CN-121970914-A - Baking process based on quality of Xiangyan No. 7 fresh tobacco leaves

CN121970914ACN 121970914 ACN121970914 ACN 121970914ACN-121970914-A

Abstract

The invention relates to the technical field of tobacco baking, and provides a baking process based on quality of Xiang tobacco No. 7 fresh tobacco, which comprises the following steps of S1, performing near infrared spectrum scanning on picked fresh tobacco to obtain near infrared spectrum data of each tobacco; the method comprises the steps of S2, dividing near infrared spectrum data into a training set and a testing set, S3, taking the near infrared spectrum data of tobacco leaves in the training set as input vectors, inputting the near infrared spectrum data into a self-organizing neural network model for clustering, dividing the tobacco leaves into 3 subsets, S4, respectively optimizing baking process parameters according to a three-section baking process aiming at the 3 subsets, wherein the three-section baking process comprises a yellowing stage, a fixed color stage and a dry rib stage, S5, carrying out near infrared spectrum scanning on newly collected tobacco leaves, classifying the newly collected tobacco leaves through the self-organizing neural network model, determining the subset to which the newly collected tobacco leaves belong, and sequentially executing baking processes of the yellowing stage, the fixed color stage and the dry rib stage according to baking process curves corresponding to the subset.

Inventors

QIU ZHENWU
WANG WEI
SUN JINGZHAO
ZHANG ZHIHUA
LI ZHAO
WANG KE
WEN LIANG
YI LIJUN
LI RUOSHI

Assignees

湖南省烟草公司常德市公司

Dates

Publication Date: 20260505
Application Date: 20251229

Claims (10)

1. A baking process based on the quality of Xiangyan No. 7 fresh tobacco leaves is characterized by comprising the following steps of: S1, carrying out near infrared spectrum scanning on the collected fresh tobacco leaves to obtain near infrared spectrum data of each tobacco leaf; S2, dividing the near infrared spectrum data into a training set and a testing set; s3, taking near infrared spectrum data of each tobacco leaf of the training set as an input vector, inputting the input vector into a self-organizing neural network model for clustering, and dividing the tobacco leaves into 3 subsets; S4, respectively optimizing baking process parameters according to a three-section baking process aiming at the divided 3 subsets, wherein the three-section baking process comprises a yellowing period, a fixed color period and a dry rib period; The optimization parameters comprise temperature, humidity, ventilation quantity and temperature stabilizing time of each stage, the quality of the cured tobacco leaves is taken as a target, and a curing process curve of a corresponding subset is established; s5, carrying out near infrared spectrum scanning on the newly collected tobacco leaves, classifying by the self-organizing neural network model, determining a subset to which the newly collected tobacco leaves belong, and sequentially executing baking processes of a yellowing period, a fixed color period and a dry rib period according to baking process curves corresponding to the subset.
2. The flue-curing process based on the quality of fresh tobacco leaves No. 7 of Hunan tobacco as claimed in claim 1, wherein the flue-curing process further comprises the step of adopting a wave band of 900-1100 nm for near infrared spectrum scanning in the step S1.
3. The flue-curing process based on the quality of fresh tobacco leaves No. 7 of Hunan tobacco as claimed in claim 2, further comprising dividing near infrared spectrum data of each tobacco leaf into a training set and a test set according to a ratio of 7:3 in step S2.
4. The flue-curing process based on fresh tobacco quality No. 7 of claim 3 further comprising dividing tobacco into 3 subsets with a number of competing layer neurons of the self-organizing neural network model of 3 in step S3.
5. The flue-curing process based on fresh tobacco quality No. 7 of Xiangyan according to claim 4, wherein the working process of the self-organizing neural network model comprises the following steps: s31, initializing, namely setting a weight initial value between an input layer and a competition layer by using a random number; S32, performing Euclidean distance calculation between each neuron of the input layer and each neuron of the competition layer, and calculating the distance between the j-th neuron of the competition layer and the input vector xi as follows: ; Wherein x i is an input vector, W j is a weight vector of a j-th neuron of the competitive layer, and the neuron with the smallest distance is obtained through calculation to be taken as a winning neuron and is marked as j; And S33, correcting the weight of the winning neuron j and adjacent neurons by the following formula: ; Wherein W j (t+1) is the corrected weight, W j (t) is the current weight, alpha (t) is the learning rate, Is a neighborhood function; S34, repeating the steps S32-S33 until the preset iteration times or the weight change is smaller than a preset threshold value.
6. The process for curing quality of fresh tobacco leaves No. 7 based on xiang tobacco of claim 5, further comprising: in step S34, the preset number of iterations is 500.
7. The process for curing the quality of fresh tobacco leaves No. 7 based on xiang tobacco according to claim 6, further comprising the steps of: the yellowing stage, namely gradually raising the temperature from the initial temperature to the yellowing termination temperature, and controlling the humidity and ventilation quantity to degrade chlorophyll of tobacco leaves, yellow the tobacco leaves and hydrolyze starch into reducing sugar; The color fixing period, namely, continuously increasing the temperature to the color fixing temperature, reducing the humidity, increasing the ventilation quantity, fixing the color of the tobacco leaves, finishing the conversion of substances and forming aroma substances; in the tendon drying period, the temperature is increased to the tendon drying temperature, the humidity is further reduced, and the maximum ventilation quantity is achieved, so that the main veins of the tobacco leaves are dried, and the overall moisture content is reduced to 12-14%.
8. The flue-curing process based on fresh tobacco quality No. 7 of Hunan tobacco of claim 7, further comprising establishing a chemical composition prediction model through the near infrared spectrum data, predicting chemical composition content of tobacco, wherein the chemical composition at least comprises reducing sugar, total sugar, nicotine, total nitrogen and potassium; Wherein, the baking difficulty index HDI is calculated according to the content of the chemical components: Wherein, the baking difficulty index HDI is calculated according to the content of the chemical components: hdi= (reducing sugar ×α+potassium ×β -nicotine ×γ)/(total nitrogen ×δ); Wherein, the value range of alpha is 0.40-0.50, the value range of beta is 0.30-0.40, the value range of gamma is 0.20-0.30, and the value range of delta is 0.25-0.35; wherein 3 of the training sets are distinguished according to the HDI values: The first subset is HDI > 8.5, which is the tobacco leaves easy to bake; the second subset is 6.0-8.5 HDI, which is the proper baking tobacco leaf; The third subset is HDI < 6.0, which is the refractory tobacco leaf.
9. The flue-curing process based on fresh tobacco quality No. 7 of xiang tobacco of claim 8, wherein optimizing the flue-curing process for different subsets further comprises a segmented temperature and humidity control strategy based on biochemical reaction kinetics: the yellowing stage is divided into four stages of chlorophyll degradation stage, starch hydrolysis stage, protein hydrolysis stage and polyphenol oxidation stage, and the temperature control interval of each stage corresponds to the optimal activity temperature interval of chlorophyllase, amylase, protease and polyphenol oxidase respectively: the chlorophyll degradation stage is carried out at a temperature of 35-40 ℃ and a residence time of 30-35% of the total duration of the yellowing stage; The starch hydrolysis stage, wherein the temperature is 40-45 ℃, and the residence time accounts for 35-40% of the total duration of the yellowing stage; the protein hydrolysis stage, wherein the temperature is 42-48 ℃, and the residence time accounts for 15-20% of the total duration of the yellowing stage; The polyphenol oxidation stage is carried out at the temperature of 45-50 ℃ and the residence time accounts for 10-15% of the total duration of the yellowing stage.
10. The process for curing the quality of fresh tobacco leaves No. 7 based on xiang tobacco of claim 9, further comprising the steps of monitoring and dynamically adjusting in real time during the curing of step S5: Monitoring yellowing degree, water content and/or volatile substance concentration of tobacco leaves in real time through a sensor; calculating the actual yellowing speed And expected yellowing speed Deviation of, and/or actual dewatering rate And the expected dehydration rate Deviation of (2); when the deviation exceeds a preset threshold, dynamically adjusting the temperature, the humidity and/or the ventilation according to a preset adjustment rule, wherein the adjustment rule comprises: at a practical yellowing speed When the expected value is 1.5 multiplied by the value, the temperature is reduced by 1-2 ℃, the humidity is increased by 1 ℃, and the ventilation quantity is reduced by 10-20%; at a practical yellowing speed When the temperature is less than 0.7 multiplied by the expected value, the temperature is increased by 1-2 ℃, the humidity is reduced by 0.5 ℃, and the ventilation quantity is increased by 5-10%; When the actual dehydration rate is When the temperature is 3%/h, the temperature is reduced by 2 ℃, the humidity is increased by 2 ℃, and the ventilation quantity is reduced by 20-30%; When the actual dehydration rate is And when the temperature is less than 0.5%/h, the temperature is increased by 2 ℃, the humidity is reduced by 1 ℃, and the ventilation quantity is increased by 15-20%.

Description

Baking process based on quality of Xiangyan No. 7 fresh tobacco leaves Technical Field The invention relates to the technical field of tobacco baking, in particular to a baking process based on fresh tobacco quality of Xiang tobacco No. 7. Background In the field of tobacco leaf baking, the design of the baking process is directly related to the baking efficiency and the tobacco leaf quality. The baking process generally comprises the stages of yellowing, color fixing, drying, and the like, and the moisture evaporation and chemical component conversion of the tobacco leaves are controlled through temperature, humidity and ventilation parameters. However, since the fresh tobacco leaves of different tobacco leaf varieties, particularly high-end varieties (such as Xiang tobacco No. 7), have remarkable differences in chemical components, moisture content and physicochemical properties, the existing baking technology cannot be accurately regulated and controlled according to the quality of the tobacco leaves, so that the problems of inconsistent baking effect, high low-quality tobacco proportion, difficulty in distinguishing flavor and the like are caused. The tobacco leaf baking method disclosed in the Chinese patent CN102499424A comprises the steps of setting a temperature control scheme of a plurality of stages according to the dry bulb temperature and the wet bulb temperature after the tobacco leaf is placed in a baking room, so that the tobacco leaf undergoes different temperature and humidity conditions in the stages of yellowing, color fixing, drying, reinforcement drying and the like. According to the method, the baking process is controlled through the temperature and humidity gradient, so that the automation and the productivity of tobacco baking are improved, the unified fixed temperature and humidity parameters for all tobacco leaves are still adopted, classification and dynamic adjustment are not carried out according to the chemical composition difference of the tobacco leaves, and the differentiated optimal control of the tobacco leaves with different qualities is difficult to realize. Another typical chinese patent CN101579144B discloses a tobacco curing method and curing barn, in which the tobacco is cured in different temperature and humidity stages in the curing barn by adjusting the air flow direction and the flow of hot air. The scheme is improved to a certain extent in the process flow, but classification treatment and parameter optimization cannot be carried out according to the physicochemical differences of different varieties of tobacco leaves, and the baking quality and the uniformity of the high-quality tobacco leaves cannot be improved obviously. In addition, the prior art has the defects that 1, most of the existing baking methods adopt unified temperature and humidity parameter control, the obvious influence of tobacco variety and chemical composition difference on the baking process is ignored, 2, the traditional process lacks an intelligent classification and dynamic parameter adjustment mechanism based on the self quality of tobacco, so that the material conversion of tobacco with different qualities in the baking process is insufficient, 3, the prior art focuses on the staged temperature control, a control system of tobacco composition, classification judgment and dynamic optimization of technological parameters is not formed, the high requirements of high-end tobacco on flavor, aroma and texture are difficult to meet, and 4, in the existing baking process, the improvement of high-end tobacco and the reduction of low-end tobacco are difficult to be simultaneously considered, and the quality stability of the whole product cannot be effectively improved. The invention is made to solve the problems of fixed baking parameters, incapability of classification control, insufficient material conversion, uneven flavor performance and the like commonly existing in the field. Disclosure of Invention The invention aims to provide a baking process based on the quality of Xiang tobacco No. 7 fresh tobacco leaves, aiming at the defects existing at present. In order to overcome the defects in the prior art, the invention adopts the following technical scheme: a baking process based on the quality of Xiangyan No. 7 fresh tobacco leaves comprises the following steps: S1, carrying out near infrared spectrum scanning on the collected fresh tobacco leaves to obtain near infrared spectrum data of each tobacco leaf; S2, dividing the near infrared spectrum data into a training set and a testing set; s3, taking near infrared spectrum data of each tobacco leaf of the training set as an input vector, inputting the input vector into a self-organizing neural network model for clustering, and dividing the tobacco leaves into 3 subsets; S4, respectively optimizing baking process parameters according to a three-section baking process aiming at the divided 3 subsets, wherein the three-section baking process co