CN108142976B - Cut tobacco drying process parameter optimization method - Google Patents
Cut tobacco drying process parameter optimization method Download PDFInfo
- Publication number
- CN108142976B CN108142976B CN201711228305.8A CN201711228305A CN108142976B CN 108142976 B CN108142976 B CN 108142976B CN 201711228305 A CN201711228305 A CN 201711228305A CN 108142976 B CN108142976 B CN 108142976B
- Authority
- CN
- China
- Prior art keywords
- cut tobacco
- drying process
- data
- process parameter
- water content
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A24—TOBACCO; CIGARS; CIGARETTES; SIMULATED SMOKING DEVICES; SMOKERS' REQUISITES
- A24B—MANUFACTURE OR PREPARATION OF TOBACCO FOR SMOKING OR CHEWING; TOBACCO; SNUFF
- A24B3/00—Preparing tobacco in the factory
- A24B3/10—Roasting or cooling tobacco
Landscapes
- Manufacture Of Tobacco Products (AREA)
Abstract
The invention discloses a method for optimizing leaf shred drying process parameters, which provides a method for optimizing and setting leaf shred drying process parameters, solves the problems that a leaf shred drying process function model cannot be established and the process parameters are difficult to optimize and set in the traditional method, improves the operation efficiency and the prediction precision of a prediction model, establishes a mapping relation between each process parameter and leaf shred water content in a leaf shred drying process by establishing a lightweight data-driven prediction model, seeks the combination of the optimal value of the leaf shred water content and the corresponding optimal leaf shred drying process parameter according to the mapping relation, and realizes the accurate optimization of the leaf shred drying process parameters and the leaf shred water content even when the leaf shred drying process parameters are more.
Description
Technical Field
The invention relates to a method for optimizing leaf shred drying process parameters, and belongs to the field of agricultural and sideline product drying.
Background
The leaf shred drying is an important process in the drying process of agricultural and sideline products, and the water content of the leaf shreds is kept stable by optimally setting the process parameters in the drying process so as to improve and control the quality of the leaf shreds. At present, the optimized setting of the cut tobacco drying process parameters mainly depends on the experience of technicians, and the optimized setting method is difficult to adopt, mainly because the cut tobacco drying process is a complex process comprising multi-field multidisciplinary coupling of physics, chemistry and the like, the relationship between each process parameter and the cut tobacco moisture content is very complex, and the functional relationship is difficult to determine by the traditional method.
Disclosure of Invention
The invention provides a cut tobacco drying process parameter optimization method, aiming at solving the problems that a cut tobacco drying process function model cannot be established, the process parameters are difficult to optimize and set and the like in the traditional method.
The technical scheme of the invention is as follows: a cut tobacco drying process parameter optimization method comprises the following steps:
step 1, eliminating abnormal data and error data in cut tobacco drying process parameter data to obtain process parameter data to be optimized;
step 2, performing dimensionality reduction on the process parameter data to be optimized to obtain lightweight parameter data;
step 3, establishing an initial BP neural network, substituting lightweight parameter data and cut tobacco moisture content training data into the initial BP neural network for training to obtain a lightweight data-driven prediction model;
step 4, randomly screening the process parameter data to be optimized to obtain a cut tobacco drying process parameter population;
step 5, performing dimensionality reduction operation on individuals in the cut tobacco drying process parameter population;
step 6, substituting the population individual data subjected to dimensionality reduction into a lightweight data driving prediction model to obtain a predicted value of the water content of the cut tobaccoy i ;
Step 7, convergence judgment: predicting the water content of the cut tobaccoy i With set value of water content of cut tobaccoy 0Making a difference if the difference is less than or equal to the convergence accuracyeAnd outputting the individuals in the cut tobacco drying process parameter population corresponding to the predicted value of the cut tobacco water content, otherwise, not outputting.
And the dimensionality reduction treatment adopts a principal component analysis method.
In the step 7, if the cut tobacco is not output, updating the cut tobacco drying process parameter population, and repeating the steps 5 to 7 until the cut tobacco drying process parameter population is output|y i -y 0 |≤e。
The invention has the beneficial effects that: the method is provided for optimizing the setting of the cut tobacco drying process parameters, the problems that a cut tobacco drying process function model cannot be established and the process parameters are difficult to optimize in the traditional method are solved, the operation efficiency and the prediction accuracy of the prediction model are improved, the mapping relation between each process parameter and the cut tobacco moisture content in the cut tobacco drying process is established by establishing the lightweight data-driven prediction model, the combination of the optimal value of the cut tobacco moisture content and the corresponding optimal cut tobacco drying process parameter is sought according to the mapping relation, and the cut tobacco drying process parameters and the cut tobacco moisture content can be accurately optimized even when the cut tobacco drying process parameters are more.
Drawings
FIG. 1 is a flow chart of the present invention.
Detailed Description
Example 1: as shown in fig. 1, a method for optimizing the drying process parameters of cut tobacco comprises the following steps:
step 1, eliminating abnormal data and error data in cut tobacco drying process parameter data to obtain process parameter data to be optimized;
step 2, performing dimensionality reduction on the process parameter data to be optimized to obtain lightweight parameter data;
step 3, establishing an initial BP neural network, substituting lightweight parameter data and cut tobacco moisture content training data into the initial BP neural network for training to obtain a lightweight data-driven prediction model;
step 4, randomly screening the process parameter data to be optimized to obtain a cut tobacco drying process parameter population;
step 5, performing dimensionality reduction operation on individuals in the cut tobacco drying process parameter population;
step 6, substituting the population individual data subjected to dimensionality reduction into a lightweight data driving prediction model to obtain a predicted value of the water content of the cut tobaccoy i ;
Step 7, convergence judgment: predicting the water content of the cut tobaccoy i With set value of water content of cut tobaccoy 0Making a difference if the difference is less than or equal to the convergence accuracyeAnd outputting the individuals in the cut tobacco drying process parameter population corresponding to the predicted value of the cut tobacco water content, otherwise, not outputting.
Further, the dimensionality reduction processing can be set to adopt a principal component analysis method.
Example 2: as shown in fig. 1, a method for optimizing the drying process parameters of cut tobacco comprises the following steps:
step 1, eliminating abnormal data and error data in cut tobacco drying process parameter data to obtain process parameter data to be optimized;
step 2, performing dimensionality reduction on the process parameter data to be optimized to obtain lightweight parameter data;
step 3, establishing an initial BP neural network, substituting lightweight parameter data and cut tobacco moisture content training data into the initial BP neural network for training to obtain a lightweight data-driven prediction model;
step 4, randomly screening the process parameter data to be optimized to obtain a cut tobacco drying process parameter population;
step 5, performing dimensionality reduction operation on individuals in the cut tobacco drying process parameter population;
step 6, substituting the population individual data subjected to dimensionality reduction into a lightweight data driving prediction model to obtain a predicted value of the water content of the cut tobaccoy i ;
Step 7, convergence judgment: predicting the water content of the cut tobaccoy i With set value of water content of cut tobaccoy 0Making a difference if the difference is less than or equal to the convergence accuracyeOutputting individuals in the cut tobacco drying process parameter population corresponding to the predicted value of the cut tobacco water content, otherwise updating the cut tobacco drying process parameter population, and repeating the steps 5 to 7 until the step is finished|y i -y 0 |≤e。
Further, the dimensionality reduction processing can be set to adopt a principal component analysis method.
Example 3: as shown in fig. 1, a method for optimizing the drying process parameters of cut tobacco comprises the following steps:
step 1, removing abnormal data and error data in the cut tobacco drying process parameter data to obtain process parameter data to be optimized, as shown in table 1:
table 1:
variable x in Table 14The data of (2) is no longer changed from the previous data starting from line 8, variable x8All the data of (1) are 0 from the 9 th row, so the data after the 8 th row are eliminated; variable x9All the data of (1) are 0, so the variable x is eliminated9Obtaining the data of the process parameters to be optimized shown in the table 2:
table 2:
step 2, performing dimensionality reduction on the process parameter data to be optimized to obtain lightweight parameter data, such as: establishing the following dimensionality reduction formula by using a principal component analysis method:
principal component 1= 0.050027 × x1+0.020110*x2-0.257952*x3-0.035438*x4+0.386489*x5
+0.443235*x6-0.000825*x7+0.144985*x8+0.014484*x10-0.004486*x11-0.027720*x12
-0.051757*x13+0.080297*x14+0.013425*x15-0.018750*x16-0.016826*x17
Principal component 2= -0.055614 × x1-0.030874*x2+0.506334*x3-0.012715*x4-0.103676*x5
-0.210429*x6-0.052316*x7+0.222817*x8+0.015285*x10-0.015888*x11+0.029754*x12
+0.049319*x13+0.310010*x14+0.283559*x15-0.037438*x16-0.005331*x17
Principal component 3= -0.186842 × x1+0.033591*x2-0.069666*x3-0.150656*x4-0.002547*x5
+0.013832*x6+0.018056*x7-0.102104*x8+0.391237*x10-0.316293*x11-0.102645*x12
+0.029974*x13+0.073417*x14+0.137466*x15+0.094621*x16+0.404656*x17
Dividing the data (x is divided) of the process parameters x 1-x 17 to be optimized, which are obtained in the step 19) Respectively substituting the dimensionality reduction formulas of the main components 1-3 to obtain the lightweight parameter data and the cut tobacco water content shown in the table 3Training data:
table 3:
step 3, establishing an initial BP neural network, substituting lightweight parameter data and cut tobacco moisture content training data into the initial BP neural network for training to obtain a lightweight data-driven prediction model, wherein the lightweight data-driven prediction model comprises the following steps: taking the principal components 1-3 as input variables, taking y as an output variable, substituting the y into a neural network for training, wherein the constructed mapping relation between the principal components 1-3 and y is a lightweight data-driven prediction model;
and 4, randomly screening the process parameter data to be optimized to obtain a cut tobacco drying process parameter population, wherein the process parameter population comprises the following steps: obtaining data (x is divided) of the process parameters x 1-x 17 to be optimized from the step 19) Respectively randomly screening one data to obtain one individual of the population, repeating the screening process for 5 times to obtain 5 individuals, wherein the 5 individuals form a cut tobacco drying process parameter population, and the process parameter population is shown in table 4:
table 4:
and 5, performing dimensionality reduction treatment on individuals in the cut tobacco drying process parameter population, such as: substituting 5 individuals of the cut leaf drying process parameter population in the step 4 into the dimensionality reduction formula in the step 2 respectively, wherein the result is shown in a table 5:
table 5:
step 6, substituting the population individual data subjected to dimensionality reduction into a lightweight data driving prediction model to obtain a predicted value of the water content of the cut tobaccoy i(ii) a Such as: each time a group of data of main components 1-3 is input, a predicted value of the water content of the cut tobacco can be obtainedy iAs shown in table 6:
table 6:
and 7, judging convergence. Specifically, the predicted value of the water content of the cut tobacco leavesy iWith set value of water content of cut tobaccoy 0Making a difference if the difference is less than or equal to the convergence accuracyeObtaining the optimal value of the water content of the cut tobacco, outputting the combination of the optimal technological parameters of cut tobacco drying corresponding to the optimal value, and ending the step; if the difference is greater than the convergence accuracyeUpdating the cut tobacco drying process parameter population, and repeating the steps 5 to 7 until the cut tobacco drying process parameter population is updated|y i-y 0 |≤e. Such as: convergence accuracy measuree=0.001, the set value of the water content of the cut tobacco leaves is takeny 0= 12.7, wheny iWhen the power is not less than = 12.699915,|y i -y 0 |if =0.000585 ≦ 0.001, theny i= 12.699915 is the optimal value of water content of cut tobacco, and the corresponding data of cut tobacco drying process parameters x 1-x 17 before individual 3 dimensionality reduction (except x)9) The combination of the optimal process parameters is obtained, and the step is finished; when in use|y i -y 0 |And (3) when the water content of the cut tobacco is more than or equal to 0.001, if the combination of the optimal value of the water content of the cut tobacco and the optimal process parameters is not obtained, selecting, crossing and mutating the individuals 1-5 by utilizing a genetic algorithm, generating new individuals and populations, and repeatedly performing the steps 5-7 by using the new individuals and the new populations until the convergence condition is met.
While the present invention has been described in detail with reference to the embodiments shown in the drawings, the present invention is not limited to the embodiments, and various changes can be made without departing from the spirit of the present invention within the knowledge of those skilled in the art.
Claims (2)
1. A cut tobacco drying process parameter optimization method is characterized by comprising the following steps: the method comprises the following steps:
step 1, eliminating abnormal data and error data in cut tobacco drying process parameter data to obtain process parameter data to be optimized;
step 2, performing dimensionality reduction on the process parameter data to be optimized to obtain lightweight parameter data;
step 3, establishing an initial BP neural network, substituting lightweight parameter data and cut tobacco moisture content training data into the initial BP neural network for training to obtain a lightweight data-driven prediction model;
step 4, randomly screening the process parameter data to be optimized to obtain a cut tobacco drying process parameter population;
step 5, performing dimensionality reduction operation on individuals in the cut tobacco drying process parameter population;
step 6, substituting the population individual data subjected to dimensionality reduction into a lightweight data driving prediction model to obtain a predicted value y of the water content of the cut tobaccoi;
Step 7, convergence judgment: predicting the water content of the cut tobaccoiAnd the set value y of the water content of the cut tobacco0Making a difference, if the difference value is less than or equal to the convergence precision e, outputting individuals in the cut tobacco drying process parameter population corresponding to the predicted value of the cut tobacco moisture content, otherwise, not outputting the individuals;
in the step 7, if the cut tobacco is not output, updating the cut tobacco drying process parameter population, and repeating the steps 5 to 7 until the value of y is up toi-y0|≤e。
2. The method for optimizing the parameters of the leaf shred drying process according to claim 1, wherein the method comprises the following steps: and the dimensionality reduction treatment adopts a principal component analysis method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711228305.8A CN108142976B (en) | 2017-11-29 | 2017-11-29 | Cut tobacco drying process parameter optimization method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711228305.8A CN108142976B (en) | 2017-11-29 | 2017-11-29 | Cut tobacco drying process parameter optimization method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108142976A CN108142976A (en) | 2018-06-12 |
CN108142976B true CN108142976B (en) | 2020-05-15 |
Family
ID=62469221
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711228305.8A Active CN108142976B (en) | 2017-11-29 | 2017-11-29 | Cut tobacco drying process parameter optimization method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108142976B (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109506450B (en) * | 2018-10-24 | 2020-06-02 | 浙江工业大学 | Neural network prediction control method for humidity in automatic drying process of traditional Chinese medicine decoction pieces |
CN110973686B (en) * | 2019-12-13 | 2022-02-08 | 红云红河烟草(集团)有限责任公司 | Method for establishing accurate moisture control model in silk making process |
CN112380760B (en) * | 2020-10-13 | 2023-01-31 | 重庆大学 | Multi-algorithm fusion based multi-target process parameter intelligent optimization method |
CN112257948A (en) * | 2020-10-30 | 2021-01-22 | 红云红河烟草(集团)有限责任公司 | Method, device and equipment for predicting moisture content of moistening leaf feeding outlet |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US3983884A (en) * | 1974-05-04 | 1976-10-05 | Eduard Gerlach Gmbh | Method for manufacturing tobacco foil |
CN1525394A (en) * | 2003-02-25 | 2004-09-01 | 颐中烟草(集团)有限公司 | Neural net prediction method for cigarette sensory evaluating smoking and fume indication |
CN1996175A (en) * | 2006-12-30 | 2007-07-11 | 辽宁省粮食科学研究所 | Corn drying prediction control system and method based on fuzzy-neural network |
EP2100524A1 (en) * | 2006-12-11 | 2009-09-16 | Japan Tobacco Inc. | Low fire spreading cigarette, wrapping paper for the cigarette, and method of producing wrapping paper |
CN106444379A (en) * | 2016-10-10 | 2017-02-22 | 重庆科技学院 | Intelligent drying remote control method and system based on internet of things recommendation |
CN106617288A (en) * | 2016-12-28 | 2017-05-10 | 广东省金叶科技开发有限公司 | Paper cigarette filter stick and manufacturing method thereof |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2003153677A (en) * | 2001-11-20 | 2003-05-27 | Japan Tobacco Inc | Manufacturing equipment for rod shaped article |
TWI267012B (en) * | 2004-06-03 | 2006-11-21 | Univ Nat Cheng Kung | Quality prognostics system and method for manufacturing processes |
-
2017
- 2017-11-29 CN CN201711228305.8A patent/CN108142976B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US3983884A (en) * | 1974-05-04 | 1976-10-05 | Eduard Gerlach Gmbh | Method for manufacturing tobacco foil |
CN1525394A (en) * | 2003-02-25 | 2004-09-01 | 颐中烟草(集团)有限公司 | Neural net prediction method for cigarette sensory evaluating smoking and fume indication |
EP2100524A1 (en) * | 2006-12-11 | 2009-09-16 | Japan Tobacco Inc. | Low fire spreading cigarette, wrapping paper for the cigarette, and method of producing wrapping paper |
CN1996175A (en) * | 2006-12-30 | 2007-07-11 | 辽宁省粮食科学研究所 | Corn drying prediction control system and method based on fuzzy-neural network |
CN106444379A (en) * | 2016-10-10 | 2017-02-22 | 重庆科技学院 | Intelligent drying remote control method and system based on internet of things recommendation |
CN106617288A (en) * | 2016-12-28 | 2017-05-10 | 广东省金叶科技开发有限公司 | Paper cigarette filter stick and manufacturing method thereof |
Non-Patent Citations (1)
Title |
---|
基于多因素分析的烘丝机入口含水率预测模型的建立与应用;钟文焱等;《烟草科技》;20150515;第48卷(第5期);第67-73页 * |
Also Published As
Publication number | Publication date |
---|---|
CN108142976A (en) | 2018-06-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108142976B (en) | Cut tobacco drying process parameter optimization method | |
CN111045326B (en) | Tobacco shred drying process moisture prediction control method and system based on recurrent neural network | |
CN108959728B (en) | Radio frequency device parameter optimization method based on deep learning | |
CN109978287B (en) | Intelligent polycrystalline silicon production method and system | |
Meng et al. | Fast genetic algorithms used for PID parameter optimization | |
CN106779068A (en) | The method and apparatus for adjusting artificial neural network | |
CN109858714B (en) | Tobacco shred quality inspection index prediction method, device and system based on improved neural network | |
CN102626557B (en) | Molecular distillation process parameter optimizing method based on GA-BP (Genetic Algorithm-Back Propagation) algorithm | |
CN107728477A (en) | A kind of industrialized aquiculture water quality dissolved oxygen prediction control method and system | |
US10918014B2 (en) | Fertilization precision control method for water and fertilizer integrated equipment and control system thereof | |
CN110347192B (en) | Glass furnace temperature intelligent prediction control method based on attention mechanism and self-encoder | |
WO2017088208A1 (en) | Data-difference-driven self-learning dynamic optimization method for batch process | |
CN104951987B (en) | Crop Breeding evaluation method based on decision tree | |
CN115529987B (en) | Air port regulating and controlling method, device, equipment and storage medium for crop facility | |
CN112273695A (en) | Method, device and equipment for predicting water content of loose moisture regain outlet | |
TWI617422B (en) | Plastic extrusion process control method and parameters adjustment system | |
CN107643684B (en) | Valve flow function optimization method and device | |
MX2022011501A (en) | System, method, and computer program product for optimizing a manufacturing process. | |
CN117495070A (en) | Technological parameter recommendation method and system for industrial assembly line | |
CN103903065A (en) | Penicillin production process product quality prediction method based on kernel learning technology | |
CN116662804A (en) | Cut tobacco dryer cut tobacco drying quality prediction model establishment method and device based on data driving | |
CN108089436B (en) | Parameter design method for rapid non-overshoot intelligent controller | |
CN110490244A (en) | A kind of data processing method and device | |
Li et al. | Optimization and application of single-point crossover and multi-offspring genetic algorithm | |
CN114139813A (en) | Flow type production equipment product quality prediction method based on self-updating model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |