CN116364198A - Tobacco leaf raw material pyrolysis characteristic prediction method - Google Patents
Tobacco leaf raw material pyrolysis characteristic prediction method Download PDFInfo
- Publication number
- CN116364198A CN116364198A CN202310363095.2A CN202310363095A CN116364198A CN 116364198 A CN116364198 A CN 116364198A CN 202310363095 A CN202310363095 A CN 202310363095A CN 116364198 A CN116364198 A CN 116364198A
- Authority
- CN
- China
- Prior art keywords
- tobacco
- pyrolysis
- model
- weight loss
- curve
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
- G16C10/00—Computational theoretical chemistry, i.e. ICT specially adapted for theoretical aspects of quantum chemistry, molecular mechanics, molecular dynamics or the like
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/08—Thermal analysis or thermal optimisation
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Investigating Or Analyzing Materials Using Thermal Means (AREA)
Abstract
The invention discloses a tobacco raw material pyrolysis characteristic prediction method, which is mainly designed in that chemical information data and reaction conditions of tobacco raw materials, namely, nonlinear complex relation of pyrolysis heating rate and thermal weight loss curve are established by means of a mature algorithm framework, and finally a general pyrolysis model capable of reflecting the influence of tobacco types and heating rate change on the pyrolysis characteristic of tobacco is established so as to be used for predicting the pyrolysis characteristic of actually measured tobacco raw materials. The method has the advantages that the general pyrolysis model is utilized, the thermal weight loss differential curve and the pyrolysis dynamics equation of any raw material at different heating rates can be accurately predicted according to the chemical characteristics of tobacco leaves by giving the chemical composition information and the heating rate values of the tobacco leaves, the workload of thermal weight loss analysis experiments is greatly reduced, the model can be used as an effective chemical reaction dynamics module, and the model can be combined with mass, energy and momentum transfer equations related to the smoking process of cigarettes to assist in constructing a cigarette combustion calculation model.
Description
Technical Field
The invention relates to the technical field of tobacco raw materials, in particular to a tobacco raw material pyrolysis characteristic prediction method.
Background
In the field, the universal pyrolysis model of the tobacco raw materials can provide an accurate dynamics module for the establishment of a cigarette combustion calculation model, and the reliability of the cigarette combustion calculation model is directly determined. By means of the cigarette combustion calculation model, the physical and chemical processes which cannot be measured or are difficult to accurately measure by a general experimental method in the cigarette smoking process can be known, the distribution rule of smoke components after relevant parameter changes is mastered, and the method has important guiding significance for the production, processing and design of cigarettes.
The temperature rising rate difference of different positions of the combustion cone in the smoking process of the cigarette is large, the temperature rising rate and the tobacco variety are the most important factors influencing the pyrolysis characteristics of the raw materials, and how to reflect the chemical composition difference of the raw materials and the influence of the temperature rising rate change on the pyrolysis characteristics of the tobacco leaves is the key point for establishing a general pyrolysis model. The pyrolysis characteristics of a feedstock are typically studied using thermal analysis methods, i.e., analysis of the thermal weight loss of the feedstock under certain temperature program conditions to characterize its pyrolysis reaction characteristics.
For example, in the scheme currently proposed in the industry, based on a thermal weight loss differential curve under the condition of a fixed heating rate, the influence of the heating rate beta on thermal weight loss curves under other heating rate conditions is calculated by using a mathematical method, so that correction based on beta values is performed on a thermal weight loss differential curve equation under the fixed heating rate, and a tobacco pyrolysis model capable of reflecting the influence of the heating rate is established. However, the method can only carry out integral optimization correction on various heating rates on the basis of the original equation, has limited prediction precision, can only analyze one fixed tobacco raw material, considers various tobacco raw materials with various types of cigarettes, has low universality, and cannot reflect the influence of the types of the tobacco raw materials on pyrolysis characteristics.
Disclosure of Invention
In view of the foregoing, the present invention aims to provide a tobacco raw material pyrolysis characteristic prediction method, so as to solve the aforementioned technical problems.
The technical scheme adopted by the invention is as follows:
the invention provides a tobacco leaf raw material pyrolysis characteristic prediction method, which comprises the following steps:
acquiring a tobacco thermal weight loss curve, and acquiring a thermal weight loss differential curve based on the tobacco thermal weight loss curve;
collecting chemical indexes of tobacco raw materials, including industrial analysis data and element analysis data;
constructing a general pyrolysis model of tobacco leaves according to the thermal weightlessness differential curve and the chemical index; the general pyrolysis model of the tobacco leaves is used for representing the influence of chemical component differences and temperature rising rate changes of tobacco leaf raw materials on pyrolysis characteristics of the tobacco leaves;
and inputting the measured chemical indexes of the tobacco sample obtained through actual measurement into the tobacco general pyrolysis model to obtain a prediction result of the measured tobacco sample under different heating rates.
In at least one possible implementation manner, the tobacco leaf general pyrolysis model adopts a preset limit random forest model architecture.
In at least one possible implementation manner, the construction of the tobacco leaf general pyrolysis model comprises:
layering the tobacco sample based on the difference of the thermal weightlessness differential curve;
randomly extracting samples according to the layering result to construct a test set and a training set;
and training a limited random forest model based on a training set, wherein model input data are the chemical index and a plurality of established heating rates.
In at least one possible implementation manner, the prediction method further includes: and verifying the tobacco leaf general pyrolysis model.
In at least one possible implementation manner, the verification manner includes: and verifying the model accuracy of the general pyrolysis model of the tobacco leaves subjected to parameter tuning and training on a test set.
In at least one possible implementation manner, the obtaining the tobacco thermal weight loss curve and obtaining the thermal weight loss differential curve based on the tobacco thermal weight loss curve includes:
carrying out dehydration pretreatment on a raw material sample in a nitrogen atmosphere;
heating the pretreated raw material sample to a preset target temperature at a plurality of heating rates within a preset range, and recording a tobacco thermal weight loss curve at each heating rate;
differentiating the tobacco thermal weight loss curve to obtain a corresponding thermal weight loss differential curve.
In at least one possible implementation thereof, the industrial analysis data includes: ash, volatiles, and fixed carbon content.
In at least one possible implementation thereof, the elemental analysis data includes: carbon element, hydrogen element, oxygen element, nitrogen element, and sulfur element.
In at least one possible implementation manner, the prediction result includes: differential thermal weight loss prediction curve and thermal weight loss prediction curve.
Compared with the prior art, the method has the main design concept that chemical information data and reaction conditions of tobacco raw materials, namely, nonlinear complex relation of pyrolysis heating rate and thermal weight loss curve are established by means of a mature algorithm framework, and finally, a general pyrolysis model capable of reflecting the influence of tobacco types and heating rate change on the pyrolysis characteristics of tobacco is established so as to be used for predicting the pyrolysis characteristics of the actually measured tobacco raw materials. Based on the general pyrolysis model provided by the invention, the thermal weight loss differential curve and the pyrolysis kinetic equation of any raw material under different heating rates can be accurately predicted according to the chemical characteristics of tobacco, namely given chemical composition information and heating rate values of the tobacco, so that the workload of thermal weight loss analysis experiments is greatly reduced, and the model can be used as an effective chemical reaction kinetic module, and is combined with mass, energy and momentum transfer equations related to the smoking process of cigarettes to assist in constructing a cigarette combustion calculation model.
Drawings
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the present invention will be further described with reference to the accompanying drawings, in which:
fig. 1 is a schematic flow chart of a tobacco raw material pyrolysis characteristic prediction method provided by an embodiment of the invention;
fig. 2 is a schematic diagram of pyrolysis DTG curves of a certain tobacco sample at a plurality of heating rates according to an embodiment of the present invention;
fig. 3 is a schematic view of a tobacco difference distribution provided by an embodiment of the present invention;
FIG. 4 is a schematic diagram of a super-parameter optimization result of a DTG curve based on multiple heating rates of a training set sample according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of performance comparison of a model on a training set and a test set at different heating rates according to an embodiment of the present invention;
FIG. 6 is a schematic diagram showing a comparison of model predictive DTG and experimental results of a test set sample provided in an embodiment of the invention at a 400K/min heating rate;
FIG. 7 is a schematic diagram showing the comparison of TG obtained by model prediction DTG integration and experimental results of test set samples at a temperature rising rate of 400K/min according to the embodiment of the invention.
Detailed Description
Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the drawings are illustrative only and are not to be construed as limiting the invention.
The invention provides an embodiment of a tobacco leaf raw material pyrolysis characteristic prediction method, specifically as shown in fig. 1, which comprises the following steps:
s1, acquiring a tobacco thermal weight loss curve;
specifically, a thermal analyzer is used for maintaining a sample at 100 ℃ for 5-30min under a nitrogen atmosphere to carry out dehydration pretreatment; then, heating the pretreated sample to 900 ℃ at various heating rates within a range of 10-400K/min, and recording a tobacco thermal loss (TG) curve under each heating rate condition; then, the curve can be differentiated to obtain a corresponding thermogravimetric differential (differential thermogravimetric analysis DTG curve, a curve obtained by first derivative of a TG curve with respect to temperature or time) curve. Further, the process can be repeated to collect TG and DTG curves of various tobaccos under various heating rate conditions.
S2, collecting chemical indexes of tobacco raw materials, including industrial analysis data and element analysis data;
in practice, the industrial analysis data (including but not limited to moisture, ash a, volatile matter V, content of fixed carbon FC) and the elemental analysis data (including but not limited to content of carbon element C, hydrogen element H, oxygen element O, nitrogen element N, sulfur element S) of tobacco can be measured with an industrial analyzer, an elemental analyzer, respectively.
Step S3, constructing a general pyrolysis model of tobacco leaves according to the chemical indexes:
specifically, the tobacco sample is firstly subjected to differential layering treatment based on a DTG curve; randomly extracting samples based on the layering result, and constructing a test set and a training set; training a preset limit random forest model based on a training set, wherein model input data are the chemical indexes (element analysis data and industrial analysis data) of tobacco leaves and a given plurality of heating rates, and model output indexes are tobacco leaf DTG curves. And integrating the DTG curve to obtain a TG curve;
and S4, inputting chemical indexes of the measured tobacco sample obtained through actual measurement into the tobacco general pyrolysis model to obtain prediction results of the measured tobacco sample under different heating rates, wherein the prediction results specifically comprise a thermal weightlessness differential curve and a thermal weightlessness curve of tobacco.
In addition, the method further comprises verifying the tobacco leaf general pyrolysis model: and verifying the model accuracy of the general pyrolysis model of the tobacco leaves subjected to parameter tuning and training on a test set.
Based on the foregoing embodiments, the following is further described in connection with examples:
(1) Tobacco thermal weight loss (differential) curve acquisition:
49 different tobacco samples were selected and numbered from 1-49 in sequence. Before the experiment, firstly, the tobacco sample is ground by a high-speed grinder and then is placed in a sample bottle for sealing for standby. About 10mg of the sample was weighed, heated to 100℃and kept at this temperature for 5 minutes to remove the free water, and the weight of the sample which had been dehydrated and pretreated was set to 100%. And then, heating up 49 different tobacco samples with the numbers of 1-49 from 100 ℃ to 900 ℃ respectively at the heating rate of 9 types of heating rates of 10 ℃/min, 50 ℃/min, 100 ℃/min, 150 ℃/min, 200 ℃/min, 250 ℃/min, 300 ℃/min, 350 ℃/min and 400 ℃/min, and recording sample weightlessness under different heating rate conditions. During the whole experiment, the carrier gas (high purity N 2 ) And a shielding gas (high purity N) 2 ) A kind of electronic deviceThe flow rates may be set to 50mL/min and 30mL/min, respectively. Each test was repeated three times under the same conditions to eliminate the influence of the systematic error on the experimental result, and the TG curve was subjected to differential processing to obtain a DTG curve. Fig. 2 shows DTG curves of a tobacco sample under different heating rates.
(2) Acquisition of tobacco industry analysis and elemental analysis data:
after the 49 tobacco powder samples are subjected to air drying, the contents of moisture, volatile (V), fixed Carbon (FC) and ash (A) of the 49 samples are measured by a 5E-MAG6700 type full-automatic industrial analyzer according to the national standard GB/T212-2008, and the results are converted into dry ash-free radical (daf,%) data. In addition, the content of carbon (C), hydrogen (H), nitrogen (N) and sulfur (S) in 49 kinds of tobacco powder samples is analyzed according to national standards GB/T476-2008, GB/T19227-2008 and GB/T214-2007 by using 5E-CHN2200 and 5E-IRS-2 element analyzers in combination with the moisture content data detected by the industrial analyzers. Wherein, the content of oxygen (O) element is calculated by subtracting C, H, N, S from 100%, moisture and ash content. The elemental analysis data were finally converted to dry ash free (daf,%) data, and the specific results are shown in Table 1.
TABLE 1 elemental analysis and Industrial analysis results for 49 tobaccos
(3) And (3) establishing a general pyrolysis model:
there are two types of pyrolysis curves of tobacco, TG and DTG, which can be converted into each other: the TG differentiation is carried out to obtain DTG, and the DTG integration can be carried out to obtain a TG curve. The differential processing amplifies the differences between different tobacco samples, reflecting the pyrolysis reaction characteristics of the different tobacco samples to a greater extent. Therefore, in order to ensure higher accuracy of the general model, the present invention preferably uses model evaluation of the predictive ability of DTG curves, rather than TG curves. The DTG curve obtained by model prediction is subjected to integral processing to obtain a corresponding TG curve, and the prediction accuracy of the TG curve is higher than that of the DTG. The specific examples are as follows:
a) Data set classification
The tobacco is subjected to differential analysis, the DTG curve of the tobacco No. 1 under the condition of 400 ℃/min is used as the basis, the difference between the DTG curve of each other tobacco under the temperature rising rate and the tobacco No. 1 is calculated, the normalized mean square error (NRMSE) is used as a measurement standard, and the difference can be calculated according to formulas (1) and (2). The characteristic heating rate of 400 ℃/min is selected because the overlapping degree of the pyrolysis process of each compound in the tobacco is increased along with the increase of the heating rate, so that the pyrolysis process is complicated, and the difference degree among different tobaccos is increased.
Tobacco is divided into four categories according to NRMSE size, as shown in fig. 3. From each layer, 30% of the tobacco samples were randomly drawn to make up the test set, and the remaining 70% of the tobacco samples were made up the training set, with the test set sample compositions shown in table 2.
Table 2 test set sample composition
b) Model construction
Training a preset limit random forest model based on a training set, wherein input data are tobacco leaf chemical indexes (element analysis and industrial analysis data) and heating rates, and output data are DTG curves of the tobacco leaf at the corresponding heating rates, namely the corresponding mass loss rates at different temperatures (500 temperature points are uniformly selected on each curve).
In order to avoid overfitting and reduce the calculation cost, the model hyper-parameters are optimized firstly, and it can be understood that the random forest algorithm has several hyper-parameters including n_est imagers (the number of trees in the forest algorithm) and max_depth (the maximum splitting depth of a single tree), and in the training process, a 5-fold cross validation method can be adopted to perform optimization on the two hyper-parameters of the number of trees and the maximum splitting depth of the single tree. The method randomly divides the training data set into 5 equal-sized subsets, each subset being used as a validation data set to test the model, and the remaining 4 subsets being used for training, the optimal superparameter being determined in dependence on the performance of the model on the validation data set. Likewise, the accuracy of the model is evaluated with normalized mean square error (NRMSE). In the training process, each parameter is searched step by step, the search interval is 1, and the search ranges of n_est imagers and max_depth are 1-150 and 1-30 respectively. The result of model hyper-parameter optimization is shown in fig. 4, and the root mean square error of the training set and the verification set is firstly reduced sharply along with the increase of the hyper-parameters, and then stabilized at a critical value. And (3) calculating accuracy and cost are considered, and two super parameters are respectively fixed at 70 and 14.
(4) And (3) verifying and evaluating the accuracy of the model:
the model after super parameter tuning and training is used for the test set sample in the step (3), namely element analysis and industrial analysis data of the test set sample are input into the model, the DTG curve of the test set sample is predicted, the predicted DTG curve is compared with the DTG curve obtained by experimental detection, and R is utilized 2 And the RMSE carries out numerical characterization on the difference between the two, thereby verifying the accuracy of the model. FIG. 5 shows the model's DTG curve predictive capability for training and test set samples at 10K/min, 400K/min and all ramp rates. Table 3 shows the model to predict R for training set and test set samples under different heating rates 2 And RMSE.
Table 3 model performance on training and test sets
As can be seen from Table 3, the model performs excellently on the training set, R as the rate of temperature rise increases 2 Slowly decreasing from 0.999 to 0.973 with a smaller magnitude and R over the entire data 2 Reaching 0.983.RMSE exhibits a similar law, gradually increasing from 0.0072 to 0.0209 with increasing ramp rate, with a smaller increase in magnitude, and RMSE reaching 0.0138 across the whole data. A series of data demonstrated that the model performed well on the training set. At the same time, it can be observed that the model has excellent performance on the test set, and R is increased along with the increase of the temperature rising rate 2 Slowly decrease from 0.995 to 0.967, decrease in width is small, and R is on the whole data 2 Reaching 0.974.RMSE slowly increased from 0.0072 to 0.0209 with little amplification and reached 0.0174 on the whole data. The model also proved to perform well on the test set.
As can be seen from comparing the performance of the model on the test set and the training set, the performance of the model on the training set is slightly better than the performance of the model on the test set, but the difference is not great, taking 10K/min as an example, and the R of the model on the training set 2 RMSE is 0.999/0.0007, R on the test set 2 RMSE is 0.995/0.0072, slightly lower than the performance on the training set, which demonstrates that the model is not over-fitted. On the other hand, the model performs optimally at 10K/min, R on the training set 2 RMSE reached 0.999/0.0007, R on test set 2 The RMSE reached 0.995/0.0072 and the model exhibited a slow decrease in performance with increasing ramp rate. The main reason for the slip is that the pyrolysis becomes more complicated as the temperature rising rate increases, and each component exhibits a peak aggregation phenomenon.
To more clearly demonstrate the predictive power of the model, FIG. 6 shows the prediction of the test set tobacco sample DTG by the model at a least well behaved ramp rate of 400K/min. Wherein the horizontal axis represents temperature (K), the vertical axis represents DTG (%/K), and the dotted line represents experimental dataThe continuous line represents the model prediction result, and No. in each sub-graph represents the specific tobacco number, R 2 And RMSE represent R of model prediction for the tobacco 2 And RMSE. As can be seen from FIG. 6, the model prediction result is closer to the experimental result, the region with inaccurate prediction is mainly concentrated in the peak region, the temperature interval is 500-700K, and the prediction of other regions is more accurate. In the peak area, the model is mainly inaccurate in peak prediction, and accurate in peak temperature (inflection point) prediction, so that the peak temperature can be accurately predicted.
In order to further confirm the influence of the peak prediction error on the final pyrolysis TG curve, the invention integrates the DTG result predicted by the model, and the TG curve is obtained again and compared with the experimental TG curve, and the result is shown in figure 7. The abscissa of fig. 7 is temperature (K), and the ordinate is TG (%), and the graph is consistent with fig. 6. As can be seen from FIG. 7, the TG results obtained by integrating the DTG results have extremely high agreement with the experimental TG results, and all R 2 All greater than 0.99, the highest R 2 Reaching 0.9998, even the worst R 2 0.9928 was also reached, indicating that the calculation and experimental results almost completely agree. Mainly because the DTG peak value is small, the influence after integration is small, so that the deviation of the DTG peak height prediction is small, and the influence on TG obtained by final integration is small.
Therefore, the general pyrolysis model established by the invention has high accuracy, and can better predict the pyrolysis DTG curve (R 2 >0.967 Further integration of model predicted DTG data gives excellent reproduction of tobacco pyrolysis TG curves (R 2 >0.990). Because the training of the model is based on a large amount of tobacco experiment pyrolysis data, the model has better generalization performance.
In summary, the main design concept of the invention is to build chemical information data and reaction conditions of tobacco raw materials, namely, nonlinear complex relation of pyrolysis heating rate and thermal weight loss curve by means of mature algorithm architecture, and finally build a general pyrolysis model capable of reflecting the influence of tobacco variety and heating rate change on pyrolysis characteristics of tobacco, so as to predict pyrolysis characteristics of actually measured tobacco raw materials. Based on the general pyrolysis model provided by the invention, the thermal weight loss differential curve and the pyrolysis kinetic equation of any raw material under different heating rates can be accurately predicted according to the chemical characteristics of tobacco, namely given chemical composition information and heating rate values of the tobacco, so that the workload of thermal weight loss analysis experiments is greatly reduced, and the model can be used as an effective chemical reaction kinetic module, and is combined with mass, energy and momentum transfer equations related to the smoking process of cigarettes to assist in constructing a cigarette combustion calculation model.
In the embodiments of the present invention, "at least one" means one or more, and "a plurality" means two or more. "and/or", describes an association relation of association objects, and indicates that there may be three kinds of relations, for example, a and/or B, and may indicate that a alone exists, a and B together, and B alone exists. Wherein A, B may be singular or plural. The character "/" generally indicates that the context-dependent object is an "or" relationship. "at least one of the following" and the like means any combination of these items, including any combination of single or plural items. For example, at least one of a, b and c may represent: a, b, c, a and b, a and c, b and c or a and b and c, wherein a, b and c can be single or multiple.
The construction, features and effects of the present invention are described in detail according to the embodiments shown in the drawings, but the above is only a preferred embodiment of the present invention, and it should be understood that the technical features of the above embodiment and the preferred mode thereof can be reasonably combined and matched into various equivalent schemes by those skilled in the art without departing from or changing the design concept and technical effects of the present invention; therefore, the invention is not limited to the embodiments shown in the drawings, but is intended to be within the scope of the invention as long as changes made in the concept of the invention or modifications to the equivalent embodiments do not depart from the spirit of the invention as covered by the specification and drawings.
Claims (9)
1. A method for predicting pyrolysis characteristics of tobacco raw materials, comprising the steps of:
acquiring a tobacco thermal weight loss curve, and acquiring a thermal weight loss differential curve based on the tobacco thermal weight loss curve;
collecting chemical indexes of tobacco raw materials, including industrial analysis data and element analysis data;
constructing a general pyrolysis model of tobacco leaves according to the thermal weightlessness differential curve and the chemical index; the general pyrolysis model of the tobacco leaves is used for representing the influence of chemical component differences and temperature rising rate changes of tobacco leaf raw materials on pyrolysis characteristics of the tobacco leaves;
and inputting the measured chemical indexes of the tobacco sample obtained through actual measurement into the tobacco general pyrolysis model to obtain a prediction result of the measured tobacco sample under different heating rates.
2. The method for predicting pyrolysis characteristics of tobacco raw materials according to claim 1, wherein the general pyrolysis model of tobacco adopts a preset limit random forest model architecture.
3. The method for predicting pyrolysis characteristics of tobacco raw materials according to claim 2, wherein the constructing a general pyrolysis model of tobacco comprises:
layering the tobacco sample based on the difference of the thermal weightlessness differential curve;
randomly extracting samples according to the layering result to construct a test set and a training set;
and training a limited random forest model based on a training set, wherein model input data are the chemical index and a plurality of established heating rates.
4. A method for predicting pyrolysis characteristics of a tobacco raw material according to claim 3, further comprising: and verifying the tobacco leaf general pyrolysis model.
5. The method for predicting pyrolysis characteristics of tobacco raw materials according to claim 4, wherein the verification method comprises: and verifying the model accuracy of the general pyrolysis model of the tobacco leaves subjected to parameter tuning and training on a test set.
6. The method of claim 1, wherein the obtaining a tobacco thermal weight loss curve and obtaining a thermal weight loss differential curve based on the tobacco thermal weight loss curve comprises:
carrying out dehydration pretreatment on a raw material sample in a nitrogen atmosphere;
heating the pretreated raw material sample to a preset target temperature at a plurality of heating rates within a preset range, and recording a tobacco thermal weight loss curve at each heating rate;
differentiating the tobacco thermal weight loss curve to obtain a corresponding thermal weight loss differential curve.
7. The method for predicting pyrolysis characteristics of tobacco raw materials according to claim 1, wherein the industrial analysis data comprises: ash, volatiles, and fixed carbon content.
8. The method for predicting pyrolysis characteristics of tobacco raw materials according to claim 1, wherein the elemental analysis data comprises: carbon element, hydrogen element, oxygen element, nitrogen element, and sulfur element.
9. The method for predicting pyrolysis characteristics of tobacco raw materials according to any one of claims 1 to 8, wherein the prediction result comprises: differential thermal weight loss prediction curve and thermal weight loss prediction curve.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310363095.2A CN116364198A (en) | 2023-04-06 | 2023-04-06 | Tobacco leaf raw material pyrolysis characteristic prediction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310363095.2A CN116364198A (en) | 2023-04-06 | 2023-04-06 | Tobacco leaf raw material pyrolysis characteristic prediction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116364198A true CN116364198A (en) | 2023-06-30 |
Family
ID=86907216
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310363095.2A Pending CN116364198A (en) | 2023-04-06 | 2023-04-06 | Tobacco leaf raw material pyrolysis characteristic prediction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116364198A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117195699A (en) * | 2023-08-22 | 2023-12-08 | 润城帝景(北京)科技发展有限公司 | Method for predicting biomass pyrolysis kinetic parameters |
-
2023
- 2023-04-06 CN CN202310363095.2A patent/CN116364198A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117195699A (en) * | 2023-08-22 | 2023-12-08 | 润城帝景(北京)科技发展有限公司 | Method for predicting biomass pyrolysis kinetic parameters |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Improved PLS regression based on SVM classification for rapid analysis of coal properties by near-infrared reflectance spectroscopy | |
| Zhou et al. | A novel method for kinetics analysis of pyrolysis of hemicellulose, cellulose, and lignin in TGA and macro-TGA | |
| CN102539377B (en) | Intermediate infrared absorption spectra based method for multi-component mixed gas qualitative and quantitative analysis | |
| Huang et al. | Improved generalization of spectral models associated with Vis-NIR spectroscopy for determining the moisture content of different tea leaves | |
| AU2018337131B2 (en) | Method for detecting raw coal moisture and volatile matter using amount of baseline drift | |
| CN116364198A (en) | Tobacco leaf raw material pyrolysis characteristic prediction method | |
| CN113049438B (en) | Method for rapidly identifying heat conversion characteristics of different tobaccos based on macroscopic quantity thermogravimetry | |
| CN114624142B (en) | Tobacco total sugar and reducing sugar quantitative analysis method based on pyrolysis kinetic parameters | |
| Soria-Verdugo et al. | Combining the lumped capacitance method and the simplified distributed activation energy model to describe the pyrolysis of thermally small biomass particles | |
| CN106501121A (en) | A kind of assay method of the volatile matters of coal and application | |
| Kandala et al. | Capacitance sensing of moisture content in fuel wood chips | |
| CN110412115A (en) | Unknown time green tea source area prediction technique based on stable isotope and multielement | |
| CN107797965A (en) | A kind of update method and system of pulverizer outlet temperature secure setting | |
| CN114923869A (en) | Method for identifying recycled plastic based on combination of spectroscopy, thermal analysis and data fusion strategy and chemometric method | |
| CN113155774A (en) | Textile material terahertz spectrum quantitative detection method | |
| CN108303493A (en) | The prediction technique of battery water content | |
| CN114813635B (en) | Method for optimizing combustion parameters of coal stove and electronic equipment | |
| CN112861415A (en) | Biomass ash content measurement and modeling method based on near infrared spectrum principal component and neural network | |
| CN112861412A (en) | Biomass volatile component content measurement and modeling method based on near infrared spectrum principal component and neural network | |
| CN112861413A (en) | Biomass water content measurement and modeling method based on near infrared spectrum principal component and neural network | |
| Yin et al. | Optimization of coal blending process control based on fuzzy self-tuning | |
| CN112861299A (en) | Biomass chlorine content measurement and modeling method based on infrared spectrum principal component and neural network | |
| CN116773390A (en) | Tobacco element content detection method based on non-isothermal thermal analysis under inert atmosphere | |
| CN114646746B (en) | Method for determining volatile components of coal for power | |
| CN112858205A (en) | Biomass hydrogen content measurement and modeling method based on infrared spectrum principal component and neural network |
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 |