical indicators according to an embodiment of the present invention.

[0069] Table 1 shows the parameter values ​​of the quality score evaluation function of key chemical indicators.

[0070] Table 2 shows the quality evaluation scores of industrial graded tobacco leaves in Yuxi, Yunnan.

[0071] Table 3 shows the chemical indicators and quality evaluation results of tobacco leaves after blending. DETAILED DESCRIPTION

[0072] like Figure 1-3 As shown in Tables 1-3, in order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0073] Example 1

[0074] In the process of tobacco leaf quality evaluation and formula design, there are still some key issues that need to be solved urgently. For example, among the numerous chemical indicators in tobacco leaves, which specific chemical indicators should be used as the important basis for efficient grading and formula design? So far, there has been no unified and effective industry norms or standards for intrinsic quality evaluation. Although tobacco leaf quality evaluation methods based on chemical indicators have begun to attract attention, the tobacco leaf quality evaluation results have not been closely integrated with tobacco leaf formula design to achieve systematic optimization of tobacco leaf intrinsic quality and achieve optimal tobacco leaf quality.

[0075] Therefore, this application proposes a formula design method based on key chemical indicators to solve the problem that when designing tobacco formulas, the tobacco quality evaluation results are not closely integrated with the tobacco formula design, resulting in the failure to optimize the intrinsic quality of tobacco and significantly improve the industrial availability of low-quality tobacco.

[0076] For details, please refer to Figure 1 , Figure 1 This is a flow chart of a tobacco leaf formulation design method based on key chemical indicators according to an embodiment of the present invention. The formulation design method based on key chemical indicators comprises the following steps:

[0077] The first step is to obtain the conventional chemical composition data of different categories of new tobacco leaf raw materials, and based on the conventional composition data of different categories of new tobacco leaf raw materials, calculate a variety of derived chemical indicators of different categories of new tobacco leaf raw materials, as follows:

[0078] (1) Collecting a plurality of different varieties of newly cured tobacco leaves from different tobacco producing areas, and selecting newly cured tobacco leaves from different growing parts from among the newly cured tobacco leaves of different varieties to obtain new tobacco leaves of different categories.

[0079] Among them, different tobacco producing areas can be understood as different tobacco producing areas in Yuxi, Yunnan; different varieties of newly cured tobacco leaves can be understood as different varieties of tobacco such as red tobacco, yellow tobacco, white tobacco, medium and high tobacco, high-growing tobacco, Ava tobacco and willow-leaf yellow tobacco; different growing parts can be understood as different growing parts such as the top, middle and bottom of the tobacco plant.

[0080] In this embodiment, multiple varieties of newly cured tobacco leaves are collected from different tobacco producing areas, and newly cured tobacco leaves from different growing parts are selected from the newly cured tobacco leaves of different varieties to obtain new tobacco leaf raw materials of different categories, thereby providing support for subsequent prediction of conventional chemical components such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium, and chlorine in the new tobacco leaf raw materials of different categories.

[0081] (2) Using an online near-infrared conventional chemical composition quantitative model, the conventional chemical component content of different types of new tobacco leaf raw materials was measured to obtain conventional chemical composition data such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine of different types of new tobacco leaf raw materials;

[0082] Among them, the near-infrared conventional chemical composition quantitative model can be understood as a mathematical model that uses near-infrared spectroscopy technology (NIR) to quantitatively analyze the chemical components in the sample. This model establishes the relationship between spectral data and chemical components to achieve rapid and non-destructive detection of specific components in the sample; the conventional chemical component content can be understood as the content of total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine in new tobacco leaf raw materials.

[0083] This example uses an online near-infrared conventional chemical component quantification model to measure the conventional chemical component content of different categories of new tobacco leaf raw materials to obtain conventional chemical component data such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine of different categories of new tobacco leaf raw materials, providing support for the subsequent calculation of derived chemical indicators in different grades of new tobacco leaf raw materials.

[0084] (3) Based on the conventional chemical composition data of different types of new tobacco leaf raw materials, such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine, the derived chemical indicators such as total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, and total nitrogen / nicotine of different types of new tobacco leaf raw materials are calculated.

[0085] Among them, total sugar / total nitrogen = total sugar content of tobacco leaves / total nitrogen content of tobacco leaves; reducing sugar / total nitrogen = reducing sugar content of tobacco leaves / total nitrogen content of tobacco leaves; total sugar / nicotine = total sugar content of tobacco leaves / nicotine content of tobacco leaves; reducing sugar / nicotine = reducing sugar content of tobacco leaves / nicotine content of tobacco leaves; total nitrogen / nicotine = total nitrogen content of tobacco leaves / nicotine content of tobacco leaves; tobacco leaf derived chemical indicators are total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, total nitrogen / nicotine, etc.

[0086] This example calculates derived chemical indices such as total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, and total nitrogen / nicotine for different categories of new tobacco leaf raw materials based on conventional chemical composition data such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium, and chlorine, thereby providing support for subsequent screening of key chemical indices for new tobacco leaf raw materials.

[0087] In the second step, the information gain and Fisher criterion were used to select the iterative algorithm to screen out the key chemical indicators of new tobacco leaf raw materials from the conventional chemical components of different grades of new tobacco leaf raw materials and the derived chemical indicators of different grades of new tobacco leaf raw materials. The details are as follows:

[0088] (1) Based on the appearance and color of the new tobacco leaf raw materials, different categories of new tobacco leaf raw materials are graded to obtain new tobacco leaf raw materials of different grades.

[0089] Among them, the appearance of new tobacco leaf raw materials can be understood as the size of the new tobacco leaves, the shape of the new tobacco leaves, and the completeness of the new tobacco leaves; the color of the new tobacco leaf raw materials can be understood as the color of the new tobacco leaves; different grades of new tobacco leaf raw materials can be understood as determining the quality grade of new tobacco leaves based on the appearance and color of the new tobacco leaves, and using the quality grade of new tobacco leaves as the grading standard. The different grades of new tobacco leaf raw materials divided, for example, can be divided into different grades such as A1, A2, A3, B1, B2, B3............F1, F2, F3 based on the tobacco leaf quality grade.

[0090] This embodiment grades different types of new tobacco leaf raw materials based on their appearance and color to obtain new tobacco leaf raw materials of different grades, which can provide support for the subsequent determination of the conventional chemical composition of new tobacco leaf raw materials of different grades and the derived chemical indicators of new tobacco leaf raw materials of different grades.

[0091] (2) Determine the conventional chemical composition of different grades of new tobacco leaf raw materials and the derived chemical indicators of different grades of new tobacco leaf raw materials.

[0092] Among them, the conventional chemical composition of new tobacco leaf raw materials of different grades is determined based on the conventional chemical composition of new tobacco leaf raw materials of different categories; the conventional chemical composition of new tobacco leaf raw materials of different grades is determined based on the derived chemical indicators of new tobacco leaf raw materials of different categories; conventional component data include total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine; derived chemical indicators include total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine and total nitrogen / nicotine;

[0093] This embodiment can provide support for subsequent screening of key chemical indicators of new tobacco leaf raw materials by determining the conventional chemical components of new tobacco leaf raw materials of different grades and the derived chemical indicators of new tobacco leaf raw materials of different grades.

[0094] (3) Using information gain and Fisher criterion and then selecting iterative algorithm, the key chemical indicators of new tobacco leaf raw materials were screened from the conventional chemical components of different grades of new tobacco leaf raw materials and the derived chemical indicators of different grades of new tobacco leaf raw materials;

[0095] Among them, information gain is an important indicator used in the decision tree algorithm to select the best features for division; the Fisher criterion is to find the optimal projection direction by maximizing the inter-class distance and minimizing the intra-class distance. The goal of the Fisher criterion is to separate data of different categories as much as possible after projection, while clustering data of the same category; conventional components include total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine; derived chemical indicators include total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine and total nitrogen / nicotine; the key chemical indicators of new tobacco leaf raw materials can be understood as nicotine, reducing sugar / total nitrogen and total sugar / nicotine.

[0096] This embodiment uses information gain and Fisher criterion to select an iterative algorithm to screen out key chemical indicators of new tobacco leaf raw materials from the conventional chemical components of new tobacco leaf raw materials of different grades and the derived chemical indicators of new tobacco leaf raw materials of different grades, which can provide support for the subsequent screening of key chemical indicators of new tobacco leaf raw materials.

[0097] The third step is to use the key chemical indicators of the selected new tobacco raw materials as a reference, collect statistics on the key chemical indicators of tobacco raw materials of different grades over the years, and combine them with the sensory evaluation scores of new tobacco raw materials of different grades to construct a tobacco quality evaluation model based on the chi-square function, as follows:

[0098] (1) With reference to the key chemical indicators of the screened new tobacco leaf raw materials, the key chemical indicators of different categories of tobacco leaf raw materials in previous years were screened and statistically analyzed to obtain the statistical data of the key chemical indicators of different categories of tobacco leaf raw materials in previous years.

[0099] Among them, the statistical data of key chemical indicators in different categories of tobacco leaf raw materials over the years include data average values ​​and data standard deviations, as shown in Table 2, which shows the statistical data of key chemical indicators in different categories of tobacco leaf raw materials over the years, including data average values ​​and data standards;

[0100] Among them, taking the key chemical indicators of the screened new tobacco leaf raw materials as a reference, screening out the key chemical indicators of different categories of tobacco leaf raw materials in previous years can be understood as taking the key chemical indicators such as nicotine, reducing sugar / total nitrogen, total sugar / nicotine of the screened new tobacco leaf raw materials as a reference, screening out the key chemical indicators such as nicotine, reducing sugar / total nitrogen, total sugar / nicotine of different categories of tobacco leaf raw materials in previous years.

[0101] This embodiment uses the key chemical indicators of the screened new tobacco raw materials as a reference to screen out the key chemical indicators of different categories of tobacco raw materials in previous years and conduct data statistics to obtain the key chemical indicator data statistics of different categories of tobacco raw materials in previous years, which can provide support for the subsequent determination of the key chemical indicator data statistics of different grades of tobacco raw materials in previous years.

[0102] (2) Based on the appearance and color of different categories of tobacco leaves over the years, different categories of tobacco leaf raw materials over the years were graded to obtain tobacco leaf raw materials of different grades over the years, and the key chemical index data statistics of tobacco leaf raw materials of different grades over the years were determined, as shown in Table 2. Table 2 shows the quality evaluation scores of industrial graded tobacco leaves in Yuxi, Yunnan over the years. Among them, Table 2 involves the key chemical index data statistics of tobacco leaf raw materials of different grades over the years, that is, the average and standard deviation of the quality evaluation scores of tobacco leaves of different quality grades over the years.

[0103] Table 2 Quality evaluation scores of industrial graded tobacco leaves in Yuxi, Yunnan

[0104]

[0105] Note: “ / ” means there is no tobacco leaf of that grade in that year.

[0106] Among them, different categories of tobacco leaves over the years can be understood as collecting multiple different varieties of tobacco leaves from different origins, and selecting tobacco leaves from different growth parts among the different varieties of tobacco leaves over the years; the appearance of different categories of tobacco leaves can be understood as the size of tobacco leaves over the years, the shape of tobacco leaves over the years, and the completeness of the leaves of tobacco leaves over the years; the external color of different categories of tobacco leaves can be understood as the color of tobacco leaves over the years; different grades of tobacco raw materials over the years can be understood as determining the quality grade of tobacco leaves over the years based on the appearance and color of tobacco leaves over the years (the quality grade of tobacco leaves over the years is consistent with the grading standard of the new tobacco quality grade mentioned above), and using the quality grade of tobacco leaves over the years as the grading standard, the tobacco raw materials of different grades over the years can be divided into different grades such as A1, A2, A3, B1, B2, B3............F1, F2, F3, etc.

[0107] This embodiment classifies different categories of tobacco raw materials over the years based on their appearance and color to obtain tobacco raw materials of different grades over the years, and determines the statistical data of key chemical indicators of tobacco raw materials of different grades over the years, so as to provide support for the subsequent design of evaluation curves of the key chemical indicator content of tobacco raw materials and the mass fraction of tobacco raw materials.

[0108] (3) Based on the different appearance and sensory quality grades of different categories of new tobacco leaf raw materials, different categories of new tobacco leaf raw materials are graded to obtain new tobacco leaf raw materials of different grades.

[0109] Among them, different categories of new tobacco leaf raw materials can be understood as collecting multiple different varieties of newly cured new tobacco leaves from different tobacco producing areas. Among the newly cured new tobacco leaf raw materials of different varieties, newly cured new tobacco leaf raw materials from different growth parts are selected to obtain different categories of new tobacco leaf raw materials; the different appearances of new tobacco leaf raw materials can be understood as the shape, size and integrity of tobacco leaves; the sensory quality of new tobacco leaves can be understood as the color and smell of new tobacco leaves; different grades of new tobacco leaf raw materials can be understood as the determination of new tobacco leaf raw materials based on their different appearances and sensory quality grades. Quality grade (the grade standards of new tobacco quality grade, previous years' tobacco quality grade and new tobacco leaf quality grade may be approximately the same, or the grade standards of new tobacco leaf quality grade, new tobacco leaf quality grade and previous years' tobacco leaf quality grade may be different). New tobacco leaf quality grade is used as the grading standard, and different grades of new tobacco leaf raw materials are divided. For example, new tobacco leaf raw materials can be divided into different grades such as A1, A2, A3, B1, B2, B3............F1, F2, F3 based on the new tobacco leaf quality grade.

[0110] This embodiment classifies different categories of new tobacco leaf raw materials based on their different appearances and sensory quality grades to obtain new tobacco leaf raw materials of different grades, which can provide support for subsequent sensory evaluation and scoring of tobacco made from the graded new tobacco leaf raw materials of different grades.

[0111] (4) sensory evaluation and smoking scores are conducted on the tobacco cuts made from the graded new tobacco leaf raw materials of different grades to obtain the sensory evaluation and smoking scores of the new tobacco leaf raw materials of different grades;

[0112] Among them, the sensory evaluation scores of different grades of new tobacco leaf raw materials can be understood as the scores made after smoking the tobacco made from different grades of new tobacco leaf raw materials.

[0113] This embodiment conducts sensory evaluation and scoring on tobacco cut into pieces made from different grades of graded new tobacco raw materials to obtain sensory evaluation scores of different grades of new tobacco raw materials, which can provide support for the subsequent design of evaluation curves of the content of key chemical indicators of tobacco raw materials and the mass fraction of tobacco raw materials.

[0114] (5) Based on the sensory evaluation scores of different grades of new tobacco raw materials and the statistical data of key chemical indicators of tobacco raw materials of different grades in previous years, an evaluation curve of key chemical indicator content of tobacco raw materials and tobacco raw materials mass fraction was designed (see Figure 2 , Figure 2 is the evaluation curve of key chemical index content and mass fraction);

[0115] Among them, the key chemical indicators of tobacco leaf raw materials are nicotine, reducing sugar / total nitrogen, and total sugar / nicotine. The key chemical indicator contents of tobacco leaf raw materials are nicotine content, reducing sugar / total nitrogen content, and total sugar / nicotine content.

[0116] (6) Genetic algorithm was used to find the optimal solution, and the chi-square function was fitted to the evaluation curve of the key chemical index content of tobacco leaf raw materials and the mass fraction of tobacco leaf raw materials to obtain the key chemical index evaluation function. The key chemical index evaluation function expression is as follows:

[0117]

[0118] Among them, y is the evaluation value of key chemical indicators; M=A×(xC) 2 +B×(xC)+E, A, B, C, D and E are the parameters that constitute the evaluation function of the key chemical index of tobacco leaves. See Table 1, which shows the parameter values ​​of the evaluation function of the mass score of the key chemical index. Adjusting the parameters A, B, C, D and E in the table can change the chemical index evaluation function to evaluate different chemical indexes;

[0119] Table 1 Parameter values ​​of the key chemical index quality score evaluation function

[0120]

[0121] For example, the evaluation score of the key chemical indicator reducing sugar / total nitrogen ratio of tobacco leaves is calculated using the key chemical indicator evaluation function and the parameter values ​​corresponding to the reducing sugar / total nitrogen ratio in Table 1. The specific calculation formula is as follows:

[0122]

[0123] Among them, x1 is the reducing sugar / total nitrogen data in tobacco raw materials; y i It is the evaluation score of the reducing sugar / total nitrogen ratio, a key chemical indicator of tobacco leaves;

[0124] When the key chemical index evaluation function is used and the parameter values ​​corresponding to the total nitrogen / nicotine ratio in Table 1 are combined to calculate the evaluation score of the tobacco leaf key chemical index total nitrogen / nicotine ratio, the specific calculation formula is as follows:

[0125]

[0126] Among them, x2 is the reducing sugar / total nitrogen ratio data in tobacco leaf raw materials; y2 is the evaluation score of the total nitrogen / nicotine ratio, a key chemical indicator of tobacco leaves;

[0127] When the key chemical index evaluation function is used and the parameter values ​​corresponding to the nicotine content in Table 1 are combined to calculate the evaluation score of the key chemical index nicotine in tobacco leaves, the specific calculation formula is as follows:

[0128]

[0129]

[0130] Among them, x3 is the reducing sugar / total nitrogen ratio data in tobacco raw materials; y3 is the evaluation score of nicotine, a key chemical indicator of tobacco leaves;

[0131] Among them, genetic algorithm is a type of randomized search method evolved from the evolutionary laws of the biological world. It adopts a probabilistic optimization method, can automatically obtain and guide the optimized search space, and adaptively adjust the search direction without the need for definite rules.

[0132] In this example, a genetic algorithm is used for optimization and a chi-square function is fitted to the evaluation curve of the key chemical index content and the mass fraction of tobacco raw materials to obtain the key chemical index evaluation function, which can provide support for the subsequent construction of a tobacco quality evaluation model based on the chi-square function.

[0133] (7) Evaluation function y for each key chemical index iThe weight ratio is set, and the evaluation function of each key chemical index after the weight ratio is summed to construct a tobacco leaf quality evaluation model based on the chi-square function. The tobacco leaf quality evaluation function expression is as follows:

[0134]

[0135] Among them, Y is the comprehensive quality evaluation score of tobacco leaves; N is the number of chemical indicators involved in the evaluation; w i is the weight of the i-th chemical index evaluation function in tobacco leaf quality evaluation, y i is the evaluation function of the i-th chemical index.

[0136] For example, a. Using the key chemical indicator evaluation function, combined with the parameter values ​​corresponding to the reducing sugar / total nitrogen ratio in Table 1 and the weight w1 of reducing sugar / total nitrogen in the tobacco leaf quality evaluation model, the evaluation score of the key chemical indicator reducing sugar / total nitrogen ratio of tobacco leaves after weight processing is calculated. The specific calculation formula is as follows:

[0137]

[0138] Among them, x1 is the reducing sugar / total nitrogen data in tobacco raw materials; y1 is the evaluation score of the reducing sugar / total nitrogen ratio of tobacco key chemical indicators after weight processing; w1 is the weight of the reducing sugar / total nitrogen ratio in the tobacco quality evaluation model, w1 is 0.5;

[0139] b. Using the key chemical index evaluation function, combined with the parameter values ​​corresponding to the total nitrogen / nicotine ratio in Table 1 and the weight w2 of total nitrogen / nicotine in the tobacco leaf quality evaluation model, calculate the evaluation score of the tobacco leaf key chemical index total nitrogen / nicotine ratio after weight processing. The specific calculation formula is as follows:

[0140]

[0141] Among them, x2 is the reducing sugar / total nitrogen ratio data in tobacco raw materials; y2 is the evaluation score of the total nitrogen / nicotine ratio of tobacco key chemical indicators after weight processing; w2 is the weight of the reducing sugar / total nitrogen ratio in the tobacco quality evaluation model, w2 is 0.2;

[0142] c. Using the key chemical index evaluation function, combined with the parameter values ​​corresponding to the nicotine ratio in Table 1 and the weight w3 of nicotine in the tobacco leaf quality evaluation model, calculate the evaluation score of the tobacco leaf key chemical index nicotine ratio after weight processing. The calculation formula is as follows:

[0143]

[0144] Among them, x3 is the reducing sugar / total nitrogen ratio data in tobacco leaf raw materials; y3 is the evaluation score of the nicotine ratio of tobacco leaf key chemical indicators after weight processing; w3 is the weight of the reducing sugar / total nitrogen ratio in the tobacco leaf quality evaluation model, w3 is 0.3;

[0145] The evaluation score of the tobacco raw material tobacco leaf quality is obtained by summing up the above c. the evaluation score of the nicotine ratio of the key chemical indicator of tobacco leaves after weighting, b. the evaluation score of the reducing sugar / total nitrogen ratio of the key chemical indicator of tobacco leaves after weighting, and a. the evaluation score of the reducing sugar / total nitrogen ratio of the key chemical indicator of tobacco leaves after weighting.

[0146] In this embodiment, the key chemical index evaluation function y i The weight ratio is set, and the evaluation function of each key chemical indicator after the weight ratio is summed to construct a tobacco quality evaluation model based on the chi-square function, which provides support for the subsequent evaluation of the tobacco quality score of new tobacco raw material samples.

[0147] Step 4: Obtain key chemical index data of the new tobacco leaf raw material sample, and input the key chemical index data of the new tobacco leaf raw material sample into the tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the new tobacco leaf raw material sample, that is, the quality score of the new tobacco leaf raw material sample before proportioning, as follows:

[0148] (1) Using an online near-infrared conventional chemical component quantitative model, the conventional chemical component content of the new tobacco leaf raw material sample was measured to obtain the conventional chemical component data of the new tobacco leaf raw material sample, such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine;

[0149] (2) Based on the conventional chemical composition data of the new tobacco leaf raw material samples, such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine, the derived chemical indicators such as total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, and total nitrogen / nicotine of the new tobacco leaf raw material samples were calculated;

[0150] (3) Using information gain and Fisher criterion and then selecting iterative algorithm, the key chemical index data of the new tobacco leaf raw material samples were screened out from the conventional chemical components such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine of the new tobacco leaf raw material samples obtained in Step 4.1 and the derived chemical indicators such as total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, and total nitrogen / nicotine of the new tobacco leaf raw material samples obtained in Step 4.2;

[0151] (4) The key chemical index data of the new tobacco leaf raw material sample are input into the tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the new tobacco leaf raw material sample, that is, the quality score of the new tobacco leaf raw material sample before matching. Refer to Table 3. Taking B2 and E2 grade tobacco leaves as examples, before matching, the tobacco leaf quality score of B2 grade tobacco leaf raw material is 42, and the tobacco leaf quality score of E2 grade tobacco leaf raw material is 61.

[0152] Table 3 Chemical indexes and quality evaluation results of B2 and E2 grade tobacco leaves after blending

[0153]

[0154] Step 5: Based on the quality scores of the new tobacco leaf raw material samples before proportioning, design the tobacco leaf formula in the redrying module, mix different grades of tobacco leaves in specific proportions, and then calculate the key chemical indicators of the mixed tobacco leaf raw materials of different grades. See Table 3 for details as follows:

[0155] (1) Based on the quality score of the new tobacco leaf raw material samples before proportioning, the tobacco leaf formula in the redrying module is designed to obtain a preliminary tobacco leaf formula;

[0156] (2) Based on the quality scores of new tobacco leaf raw material samples before blending, select tobacco leaf raw materials with lower quality or less obvious style characteristics, such as B-grade tobacco and E1-grade tobacco, to form different grades of tobacco leaf raw materials;

[0157] (3) Based on the quality score of the new tobacco leaf raw material sample before proportioning, the different grades of tobacco leaf raw materials in (2) are mixed in a specific ratio. Specifically, refer to Table 3, such as the B grade tobacco leaf and the E1 grade tobacco leaf are mixed in a mass ratio of 8:2; such as the B grade tobacco leaf and the E1 grade tobacco leaf are mixed in a mass ratio of 7:3; such as the B grade tobacco leaf and the E1 grade tobacco leaf are mixed in a mass ratio of 6:4; such as the B grade tobacco leaf and the E1 grade tobacco leaf are mixed in a mass ratio of 5:5; such as the B grade tobacco leaf and the E1 grade tobacco leaf are mixed in a mass ratio of 4:6; such as the B grade tobacco leaf and the E1 grade tobacco leaf are mixed in a mass ratio of 3:7; such as the B grade tobacco leaf and the E1 grade tobacco leaf are mixed in a mass ratio of 2:8, and then calculate the key chemical index data of the different grades of tobacco leaf raw materials after the above mixing to obtain the key chemical index data of the different grades of tobacco leaf raw materials.

[0158] Among them, the key chemical indicators of different grades of tobacco leaf raw materials are reducing sugar / total nitrogen, total nitrogen / nicotine and nicotine; the key chemical indicator data of different grades of tobacco leaf raw materials are reducing sugar / total nitrogen ratio, total nitrogen / nicotine ratio and nicotine content.

[0159] Among them, when calculating the key chemical index data of different grades of tobacco leaf raw materials after mixing, the online near-infrared conventional chemical composition quantitative model can be used to first determine the conventional chemical component content of different grades of tobacco leaf raw materials to obtain the conventional chemical composition data of total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine of different grades of tobacco leaf raw materials. Then, based on the conventional chemical composition data of total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine of different grades of tobacco leaf raw materials, the derived chemical indicators such as total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, total nitrogen / nicotine of different grades of tobacco leaf raw materials are calculated, and then the results are analyzed. Information gain and Fisher criterion were used to select an iterative algorithm. Key chemical indicator data of tobacco leaf raw materials of different grades were screened out from conventional chemical components such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine, as well as derived chemical indicators such as total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, and total nitrogen / nicotine of tobacco leaf raw materials of different grades. Finally, the recalculated key chemical indicator data of tobacco leaf raw materials of different grades were input into a tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the tobacco leaf raw materials after proportioning, that is, the tobacco leaf raw material quality score after proportioning.

[0160] Table 3 Chemical indexes and quality evaluation results of B2 and E2 grade tobacco leaves after blending

[0161]

[0162] In the sixth step, the key chemical index data of different grades of tobacco raw materials that were recalculated were input into the tobacco quality evaluation model based on the chi-square function to obtain the tobacco quality scores of the tobacco raw materials after matching, that is, the tobacco raw material quality scores after matching. Referring to Table 3, after matching, the score of the tobacco leaves of grade B and grade E1 at a mass ratio of 8:2 was 66; the score of the tobacco leaves of grade B and grade E1 at a mass ratio of 7:3 was 77.8; the score of the tobacco leaves of grade B and grade E1 at a mass ratio of 7:3 was 104. The score of B-grade tobacco leaves mixed with E1-grade tobacco leaves in a mass ratio of 6:4 was 85.2; the score of B-grade tobacco leaves mixed with E1-grade tobacco leaves in a mass ratio of 5:5 was 87; the score of B-grade tobacco leaves mixed with E1-grade tobacco leaves in a mass ratio of 4:6 was 83.4; the score of B-grade tobacco leaves mixed with E1-grade tobacco leaves in a mass ratio of 3:7 was 77.2; the score of B-grade tobacco leaves mixed with E1-grade tobacco leaves in a mass ratio of 2:8 was 70.9.

[0163] Table 3 Chemical indexes and quality evaluation results of B2 and E2 grade tobacco leaves after blending

[0164]

[0165]

[0166] This embodiment inputs the recalculated key chemical indicator data of different grades of tobacco raw materials into a tobacco quality evaluation model based on the chi-square function to obtain the tobacco quality score of the tobacco raw materials after proportioning, that is, the quality score of the tobacco raw materials after proportioning. This can provide support for subsequent comparison results based on the quality score of the tobacco raw materials after proportioning and the quality score of the tobacco raw material samples before proportioning, and multiple executions of part of the fifth to sixth steps to optimize the tobacco recipe list in the redrying module.

[0167] Step 7: Based on the comparison results of the quality scores of the tobacco leaf raw materials after proportioning and the quality scores of the tobacco leaf raw material samples before proportioning, the process of steps 5 to 6 is partially performed multiple times to optimize the tobacco leaf formula list in the redrying module, thereby obtaining an optimized tobacco leaf formula list, as follows:

[0168] (1) Based on the comparison results of the quality scores of tobacco leaf raw materials after proportioning and before proportioning, the tobacco leaf formula in the redrying module is designed, and different grades of new tobacco leaf raw materials are mixed in a specific proportion, and then the key chemical indicators of the different grades of new tobacco leaf raw materials after mixing are calculated; among them, different grades of tobacco leaf raw materials are obtained by grading the tobacco leaf raw materials based on their appearance and color.

[0169] (3) The recalculated key chemical index data of different grades of tobacco raw materials are input into the tobacco quality evaluation model based on the chi-square function to obtain the tobacco quality score of the tobacco raw materials after proportioning, that is, the tobacco raw material quality score after proportioning,

[0170] (4) The above steps (1)-(2) are repeated multiple times to optimize the tobacco leaf formula in the redrying module, thereby obtaining an optimized tobacco leaf formula.

[0171] The technical solution implemented by the present invention first obtains the conventional chemical composition data of new tobacco leaf raw materials of different categories, and based on the conventional component data of new tobacco leaf raw materials of different categories, calculates a variety of derived chemical indicators of new tobacco leaf raw materials of different categories, and then uses information gain and Fisher criterion to select an iterative algorithm to screen out the key chemical indicators of new tobacco leaf raw materials from the conventional chemical components of new tobacco leaf raw materials of different grades and the derived chemical indicators of new tobacco leaf raw materials of different grades. Then, with the key chemical indicators of the screened new tobacco leaf raw materials as a reference, the key chemical indicators of tobacco leaf raw materials of different grades over the years are statistically analyzed, and combined with the distribution law of the key chemical indicators of new tobacco leaf raw materials of different grades, a tobacco leaf raw material distribution model based on the chi-square function is constructed. The leaf quality evaluation model is then used to input the key chemical index data of the obtained new tobacco leaf raw material samples into the tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the new tobacco leaf raw material sample, that is, the quality score of the new tobacco leaf raw material sample before proportioning. Finally, based on the quality score of the new tobacco leaf raw material sample before proportioning, the tobacco leaf formula in the redrying module is designed. In this way, not only can the intrinsic quality of tobacco leaves be optimized and the industrial availability of low-quality tobacco leaves be improved, but also a scientific and reasonable formula design basis can be provided for the tobacco industry to solve the problem that when designing tobacco leaf formulas, the tobacco leaf quality evaluation results are not closely integrated with the tobacco leaf formula design, the intrinsic quality of tobacco leaves cannot be optimized, and the industrial availability of low-quality tobacco leaves cannot be significantly improved.

[0172] Example 2

[0173] In the process of tobacco leaf quality evaluation and formula design, there are still some key issues that need to be solved urgently. For example, among the numerous chemical indicators in tobacco leaves, which specific chemical indicators should be used as the important basis for efficient grading and formula design? So far, there has been no unified and effective industry norms or standards for intrinsic quality evaluation. Although tobacco leaf quality evaluation methods based on chemical indicators have begun to attract attention, the tobacco leaf quality evaluation results have not been closely integrated with tobacco leaf formula design to achieve systematic optimization of tobacco leaf intrinsic quality and achieve optimal tobacco leaf quality.

[0174] Therefore, this application proposes a formula design device based on key chemical indicators to solve the problem that when designing tobacco formulas, the tobacco quality evaluation results are not closely integrated with the tobacco formula design, resulting in the failure to optimize the intrinsic quality of tobacco and significantly improve the industrial availability of low-quality tobacco.

[0175] For details, please refer to Figure 3 , Figure 3This is a schematic structural diagram of a tobacco leaf formulation design device based on key chemical indicators according to an embodiment of the present invention. A formulation device based on key chemical indicators includes a data acquisition and calculation module, a screening and grading module, a quality evaluation module, a tobacco leaf formulation design module, and a tobacco leaf formulation optimization module. The data acquisition and calculation module is used to obtain conventional chemical composition data of different categories of new tobacco leaf raw materials, and based on the conventional composition data of different categories of new tobacco leaf raw materials, calculate multiple derived chemical indicators of different categories of new tobacco leaf raw materials. The conventional composition data includes total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium, and chlorine; the derived chemical indicators include total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, and total nitrogen / nicotine.

[0176] The screening and grading module and the data acquisition and calculation module are used to select the key chemical indicators of new tobacco leaf raw materials from the conventional chemical components of different grades of new tobacco leaf raw materials and the derived chemical indicators of different grades of new tobacco leaf raw materials by using information gain and Fisher criterion and then selecting iterative algorithms; among them, different grades of new tobacco leaf raw materials are classified based on the appearance and color of the new tobacco leaf raw materials;

[0177] The quality evaluation module is connected to the screening and grading module and is used to use the key chemical indicators of the screened new tobacco leaf raw materials as a reference, calculate the key chemical indicators of tobacco leaf raw materials of different grades over the years, and combine the sensory evaluation scores of new tobacco leaf raw materials of different grades to construct a tobacco leaf quality evaluation model based on the chi-square function. The key chemical indicator data of the new tobacco leaf raw material samples are then input into the tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the new tobacco leaf raw material samples, that is, the quality score of the new tobacco leaf raw material samples before proportioning;

[0178] The tobacco recipe design module, connected to the quality evaluation module, is used to design the tobacco recipe in the redrying module based on the quality score of the new tobacco raw material samples before proportioning, mix different grades of tobacco raw materials in specific proportions, and then calculate the key chemical indicators of the mixed tobacco raw materials of different grades. Among them, different grades of tobacco raw materials are classified based on their appearance and color.

[0179] The tobacco leaf formula optimization module is connected to the tobacco leaf formula design module and is used to input the recalculated key chemical index data of different grades of tobacco leaf raw materials into the tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the tobacco leaf raw materials after proportioning, that is, the quality score of the tobacco leaf raw materials after proportioning, and based on the comparison results of the quality score of the tobacco leaf raw materials after proportioning and the quality score of the tobacco leaf raw material samples before proportioning, the processing process from the tobacco leaf formula design module to the tobacco leaf formula optimization module is executed in a loop multiple times to optimize the tobacco leaf formula list in the redrying module, and then obtain the optimized tobacco leaf formula list.

[0180] The technical solution of the embodiment of the present invention first obtains the conventional chemical composition data of different categories of new tobacco leaf raw materials through the data acquisition and calculation module, and calculates a variety of derived chemical indicators of different categories of new tobacco leaf raw materials based on the conventional component data of different categories of new tobacco leaf raw materials. Then, the screening and grading module adopts information gain and Fisher criterion and then selects an iterative algorithm to screen out the key chemical indicators of the new tobacco leaf raw materials from the conventional chemical components of different grades of new tobacco leaf raw materials and the derived chemical indicators of different grades of new tobacco leaf raw materials. Then, the quality evaluation module uses the screened key chemical indicators of the new tobacco leaf raw materials as a reference, and statistics the key chemical indicators of tobacco leaf raw materials of different grades over the years. In combination with the distribution law of the key chemical indicators of new tobacco leaf raw materials of different grades, a tobacco leaf quality evaluation model based on the chi-square function is constructed. The key chemical indicator data of the new tobacco leaf raw material sample is input into the tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco quality score of the new tobacco leaf raw material sample, that is, the quality score of the new tobacco leaf raw material sample before proportioning. Then, the tobacco leaf formula design module uses the key chemical indicators of the new tobacco leaf raw materials before proportioning to evaluate the tobacco leaf quality. Raw material sample quality scoring, designing the tobacco leaf formula in the redrying module, and mixing different grades of tobacco leaf raw materials in a specific proportion, then calculating the key chemical indicators of the mixed tobacco leaf raw materials of different grades, and finally inputting the recalculated key chemical indicator data of different grades of tobacco leaf raw materials into the tobacco leaf quality evaluation model based on the chi-square function through the tobacco leaf formula optimization module to obtain the tobacco leaf quality score of the tobacco leaf raw materials after proportioning, that is, the tobacco leaf raw material quality score after proportioning, and based on the comparison result of the tobacco leaf raw material quality score after proportioning and the tobacco leaf raw material sample quality score before proportioning, the processing process from the tobacco leaf formula design module to the tobacco leaf formula optimization module is executed repeatedly to optimize the tobacco leaf formula list in the redrying module, and then the optimized tobacco leaf formula list is obtained. This not only optimizes the intrinsic quality of tobacco leaves and improves the industrial availability of low-quality tobacco leaves, but also provides the tobacco industry with a scientific and reasonable formula design basis to solve the problem that when designing tobacco leaf formulas, the tobacco leaf quality evaluation results are not closely integrated with the tobacco leaf formula design, the intrinsic quality of tobacco leaves cannot be optimized, and the industrial availability of low-quality tobacco leaves cannot be significantly improved.

[0181] See also Figure 3 To facilitate printing of labels with information about the formulated tobacco leaves, a recipe design device based on key chemical indicators also includes a recipe execution module and a label printing module. The recipe execution module is connected to the tobacco recipe optimization module and is used to find different grades of tobacco leaves based on the optimized tobacco recipe and evenly mix them multiple times in a specific proportion.

[0182] The label printing module is connected to the ratio execution module and is used to print the tobacco leaf formula sheet corresponding to the tobacco leaves after mixing to obtain a tobacco leaf formula information label.

[0183] Although the present invention has been described herein with reference to a number of illustrative embodiments thereof, it will be understood that numerous other modifications and implementations may be devised by those skilled in the art that fall within the scope and spirit of the principles disclosed herein. More specifically, within the scope of the present disclosure, the drawings, and the claims, numerous variations and modifications may be made to the components and / or layout of the subject combination arrangement. In addition to variations and modifications to the components and / or layout, other uses will also be apparent to those skilled in the art.

Claims

1. A formulation design method based on key chemical indicators, characterized by: The following steps are involved: Step 1. Obtain conventional chemical composition data of different types of new tobacco leaf raw materials, and based on the conventional composition data of different types of new tobacco leaf raw materials, calculate multiple derived chemical indicators of different types of new tobacco leaf raw materials; wherein the conventional composition data include total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine; the derived chemical indicators include total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine and total nitrogen / nicotine; Step 2: Using information gain and Fisher criterion and then selecting an iterative algorithm, the key chemical indicators of new tobacco leaf raw materials are screened from the conventional chemical components of different grades of new tobacco leaf raw materials and the derived chemical indicators of different grades of new tobacco leaf raw materials. Among them, different grades of new tobacco leaf raw materials are classified based on the appearance and color of the new tobacco leaf raw materials. Step 3. Using the key chemical indicators of the selected new tobacco leaf raw materials as a reference, the key chemical indicators of tobacco leaf raw materials of different grades over the years were counted, and combined with the sensory evaluation scores of new tobacco leaf raw materials of different grades, a tobacco leaf quality evaluation model based on the chi-square function was constructed; wherein, different grades of new tobacco leaf raw materials are classified based on the different appearance and sensory quality grades of new tobacco leaves; Step 4. Obtain key chemical index data of the new tobacco leaf raw material sample, and input the key chemical index data of the new tobacco leaf raw material sample into the tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the new tobacco leaf raw material sample, that is, the quality score of the new tobacco leaf raw material sample before proportioning; Step 5: Based on the quality scores of the new tobacco leaf raw material samples before blending, the tobacco leaf recipe in the redrying module is designed, and different grades of tobacco leaves are mixed in specific proportions. The key chemical indicators of the mixed tobacco leaf raw materials of different grades are then calculated. The different grades of tobacco leaves are graded based on the appearance and color of the tobacco leaf raw materials. Step 6. Input the recalculated key chemical index data of different grades of tobacco leaf raw materials into the tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the tobacco leaf raw materials after proportioning, that is, the tobacco leaf raw material quality score after proportioning; Step 7. Based on the comparison results of the quality score of the tobacco raw materials after proportioning in Step 6 and the quality score of the tobacco raw material samples before proportioning in Step 4, the process of Step 5 to Step 6 is executed sequentially multiple times to optimize the tobacco leaf formula list in the redrying module, and then obtain the optimized tobacco leaf formula list.

2. The method for formula design based on key chemical indicators according to claim 1, wherein: The step 1 of obtaining conventional chemical composition data of different categories of new tobacco leaf raw materials and calculating a plurality of derived chemical indicators of different categories of new tobacco leaf raw materials based on the conventional composition data of different categories of new tobacco leaf raw materials includes the following steps: Step 1.

1. Collect multiple varieties of freshly-cured tobacco leaves from different tobacco producing areas. Select freshly-cured tobacco leaves from different growing parts among the freshly-cured tobacco leaves of different varieties to obtain freshly-cured tobacco leaves of different categories. Step 1.2, using an online near-infrared conventional chemical composition quantitative model, the conventional chemical component content of different types of new tobacco raw materials is measured to obtain conventional chemical composition data such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine of different types of new tobacco raw materials; Step 1.

3. Based on the conventional chemical composition data of different categories of new tobacco leaf raw materials, such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine, the derived chemical indicators such as total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, total nitrogen / nicotine of different categories of new tobacco leaf raw materials are calculated.

3. The method for formula design based on key chemical indicators according to claim 1, wherein: The step 2, using information gain and Fisher criterion to select an iterative algorithm, to screen out key chemical indicators of new tobacco leaf raw materials from conventional chemical components of different grades of new tobacco leaf raw materials and derived chemical indicators of different grades of new tobacco leaf raw materials, comprises the following steps: Step 2.

1. Based on the appearance and color of the new tobacco leaf raw materials, different types of new tobacco leaf raw materials are graded to obtain new tobacco leaf raw materials of different grades; Step 2.2, determine the conventional chemical composition of different grades of new tobacco leaf raw materials and the derived chemical indicators of different grades of new tobacco leaf raw materials; wherein, conventional component data include total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine; derived chemical indicators include total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine and total nitrogen / nicotine; Step 2.

3. Use information gain and Fisher criterion to select iterative algorithm to screen out key chemical indicators of new tobacco leaf raw materials from the conventional chemical components of different grades of new tobacco leaf raw materials and the derived chemical indicators of different grades of new tobacco leaf raw materials; among them, conventional components include total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine; derived chemical indicators include total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine and total nitrogen / nicotine.

4. The method for formula design based on key chemical indicators according to claim 1, wherein: Step 3, using the key chemical indicators of the screened new tobacco raw materials as a reference, statistically analyzing the key chemical indicators of tobacco raw materials of different grades over the years, and combining the sensory evaluation scores of new tobacco raw materials of different grades to construct a tobacco quality evaluation model based on the chi-square function, includes the following steps: Step 3.

1. Using the selected key chemical indicators of the new tobacco leaf raw materials as a reference, select the key chemical indicators of different categories of tobacco leaf raw materials over the years and perform data statistics to obtain the key chemical indicator data statistics of different categories of tobacco leaf raw materials over the years; wherein the key chemical indicator data statistics of different categories of tobacco leaf raw materials over the years include the data mean and the data standard deviation; Step 3.2: Based on the appearance and color of different categories of tobacco leaves over the years, classify the different categories of tobacco leaf raw materials over the years to obtain tobacco leaf raw materials of different grades over the years, and determine the statistical data of key chemical indicators of tobacco leaf raw materials of different grades over the years; Step 3.3, based on the different appearance and sensory quality grades of different categories of new tobacco leaf raw materials, different categories of new tobacco leaf raw materials are graded to obtain new tobacco leaf raw materials of different grades; Step 3.4, sensory evaluation and smoking scores are conducted on the cut tobacco made from the graded new tobacco leaves of different grades, so as to obtain sensory evaluation scores of the new tobacco leaves of different grades; Step 3.5: Based on the sensory evaluation scores of different grades of new tobacco raw materials obtained in Step 3.3 and the statistical data of key chemical indicators of tobacco raw materials of different grades in previous years obtained in Step 3.2, design an evaluation curve of the key chemical indicator content of tobacco raw materials and the mass fraction of tobacco raw materials; Step 3.6: Use genetic algorithm to find the optimal solution and fit the chi-square function to the evaluation curve of key chemical index content and tobacco leaf raw material mass fraction to obtain the key chemical index evaluation function. The key chemical index evaluation function expression is as follows: Among them, y is the evaluation value of key chemical indicators; M=A×(xC) 2 +B×(xC)+E, where A, B, C, D, and E are the parameters that constitute the evaluation function of the key chemical indicators of tobacco leaves. Adjusting these parameters A, B, C, D, and E can change the chemical indicator evaluation function to evaluate different chemical indicators; Step 3.7, evaluate the function y for each key chemical index respectively i The weight ratio is set, and the evaluation function of each key chemical index after the weight ratio is summed to construct a tobacco leaf quality evaluation model based on the chi-square function. The tobacco leaf quality evaluation function expression is as follows: Among them, Y is the comprehensive quality evaluation score of tobacco leaves; N is the number of chemical indicators involved in the evaluation; w i is the weight of the i-th chemical index evaluation function in tobacco leaf quality evaluation, y i is the evaluation function of the i-th chemical index.

5. The method for formula design based on key chemical indicators according to claim 1, wherein: Step 4, obtaining key chemical index data of the new tobacco leaf raw material sample, and inputting the key chemical index data of the new tobacco leaf raw material sample into a tobacco leaf quality evaluation model based on a chi-square function to obtain a tobacco leaf quality score of the new tobacco leaf raw material sample, that is, a score of the new tobacco leaf raw material sample before proportioning, comprises the following steps: Step 4.

1. Use an online near-infrared conventional chemical composition quantitative model to measure the conventional chemical composition content of the new tobacco leaf raw material sample to obtain the conventional chemical composition data of the new tobacco leaf raw material sample, such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine; Step 4.

2. Based on the conventional chemical composition data of the new tobacco leaf raw material sample, such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium, and chlorine, calculate the derived chemical indicators of the new tobacco leaf raw material sample, such as total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, and total nitrogen / nicotine; Step 4.3, using information gain and Fisher criterion to select iterative algorithm, from the conventional chemical components of the new tobacco leaf raw material samples obtained in Step 4.1, such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine, and the derived chemical indicators of the new tobacco leaf raw material samples obtained in Step 4.2, such as total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, and total nitrogen / nicotine, to screen out the key chemical indicator data of the new tobacco leaf raw material samples; Step 4.4: Input the key chemical index data of the new tobacco leaf raw material sample in Step 4.3 into the tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the new tobacco leaf raw material sample, that is, the quality score of the new tobacco leaf raw material sample before proportioning.

6. The method for formula design based on key chemical indicators according to claim 1, wherein: The step 5, based on the quality score of the new tobacco raw material sample before proportioning, designs the tobacco recipe in the redrying module, mixes different grades of tobacco raw materials in a specific proportion, and then calculates the key chemical indicators of the mixed tobacco raw materials of different grades, includes the following steps: Step 5.

1. Based on the quality scores of the new tobacco leaf raw material samples before blending, design the tobacco leaf formula in the redrying module to obtain a preliminary tobacco leaf formula; Step 5.2: Based on the quality scores of the new tobacco leaf raw material samples before blending, select tobacco leaf raw materials with lower quality or less obvious style characteristics to form different grades of tobacco leaf raw materials; Step 5.3: Based on the quality score of the new tobacco leaf raw material sample before proportioning, the different grades of tobacco leaf raw materials in Step 5.2 are mixed in a specific proportion, and the key chemical index data of the different grades of tobacco leaf raw materials after mixing are calculated to obtain the key chemical index data of the different grades of tobacco leaf raw materials.

7. The method for formula design based on key chemical indicators according to claim 6, characterized in that: Step 5.4, calculating the key chemical index data of the mixed tobacco raw materials of different grades, includes the following steps: Step 5.

41. Use an online near-infrared conventional chemical composition quantitative model to measure the conventional chemical composition content of different grades of tobacco leaf raw materials to obtain conventional chemical composition data such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium, and chlorine of different grades of tobacco leaf raw materials; Step 5.

42. Based on the conventional chemical composition data of different grades of tobacco leaf raw materials, such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium, and chlorine, calculate the derived chemical indicators such as total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, and total nitrogen / nicotine of different grades of tobacco leaf raw materials; Step 5.

43. Use information gain and Fisher criterion to select iterative algorithm again to screen out key chemical indicator data of different grades of tobacco leaf raw materials from the conventional chemical components such as total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine of different grades of tobacco leaf raw materials obtained in Step 5.41 and the derived chemical indicators such as total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine, total nitrogen / nicotine of different grades of tobacco leaf raw materials obtained in Step 5.

42.

8. The method for formula design based on key chemical indicators according to claim 1, wherein: The step 7, based on the comparison result of the quality score of the tobacco raw material after proportioning in step 6 and the quality score of the tobacco raw material sample before proportioning in step 4, performs part of the process from step 5 to step 6 multiple times to optimize the tobacco leaf formula list in the redrying module, thereby obtaining the optimized tobacco leaf formula list, including the following steps: Step 7.

1. Based on the comparison of the quality scores of the tobacco raw materials after blending in Step 6 and before blending in Step 4, design the tobacco recipe for the redrying module, mix different grades of new tobacco raw materials in specific proportions, and then calculate the key chemical indicators of the different grades of mixed new tobacco raw materials; the different grades of tobacco raw materials are classified based on their appearance and color; Step 7.2, inputting the recalculated key chemical index data of different grades of tobacco leaf raw materials into a tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the tobacco leaf raw materials after blending, i.e., the tobacco leaf raw material quality score after blending; Step 7.3: Repeat Step 7.1-Step 7.2 multiple times to optimize the tobacco leaf formula in the redrying module, and then obtain the optimized tobacco leaf formula.

9. A formulation design device based on key chemical indicators according to any one of claims 1 to 8, comprising: The data acquisition and calculation module is used to obtain conventional chemical composition data of different types of new tobacco leaf raw materials, and based on the conventional composition data of different types of new tobacco leaf raw materials, calculate a variety of derived chemical indicators of different types of new tobacco leaf raw materials; wherein, conventional composition data include total sugar, reducing sugar, total nitrogen, nicotine, moisture, potassium and chlorine; derived chemical indicators include total sugar / total nitrogen, reducing sugar / total nitrogen, total sugar / nicotine, reducing sugar / nicotine and total nitrogen / nicotine; The screening and grading module and the data acquisition and calculation module are used to select the key chemical indicators of new tobacco leaf raw materials from the conventional chemical components of different grades of new tobacco leaf raw materials and the derived chemical indicators of different grades of new tobacco leaf raw materials by using information gain and Fisher criterion and then selecting iterative algorithms; among them, different grades of new tobacco leaf raw materials are classified based on the appearance and color of the new tobacco leaf raw materials; The quality evaluation module is connected to the screening and grading module and is used to use the key chemical indicators of the screened new tobacco raw materials as a reference, calculate the key chemical indicators of tobacco raw materials of different grades over the years, and combine the sensory evaluation scores of new tobacco raw materials of different grades to construct a tobacco quality evaluation model based on the chi-square function. The key chemical indicator data of the new tobacco raw material samples are then input into the tobacco quality evaluation model based on the chi-square function to obtain the tobacco quality score of the new tobacco raw material samples, that is, the quality score of the new tobacco raw material samples before proportioning; The tobacco recipe design module, connected to the quality evaluation module, is used to design the tobacco recipe in the redrying module based on the quality score of the new tobacco raw material samples before proportioning, mix different grades of tobacco raw materials in specific proportions, and then calculate the key chemical indicators of the mixed tobacco raw materials of different grades. Among them, different grades of tobacco raw materials are classified based on their appearance and color. The tobacco leaf formula optimization module is connected to the tobacco leaf formula design module and is used to input the recalculated key chemical index data of different grades of tobacco leaf raw materials into the tobacco leaf quality evaluation model based on the chi-square function to obtain the tobacco leaf quality score of the tobacco leaf raw materials after proportioning, that is, the quality score of the tobacco leaf raw materials after proportioning, and based on the comparison results of the quality score of the tobacco leaf raw materials after proportioning and the quality score of the tobacco leaf raw material samples before proportioning, the processing process from the tobacco leaf formula design module to the tobacco leaf formula optimization module is executed in a loop multiple times to optimize the tobacco leaf formula list in the redrying module, and then obtain the optimized tobacco leaf formula list.

10. The formulation design device based on key chemical indicators according to claim 9, further comprising: The proportion execution module is connected to the tobacco leaf formula optimization module and is used to find different grades of tobacco leaves based on the optimized tobacco leaf formula and mix them evenly multiple times in a specific proportion; as well as The label printing module is connected to the ratio execution module and is used to print the tobacco leaf formula sheet corresponding to the tobacco leaves after mixing to obtain a tobacco leaf formula information label.

Images