CN112021626B - Intelligent control system and method for tobacco shred making link - Google Patents
Intelligent control system and method for tobacco shred making link Download PDFInfo
- Publication number
- CN112021626B CN112021626B CN202010662329.XA CN202010662329A CN112021626B CN 112021626 B CN112021626 B CN 112021626B CN 202010662329 A CN202010662329 A CN 202010662329A CN 112021626 B CN112021626 B CN 112021626B
- Authority
- CN
- China
- Prior art keywords
- parameters
- production
- prediction
- parameter
- model
- 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
- A24B3/04—Humidifying or drying tobacco bunches or cut tobacco
-
- 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
- A24B3/06—Loosening tobacco leaves or cut tobacco
-
- 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
-
- A—HUMAN NECESSITIES
- A24—TOBACCO; CIGARS; CIGARETTES; SIMULATED SMOKING DEVICES; SMOKERS' REQUISITES
- A24B—MANUFACTURE OR PREPARATION OF TOBACCO FOR SMOKING OR CHEWING; TOBACCO; SNUFF
- A24B9/00—Control of the moisture content of tobacco products, e.g. cigars, cigarettes, pipe tobacco
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
Abstract
The invention discloses an intelligent control system and method for a tobacco shred making link, wherein the system comprises an information platform, and a trend monitoring module, a key parameter mutual-pushing module and a historical data query module which are in communication connection with the information platform, wherein the trend monitoring module comprises a real-time prediction unit, an intelligent control unit, an overproof early warning unit, an intelligent recommendation unit and a self-learning unit. According to the method, historical production data is mined and analyzed, operation methods such as a multiple regression analysis method and a neural network algorithm are integrated, a prediction model is built in stages, forward prediction and backward prediction of key parameters under the influence of different factors are obtained through model solving, so that accurate control of moisture at a cut tobacco drying inlet is realized, and functions such as online quality early warning, correlation comparison, real-time prediction, trend prediction, error prevention, upstream prediction downstream, downstream feedback upstream and intelligent recommendation are realized through a modeling-verification-application-optimization process.
Description
Technical Field
The invention relates to the tobacco industry, in particular to the field of intelligent control of a tobacco shred production line, and particularly relates to an intelligent control system and method for a tobacco shred production link.
Background
In the tobacco production process, the tobacco shred production process is a very important link, and is a process for making the tobacco leaves into qualified tobacco shreds gradually through various processing procedures according to the physicochemical characteristics of the tobacco leaf raw materials and a certain program. In the cigarette production process, the process flow of the tobacco shred manufacturing is longest, the working procedures are most complicated, and the equipment types are most. The production operation of the existing silk-making workshop is in a production line type, and the existing silk-making workshop comprises key equipment such as a vacuum moisture regaining machine, a loosening moisture regaining machine, a temporary storage cabinet, a feeding moisture regaining machine, a hot air leaf moistening machine, a filament cutter, a silk dryer and the like.
At present, production planning and scheduling of various domestic large tobacco enterprises are still manually completed by operators through experience, and all the devices are manually or semi-automatically operated, so that quality difference among product batches is large, and homogenization of products is not facilitated. Namely, a great deal of manual coordination and resource balance problems exist in the tobacco shred manufacturing process, limited manpower is difficult to ensure the accuracy of coordination and balance, production pause and feeding interruption greatly influence the production quality of enterprises, and therefore intelligent control of the tobacco shred manufacturing process greatly influences the production cost and economic benefits of the enterprises.
Publication No.: the invention application of CN111103854A discloses a system and a method for improving production stability of a tobacco shred drying machine, belonging to the technical field of tobacco shred production. The system comprises a data acquisition module, a data processing module, a stability evaluation module, an early warning and optimization module, a steady state real-time analysis and evaluation module, an unsteady state automatic control judgment module, an unsteady state real-time analysis and evaluation module and the like. The invention takes the new generation information technology such as industrial big data, artificial intelligence and the like as support, reduces the unsteady state time and improves the steady state control quality, thereby effectively improving the quality stability of the whole production process of the cut tobacco dryer.
The notice number is: the invention patent of CN105759764A discloses a cigarette production process parameter control system and a control method thereof. The system consists of an MC2 cigarette physical detector, a physical index acquisition device, a cigarette making machine electric control system, a process parameter acquisition device, an application server, a real-time history/relational database, a data analysis server, a field operation terminal, a process parameter control server and an industrial Ethernet, wherein the real-time history/relational database consists of a relational database server and a real-time history database server; the invention forms a complete closed loop system comprising acquisition, analysis and control, and really realizes the intelligent control of the cigarette production process parameters.
Publication No.: the invention application of CN101488024A discloses an online quality evaluation and real-time intelligent control method for tobacco processing process parameters, which comprises the following steps: presetting parameter values between the operating parameters of the raw material processing equipment and the internal quality of a processed product through a data management system; online acquisition is carried out in real time through a data management acquisition system; comparing the real-time online acquired parameters with preset parameter values through a data management layer intelligent control system; the accurate and dynamic control of the raw material quality change trend in the tobacco online processing process is realized by adjusting the running parameters of the PLC control system and the set values of the PID opening and closing ring automatic control parameters, controlling the valves, fans and air door executing elements on the equipment, and controlling the parameters of the cut tobacco moisture regaining and the rotating speed of the cylinder body of the tobacco moistening machine, and the real-time control of the inherent quality of the product in a high-quality and stable range is ensured.
The above patent documents collect data parameters in tobacco processing or tobacco shred processing links, and analyze and compare the collected data parameters, so as to improve the quality stability of the tobacco processing process and realize the intelligent control of process parameters. However, the intelligent control of the method still stays at a relatively primary stage, prediction, early warning, recommendation and learning of process parameters cannot be realized, the interactivity among the process parameters is poor, the accurate control of the moisture of the cut tobacco drying inlet cannot be realized, and the product homogeneity level is improved from the source.
Disclosure of Invention
In order to overcome the defects in the prior art, the application provides an intelligent control system and method for a tobacco shred making link.
The method relies on massive historical production data collected in an information platform, and by using a data model concept for reference, by mining and analyzing the historical production data, a multivariate regression analysis method, a neural network algorithm and other operation methods are integrated, a production parameter prediction model is established in stages, and forward and backward predictions of key parameters under the influence of different factors are obtained through model solving, so that the accurate control of the moisture of a cut tobacco drying inlet of the cut tobacco making process and the monitoring of the production trend are realized; establishing an intelligent control model and a parameter mutual-pushing model, and finally realizing the functions of online quality early warning, correlation comparison, real-time prediction, trend prediction, error prevention, upstream prediction downstream, downstream feedback upstream, intelligent recommendation and the like through the processes of modeling, verification, application and optimization.
The invention aims to solve the product homogeneity level, improve the accuracy of tobacco shred moisture control, reduce the number of operators and realize real intelligent control of a tobacco shred production line.
The technical scheme adopted by the invention for solving the technical problems is as follows:
the tobacco shred making link intelligent control system comprises an information platform, wherein historical production data are stored in the information platform, the historical production data comprise historical batch parameters and parameter relations, and the system further comprises a control system which is in communication connection with the information platform
The trend monitoring module comprises a real-time prediction unit, an intelligent control unit, an overproof early warning unit, an intelligent recommendation unit and a self-learning unit;
the key parameter mutual-pushing module is used for establishing a parameter mutual-pushing model through a parameter relation on the basis of historical batch parameters, predicting and mutually pushing the key parameters, verifying the production correctness of the key parameters and searching abnormal parameters through parameter reverse pushing;
the historical data query module is internally stored with historical prediction data and actual parameter data and can be directly consulted when a corresponding batch needs to be queried;
wherein:
the trend monitoring module builds a nested production parameter prediction model in stages according to process working sections based on historical production data and combined with a multiple regression analysis algorithm and a neural network algorithm, predicts corresponding production parameters, builds an intelligent control model in stages according to the process working sections based on production parameter prediction values through the neural network algorithm, predicts corresponding control parameters, and intelligently controls equipment of each process working section based on the control parameter prediction values.
As an improvement of the above technical solution, the nested production parameter prediction model includes a linear equation prediction model and a neural network prediction model respectively constructed by a multiple regression analysis algorithm and a neural network algorithm, wherein:
predicting the production parameters with the batch standard deviation less than or equal to 0.2 through a linear equation prediction model;
the production parameters with batch standard deviation larger than 0.2 are predicted through a neural network prediction model.
As an improvement of the technical scheme, the parameter mutual-deduction model is a linear or nonlinear mutual-deduction model of key parameters, which is constructed through a multiple regression analysis algorithm based on a production parameter predicted value and a parameter relation of a real-time prediction unit.
As an improvement of the technical scheme, the overproof early warning unit is provided with a parameter standard upper limit and a parameter standard lower limit which are used for judging production parameters and control parameters which exceed a process range and giving an alarm.
As an improvement of the technical scheme, the intelligent recommendation unit adopts a similarity matching algorithm, realizes intelligent real-time recommendation of the control parameters of the whole process of the silk making process through association and integration of cigarette production process conditions and historical production data and through construction, matching, optimization and recommendation steps.
As an improvement of the technical scheme, the self-learning unit improves the model prediction precision through self-learning according to the continuous accumulation of historical production data.
As an improvement of the technical scheme, the working procedure working section comprises a loosening and moisture regaining working section, a leaf moistening and feeding working section and a hot air leaf moistening working section.
As an improvement of the technical scheme, the intelligent control model of the intelligent control unit is provided with a feedback loop, and the feedback loop is connected with a monitoring sensor.
The invention also provides an intelligent control method for the tobacco shred making link, which is applied to any one of the intelligent control systems for the tobacco shred making link, and comprises the following steps:
step one, data acquisition
The data source is as follows: corresponding historical production data in the silk-making line information management system;
step two, establishing a prediction model
Establishing a nested production parameter prediction model by stages according to process sections by combining a multiple regression analysis algorithm and a neural network algorithm, and predicting production parameters of each section;
step three, establishing a control model
Establishing an intelligent control model by stages according to the working procedure working sections through a neural network algorithm based on the predicted value in the second step, predicting control parameters of each working section, and intelligently controlling equipment of each working procedure working section based on the predicted values of the control parameters;
step four, mutually pushing key parameters
Establishing a parameter mutual-pushing model based on the predicted value in the step two and the historical batch parameters through a parameter relation, and performing prediction mutual-pushing on key parameters;
the production correctness of the key parameters can be verified through a parameter mutual-deduction model, and abnormal key parameters can be searched through parameter reverse deduction;
step five, controlling parameters to recommend in real time
The method adopts a similarity matching algorithm, realizes intelligent real-time recommendation of the control parameters of the whole process of the tobacco making process through association and integration of the conditions of the cigarette production process and historical production data and through the steps of construction, matching, optimization and recommendation;
step six, exceeding the standard early warning
Based on the upper and lower limits of the parameter standard, the predicted values of the production parameters and the recommended values of the control parameters are automatically judged, and the parameters exceeding the process range are alarmed;
step seven, self-learning
According to the continuous accumulation of historical production data, the model is continuously optimized through self-learning, and the model prediction precision is improved;
step eight, trend monitoring and historical data query
Collecting the prediction data of the current time period into a prediction curve for production trend monitoring;
and summarizing the historical prediction data and the actual parameter data into a data curve, and directly consulting when a corresponding batch needs to be queried.
As an improvement of the above technical solution, in the third step:
the intelligent control model is connected with the feedback loop, optimizes control parameters in real time according to feedback data, and achieves intelligent and accurate control.
The invention has the following beneficial effects:
compared with the prior art, the method has the advantages that:
(1) the real-time prediction function of the invention can ensure that the moisture display points of each process section are matched with the time points of the control process in the production process of the silk making, so that the prediction, the control and the feedback are synchronous, and the accuracy of an intelligent control system is improved;
(2) the intelligent control function of the invention is based on the real-time prediction function, an intelligent control model is established, and the intelligent accurate control of the tobacco shred moisture in the shred manufacturing process is realized by calculating and continuously optimizing the control parameters of the corresponding sections in real time according to the prediction data and the feedback data;
(3) according to the tobacco shred production monitoring system, the standard exceeding early warning unit is arranged, so that the instability and batch difference of tobacco shred moisture in the production process can be prevented, the production abnormity can be foreseen in advance, abnormity early warning can be carried out, the warning history can be displayed in the warning area of the standard exceeding early warning unit, and an operator can carry out pretreatment according to the warning prompt, so that the production parameters meet the process standard;
(4) the intelligent recommendation unit adopts a similarity matching algorithm, and realizes intelligent real-time recommendation of full-process optimal control parameters and optimization of the intelligent control mode and capacity of the system by optimizing operation modes of key steps such as construction, matching, optimization, recommendation and the like;
(5) by setting the self-learning unit, the model prediction precision can be improved and the upgrading of the model prediction capability can be realized by self-learning according to historical production data, such as historical predicted values and historical measured values of the moisture of the cut tobacco inlet in different temperature and humidity environments all year round in the production process;
(6) in the specific implementation process of the invention, the functions of online quality early warning, correlation comparison, real-time prediction, trend prediction, error prevention, upstream prediction downstream, downstream feedback upstream, intelligent recommendation and the like can be finally realized by following the IS-PDCA methodology and through the flow of modeling, verification, application and optimization.
Drawings
The invention will be further described with reference to the accompanying drawings and specific embodiments,
FIG. 1 is a block diagram of the system of the present invention;
FIG. 2 is a process flow diagram of a possible intelligent control unit of the present invention;
fig. 3 is a process flow diagram of a possible out-of-compliance warning unit of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the present invention, unless otherwise expressly specified or limited, the terms "disposed," "mounted," "connected," and "fixed" are to be construed broadly and may, for example, be fixedly connected or detachably connected; may be a mechanical connection; may be directly connected or indirectly connected through an intermediate. The specific meanings of the above terms in the present invention can be understood by those skilled in the art according to specific situations.
Example 1
Referring to fig. 1, the embodiment is an intelligent control system for tobacco shred making links, which comprises an information platform, wherein historical production data are stored in the information platform: the historical production data comprises historical batch parameters, parameter relations, key parameters (production key indexes such as moisture at a feeding and moisture regaining outlet, moisture at a hot air leaf moistening outlet and moisture at a cut tobacco drying inlet) and historical prediction data, actual parameter data (measured data obtained by monitoring sensors such as a moisture meter, a flowmeter, an electronic scale and the like), influence factors such as environmental temperature and humidity changing along with time and seasons and other relevant historical production data which can be used in an intelligent control system.
The system also comprises a trend monitoring module, a key parameter prediction module and a historical data query module which are in communication connection with the information platform, wherein:
the trend monitoring module comprises a real-time prediction unit, an intelligent control unit, an overproof early warning unit, an intelligent recommendation unit and a self-learning unit; the trend monitoring module builds a nested production parameter prediction model in stages according to process sections based on historical production data by combining a multiple regression analysis algorithm and a neural network algorithm, predicts corresponding production parameters, builds an intelligent control model in stages according to the process sections through the neural network algorithm based on production parameter prediction values, predicts corresponding control parameters, and intelligently controls equipment of each process section based on the control parameter prediction values;
the key parameter mutual-pushing module is used for establishing a parameter mutual-pushing model through a parameter relation on the basis of historical batch parameters, carrying out prediction mutual pushing on the key parameters, verifying the production correctness of the key parameters and searching abnormal parameters through parameter reverse pushing;
and the historical data query module is internally stored with historical prediction data and actual parameter data and can directly refer to the historical prediction data and the actual parameter data when a corresponding batch needs to be queried.
In this embodiment:
and the nested production parameter prediction model is built by stages according to the working procedure sections and is used for predicting the production parameters of the corresponding working sections.
Each working procedure section comprises (but not limited to) a loose moisture regaining section, a leaf moistening and feeding section and a hot air leaf moistening section which are defined according to the silk making production line.
The multi-module nested production parameter prediction model is a linear equation prediction model and a neural network prediction model which are respectively constructed by a multiple regression analysis algorithm and a neural network algorithm, wherein:
predicting the production parameters with the batch standard deviation less than or equal to 0.2 through a linear equation prediction model;
the production parameters with batch standard deviation larger than 0.2 are predicted through a neural network prediction model.
Specifically, the standard deviation is (measured value-predicted value)/predicted value;
the batch standard deviation is less than or equal to 0.2, such as: temperature at a loosening and moisture regaining outlet, moisture at a hot air leaf moistening outlet and the like;
batch bias > 0.2 production parameters, such as: temperature of loose moisture regaining hot air, temperature and flow rate of loose moisture regaining circulating air, moisture at a feeding moisture regaining outlet, temperature and flow rate of feeding moisture regaining circulating air, moisture at a hot air leaf moistening inlet and the like.
The explanation is given by taking a model for predicting the moisture at the outlet of a loosening and conditioning workshop section as an example:
the outlet water prediction model of the loosening and conditioning section is a multiple regression linear equation prediction model established by taking the inlet water (X1) of the loosening and conditioning, the total pumping amount (X3), the compensation steam opening (X4), the time (X5) from vacuum conditioning to loosening and conditioning as influence factors and the outlet water (X2) of the loosening and conditioning as a dependent variable.
The coefficients of the various influencing factors are shown in table 1:
table 1: influence factor coefficient of loose moisture regaining moisture meter calibration model
Establishing a regression equation according to each influence factor coefficient:
y (loose moisture regain outlet moisture X2) ═ 2.802X1+0.351X3-0.024X4+0.003X5-0.922
The equation goodness of fit is 0.903.
By adopting the formula, the process outlet water content can be predicted by combining known parameters (average value or measured value of historical production data and intelligent recommended value). And the prediction result is used for constructing the intelligent control model.
The explanation is given by taking a prediction model of moisture at the moisture regain outlet of the feeding material as an example:
the calibration model of the feeding and conditioning moisture meter is a neural network prediction model established by taking feeding and conditioning inlet moisture as input, feeding and conditioning outlet moisture as output, and the storage time, the compensation steam opening and the feeding and conditioning moisture discharge opening of a temporary storage cabinet as influence factors.
The coefficients of the various influencing factors are shown in Table 2:
table 2: neural network structure information of calibration model of feeding moisture regaining instrument
In table 2: v17-temporary storage cabinet storage time; v20-feed moisture regain inlet moisture; v22-moisture at the moisture regain of the feed; v24-feeding and dampening to compensate the steam opening; v26-moisture removal opening degree of feeding and moisture regaining.
The importance of each of the above influencing factors is shown in Table 3:
table 3: importance of each influence factor of charging moisture regaining instrument calibration model
| Nodes | Importance | Importance | V22 |
| V26 | 0.104 | 0.104 | 0.104 |
| V17 | 0.1651 | 0.1651 | 0.1651 |
| V24 | 0.2469 | 0.2469 | 0.2469 |
| V20 | 0.484 | 0.484 | 0.484 |
The outlet moisture of the process can be predicted by using the neural network prediction model and combining with known parameters. And the prediction result is used for constructing the intelligent control model.
Based on the predicted values of the production parameters of all the sections, an intelligent control model of the corresponding section is established through a neural network algorithm, the corresponding control parameters are predicted, and intelligent control can be performed on equipment of the section based on the predicted values of the control parameters.
Further:
the trend monitoring module comprises a real-time prediction unit, an intelligent control unit, a standard exceeding early warning unit, an intelligent recommendation unit and a self-learning unit:
the real-time prediction unit establishes a production parameter prediction model through a mathematical modeling method, calculates predicted values of various production parameters on line and in real time, and has a real-time prediction function, so that moisture display points of various process sections in a silk making production process are matched with time points of control processes, prediction, control and feedback are synchronous, and the accuracy of an intelligent control system is improved;
the intelligent control unit establishes an intelligent control model based on the predicted value of the real-time prediction unit, the intelligent control model is provided with a feedback loop, the feedback loop is connected with a monitoring sensor, and control parameters are calculated and optimized according to feedback data in real time to realize intelligent accurate control;
taking a hot air leaf moistening section as an example:
referring to fig. 2, the hot air leaf moistening machine section establishes an intelligent control model based on a prediction model thereof. This intelligent control model is mainly based on historical production data and predicted value, and the hot-blast moist leaf of application entry moisture, the hot-blast moist leaf export moisture of hot-blast moist shred entry moisture of application construct through neural network algorithm to the hot-blast moist leaf machine of intelligent control compensates the steam valve opening, makes the moisture of the inlet of dried shred stable:
when a material reaches a hot air leaf moistening inlet, presetting a compensation steam opening by an intelligent control model according to historical production data, running a linear equation prediction model before the material reaches a hot air leaf moistening outlet, predicting the moisture of the hot air leaf moistening outlet according to production conditions and other influence factors, and adjusting the compensation steam opening in real time by the intelligent control model according to the predicted value so as to enable the outlet moisture to approach the predicted value;
when the material reaches the hot air leaf-moistening outlet, the intelligent control model carries out correlation comparison with a predicted value and key parameter historical production data according to the measured value of the moisture meter at the position to generate feedback data, whether the setting of the feedback control parameters is proper or not is fed back through a feedback loop, and the improper control parameters are adjusted in real time to intelligently control the hot air leaf-moistening to compensate the steam opening, so that the moisture at the cut tobacco drying inlet tends to be stable, the stability and consistency of the moisture of the cut tobacco incoming material are improved, and the product homogenization level is further improved.
The standard exceeding early warning unit is provided with standard upper and lower limits of each parameter according to the principle of 'standard based and strict standard', and is used for automatically judging the production process parameters exceeding the process range and giving an alarm.
Referring to fig. 3, the excessive early warning unit is disposed between the real-time prediction unit and the intelligent control unit of each section. Based on the predicted value and the historical production data of the real-time prediction unit, when the predicted value and the historical production data (such as the average value of moisture at a cut tobacco drying inlet and the standard value of moisture at a cut tobacco drying inlet) at the same time section exceed the set upper and lower limits of the standard, the abnormal problem of a certain production parameter/certain production parameters and control parameters in the process range of a cut tobacco production line is indicated.
In order to accurately find the abnormal parameters, the abnormal parameters can be timely found through parameter relations (such as linear/nonlinear relations, logical relations, data association and the like), upstream tracing and parameter back-stepping by combining prediction models of all sections and moisture prediction values.
The overproof early warning unit can also carry out overproof early warning on the measured water value of each section, each production parameter value in the process range, the abnormal prediction curve and the like which exceed the upper limit and the lower limit of the corresponding standard so as to enable each parameter to accord with the process standard.
The intelligent recommendation unit adopts a similarity matching algorithm, and realizes intelligent real-time recommendation of the optimal operating parameters of the whole process of the silk making process through association and integration of cigarette production process conditions (such as incoming material moisture, loose moisture regain and water adding quantity) and historical production data and through steps of parameter recommendation model construction, similar parameter matching, parameter optimization and optimal parameter recommendation.
For example, when the intelligent control unit predicts and controls the control parameter of the opening degree of the compensation steam valve of the hot air leaf moistening machine in real time, the intelligent recommendation unit recommends the control parameter in real time based on the linear relation (or similarity) between the control parameter and some other control parameter.
The self-learning unit improves the model prediction precision through self-learning according to the continuous accumulation of historical production data. The self-learning unit and the intelligent recommendation unit supplement each other and can be regarded as a model optimization unit.
For example, according to historical production data, historical predicted values and historical measured values of moisture at the inlet of the cut tobacco drying machine in different temperature and humidity environments all year around in the production process can be intelligently summarized, the influence of the environmental temperature and humidity caused by seasonal changes is eliminated, and the prediction model is more accurate and complete.
Further:
the key parameter mutual-pushing module establishes a parameter mutual-pushing model through a parameter relation on the basis of historical batch parameters, performs prediction mutual-pushing aiming at the key parameters, and can also verify the production correctness of the key parameters and search abnormal parameters through parameter reverse-pushing.
The parameter mutual-deducing model is a linear or nonlinear mutual-deducing model of key parameters constructed by a multiple regression analysis algorithm based on the production parameter predicted value and parameter relation of the real-time prediction unit.
For example, when the intelligent recommendation unit cannot perform parameter similarity matching and intelligent recommendation on the nonlinear parameters, the function can be realized through a parameter mutual-pushing model.
If so, obtaining a predicted value of the outlet moisture of the section based on a feed moisture regain outlet moisture prediction model, carrying out parameter mutual-deduction (prediction) on the outlet moisture of the hot-air leaf moistening by virtue of a parameter relation between the value of the outlet moisture of the hot-air leaf moistening of the next section and the value of the outlet moisture of the hot-air leaf moistening of the next section based on historical production data of the moisture value, predicting or monitoring the production trend of the hot-air leaf moistening section based on the recommended value, and verifying the production correctness of key parameters;
for example, when the compensation steam opening is predicted, controlled and adjusted in real time based on an intelligent control model of a hot air leaf moistening machine section, the control parameters are mutually inferred in real time based on the nonlinear relation (obtained by analyzing historical production data) between the compensation steam opening and some other control parameters.
By means of the parameter mutual-deduction model, based on the key parameters and the parameter relation, the abnormal key parameters can be found in time through parameter reverse deduction and upstream tracing.
In this embodiment:
historical prediction data and actual parameter data are stored in the historical data query module, can be summarized into a curve, and can be directly consulted when a corresponding batch needs to be queried.
The prediction curve of the predicted value assembly of the real-time prediction unit, the adjustment history of the control parameters of the intelligent control unit, the actual curve of the moisture measured value assembly of each section such as a cut tobacco inlet, the production trend of key parameters and the like can be stored in the module and can be consulted at any time.
Example 2
The method is applied to the intelligent control system for the tobacco shred making link in the embodiment 1, and comprises the following steps:
step one, data acquisition
The data source is as follows: corresponding historical production data in the silk-making line information management system;
step two, establishing a prediction model
Establishing a nested production parameter prediction model by stages according to process sections by combining a multiple regression analysis algorithm and a neural network algorithm, and predicting production parameters of each section;
step three, establishing a control model
Establishing an intelligent control model by stages according to the working procedure working sections through a neural network algorithm based on the predicted value in the second step, predicting control parameters of each working section, and intelligently controlling equipment of each working procedure working section based on the predicted values of the control parameters;
the intelligent control model is connected with the feedback loop, optimizes control parameters according to feedback data in real time and realizes intelligent and accurate control;
step four, mutually pushing key parameters
Establishing a parameter mutual-pushing model based on the predicted value in the step two and the historical batch parameters through a parameter relation, and performing prediction mutual-pushing on key parameters;
the production correctness of the key parameters can be verified through a parameter mutual-deduction model, and abnormal key parameters can be searched through parameter reverse deduction;
step five, controlling parameters to recommend in real time
The method adopts a similarity matching algorithm, realizes intelligent real-time recommendation of the control parameters of the whole process of the tobacco making process through association and integration of the conditions of the cigarette production process and historical production data and through the steps of construction, matching, optimization and recommendation;
step six, exceeding the standard early warning
Based on the upper and lower limits of the parameter standard, the predicted values of the production parameters and the recommended values of the control parameters are automatically judged, and the parameters exceeding the process range are alarmed;
step seven, self-learning
According to the continuous accumulation of historical production data, the model is continuously optimized through self-learning, and the model prediction precision is improved;
step eight, trend monitoring and historical data query
Collecting the prediction data of the current time period into a prediction curve for production trend monitoring;
and summarizing the historical prediction data and the actual parameter data into a data curve, and directly consulting when a corresponding batch needs to be queried.
Finally, it should be noted that: although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that modifications may be made to the embodiments or portions thereof without departing from the spirit and scope of the invention.
Claims (9)
1. The intelligent control system for the tobacco shred production link comprises an information platform, wherein historical production data are stored in the information platform, and the historical production data comprise historical batch parameters, parameter relationships, key parameters, historical prediction data, actual parameter data and influence factors, and is characterized in that: the system also includes a communication interface with the information platform
The trend monitoring module comprises a real-time prediction unit, an intelligent control unit, an overproof early warning unit, an intelligent recommendation unit and a self-learning unit;
the key parameter mutual-pushing module is used for establishing a parameter mutual-pushing model through a parameter relation on the basis of historical batch parameters, predicting and mutually pushing the key parameters, verifying the production correctness of the key parameters, and searching abnormal parameters through parameter reverse-pushing, wherein the key parameters comprise feed moisture regain outlet moisture, hot air leaf moistening outlet moisture and cut tobacco drying inlet moisture;
the historical data query module is internally stored with historical prediction data and actual parameter data and can be directly consulted when a corresponding batch needs to be queried;
wherein:
the trend monitoring module builds a nested production parameter prediction model by stages according to process working sections based on historical production data by combining a multiple regression analysis algorithm and a neural network algorithm, predicts corresponding production parameters, builds an intelligent control model by stages according to the process working sections based on production parameter prediction values through the neural network algorithm, predicts corresponding control parameters, and intelligently controls equipment of each process working section based on the control parameter prediction values;
the nested production parameter prediction model comprises a linear equation prediction model and a neural network prediction model which are respectively constructed by a multiple regression analysis algorithm and a neural network algorithm, wherein:
predicting the production parameters with the batch standard deviation less than or equal to 0.2 through a linear equation prediction model;
predicting the production parameters with the batch standard deviation larger than 0.2 through a neural network prediction model;
the production parameters with the batch standard deviation less than or equal to 0.2 comprise loose moisture regain outlet temperature, loose moisture regain outlet moisture and hot air leaf moistening outlet moisture;
the production parameters with batch standard deviation larger than 0.2 comprise loose moisture regaining hot air temperature, loose moisture regaining circulating air temperature and flow velocity, feeding moisture regaining outlet moisture, feeding moisture regaining circulating air temperature and flow velocity, and hot air leaf moistening inlet moisture.
2. The tobacco shred making link intelligent control system according to claim 1, wherein: the parameter mutual-deducing model is a linear or nonlinear mutual-deducing model of key parameters, which is constructed by a multiple regression analysis algorithm based on the production parameter predicted value and the parameter relation of the real-time prediction unit.
3. The tobacco shred making link intelligent control system according to claim 1, wherein: the overproof early warning unit is provided with parameter standard upper and lower limits which are used for judging production parameters and control parameters which exceed the process range and giving an alarm.
4. The tobacco shred making link intelligent control system according to claim 1, wherein: the intelligent recommendation unit adopts a similarity matching algorithm, realizes intelligent real-time recommendation of the control parameters of the whole process of the tobacco making process through association and integration of cigarette production process conditions and historical production data and through the steps of construction, matching, optimization and recommendation.
5. The tobacco shred making link intelligent control system according to claim 1, wherein: the self-learning unit improves the model prediction precision through self-learning according to the continuous accumulation of historical production data.
6. The tobacco shred making link intelligent control system according to claim 1, wherein: the working procedure section comprises a loosening and moisture regaining section, a leaf moistening and feeding section and a hot air leaf moistening section.
7. The tobacco shred making link intelligent control system according to claim 1, wherein: the intelligent control model of the intelligent control unit is provided with a feedback loop, and the feedback loop is connected with a monitoring sensor.
8. The intelligent control method for the tobacco shred making link is characterized by comprising the following steps: the method is applied to the tobacco shred making link intelligent control system as claimed in any one of claims 1 to 7, and comprises the following steps:
step one, data acquisition
The data source is as follows: corresponding historical production data in the information platform;
step two, establishing a prediction model
Establishing a nested production parameter prediction model by stages according to the process sections by combining a multiple regression analysis algorithm and a neural network algorithm, and predicting the production parameters of each process section;
step three, establishing a control model
Establishing an intelligent control model by stages according to the process sections through a neural network algorithm based on the predicted values in the second step, predicting control parameters of each process section, and intelligently controlling equipment of each process section based on the predicted values of the control parameters;
step four, mutually pushing key parameters
Establishing a parameter mutual-pushing model based on the predicted value in the step two and the historical batch parameters through a parameter relation, and performing prediction mutual-pushing on key parameters;
the production correctness of the key parameters can be verified through a parameter mutual-deduction model, and abnormal key parameters can be searched through parameter reverse deduction;
step five, controlling parameters to recommend in real time
The method adopts a similarity matching algorithm, realizes intelligent real-time recommendation of the control parameters of the whole process of the tobacco making process through association and integration of the conditions of the cigarette production process and historical production data and through the steps of construction, matching, optimization and recommendation;
step six, exceeding the standard early warning
Based on the upper and lower limits of the parameter standard, the predicted values of the production parameters and the recommended values of the control parameters are automatically judged, and the parameters exceeding the process range are alarmed;
step seven, self-learning
According to the continuous accumulation of historical production data, the model is continuously optimized through self-learning, and the model prediction precision is improved;
step eight, trend monitoring and historical data query
Collecting the prediction data of the current time period into a prediction curve for production trend monitoring;
and summarizing the historical prediction data and the actual parameter data into a data curve, and directly consulting when a corresponding batch needs to be queried.
9. The intelligent control method for the cigarette tobacco shred making link according to claim 8, wherein the intelligent control method comprises the following steps: in the third step:
the intelligent control model is connected with the feedback loop, optimizes control parameters in real time according to feedback data, and achieves intelligent and accurate control.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010662329.XA CN112021626B (en) | 2020-07-10 | 2020-07-10 | Intelligent control system and method for tobacco shred making link |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010662329.XA CN112021626B (en) | 2020-07-10 | 2020-07-10 | Intelligent control system and method for tobacco shred making link |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112021626A CN112021626A (en) | 2020-12-04 |
| CN112021626B true CN112021626B (en) | 2021-08-17 |
Family
ID=73579039
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010662329.XA Active CN112021626B (en) | 2020-07-10 | 2020-07-10 | Intelligent control system and method for tobacco shred making link |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112021626B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112465274B (en) * | 2020-12-30 | 2024-03-29 | 镇江龙源铝业有限公司 | Intelligent management system for production workshop |
| CN112884212A (en) * | 2021-02-02 | 2021-06-01 | 陕西中烟工业有限责任公司 | Cigarette single gram weight deviation analysis and prediction method |
| CN113837608A (en) * | 2021-09-24 | 2021-12-24 | 浙江中烟工业有限责任公司 | Tobacco throwing management and control and early warning platform based on big data analysis |
| CN114304709B (en) * | 2021-12-30 | 2023-03-31 | 云南烟叶复烤有限责任公司 | Automatic control method, storage medium and system for operation parameters of threshing and redrying equipment |
| CN114403486B (en) * | 2022-02-17 | 2022-11-22 | 四川大学 | Intelligent control method of airflow type cut-tobacco drier based on local peak value coding circulation network |
| CN114642268B (en) * | 2022-03-31 | 2023-05-16 | 张家口卷烟厂有限责任公司 | Intelligent control system and method for drying silk based on volatile moisture content |
| CN115167552B (en) * | 2022-06-28 | 2023-09-26 | 张家口卷烟厂有限责任公司 | Automatic control method for optimizing charging circulating air temperature based on response surface method |
| CN116382219B (en) * | 2023-05-16 | 2023-08-11 | 苏州海卓伺服驱动技术有限公司 | Motor production process optimization method and system based on online measurement technology |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1605958A (en) * | 2004-11-16 | 2005-04-13 | 冶金自动化研究设计院 | Combined modeling method and system for complex industrial process |
| CN103473438A (en) * | 2013-08-15 | 2013-12-25 | 国家电网公司 | Method for optimizing and correcting wind power prediction models |
| CN108720069A (en) * | 2018-06-20 | 2018-11-02 | 福建中烟工业有限责任公司 | The method of dry pipe tobacco and the device of prediction outlet moisture content of cut tobaccos |
| CN109674080A (en) * | 2019-03-07 | 2019-04-26 | 山东中烟工业有限责任公司 | Tobacco leaf conditioning amount of water prediction technique, storage medium and terminal device |
| CN110101106A (en) * | 2019-05-31 | 2019-08-09 | 杭州安脉盛智能技术有限公司 | Resurgence humidification humidity control method and system based on fuzzy feedforward feedback algorithm |
| CN110533082A (en) * | 2019-08-09 | 2019-12-03 | 东北大学秦皇岛分校 | A kind of sintering, mixing and water adding control method based on dual model collaborative forecasting |
| CN111144667A (en) * | 2020-01-16 | 2020-05-12 | 红云红河烟草(集团)有限责任公司 | Tobacco conditioner discharged material water content prediction method based on gradient lifting tree |
| CN111191854A (en) * | 2020-01-10 | 2020-05-22 | 上海积成能源科技有限公司 | Photovoltaic power generation prediction model and method based on linear regression and neural network |
| CN111184246A (en) * | 2018-11-14 | 2020-05-22 | 厦门邑通软件科技有限公司 | Method and system for controlling moisture content of cut tobacco drying inlet |
| CN111248485A (en) * | 2018-11-30 | 2020-06-09 | 湖南中烟工业有限责任公司 | Method for online adjusting perfuming moisture of mixed silk |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE2135637C3 (en) * | 1971-07-16 | 1980-05-29 | Hauni-Werke Koerber & Co Kg, 2050 Hamburg | Method and device for adding an admixture to tobacco |
| CN105341985A (en) * | 2015-12-10 | 2016-02-24 | 龙岩烟草工业有限责任公司 | Moisture content control method and system of inlet cut tobaccos of cut-tobacco drier |
-
2020
- 2020-07-10 CN CN202010662329.XA patent/CN112021626B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1605958A (en) * | 2004-11-16 | 2005-04-13 | 冶金自动化研究设计院 | Combined modeling method and system for complex industrial process |
| CN103473438A (en) * | 2013-08-15 | 2013-12-25 | 国家电网公司 | Method for optimizing and correcting wind power prediction models |
| CN108720069A (en) * | 2018-06-20 | 2018-11-02 | 福建中烟工业有限责任公司 | The method of dry pipe tobacco and the device of prediction outlet moisture content of cut tobaccos |
| CN111184246A (en) * | 2018-11-14 | 2020-05-22 | 厦门邑通软件科技有限公司 | Method and system for controlling moisture content of cut tobacco drying inlet |
| CN111248485A (en) * | 2018-11-30 | 2020-06-09 | 湖南中烟工业有限责任公司 | Method for online adjusting perfuming moisture of mixed silk |
| CN109674080A (en) * | 2019-03-07 | 2019-04-26 | 山东中烟工业有限责任公司 | Tobacco leaf conditioning amount of water prediction technique, storage medium and terminal device |
| CN110101106A (en) * | 2019-05-31 | 2019-08-09 | 杭州安脉盛智能技术有限公司 | Resurgence humidification humidity control method and system based on fuzzy feedforward feedback algorithm |
| CN110533082A (en) * | 2019-08-09 | 2019-12-03 | 东北大学秦皇岛分校 | A kind of sintering, mixing and water adding control method based on dual model collaborative forecasting |
| CN111191854A (en) * | 2020-01-10 | 2020-05-22 | 上海积成能源科技有限公司 | Photovoltaic power generation prediction model and method based on linear regression and neural network |
| CN111144667A (en) * | 2020-01-16 | 2020-05-12 | 红云红河烟草(集团)有限责任公司 | Tobacco conditioner discharged material water content prediction method based on gradient lifting tree |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112021626A (en) | 2020-12-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112021626B (en) | Intelligent control system and method for tobacco shred making link | |
| CN113017132A (en) | Cut tobacco quality optimization method based on cut tobacco dryer process parameter prediction | |
| CN116341771B (en) | Intelligent optimization method and system for low-temperature-resistant cable production process | |
| Ko et al. | Assessment of achievable PI control performance for linear processes with dead time | |
| CN112034791B (en) | Intelligent control system and method for sheet cut-tobacco drier | |
| CN112021641B (en) | On-line moisture meter calibration system for tobacco shred making link | |
| CN111144667A (en) | Tobacco conditioner discharged material water content prediction method based on gradient lifting tree | |
| CN100585303C (en) | Energy control method of cyclone compressor modularizing wind cold-hot pump unit | |
| CN108652066A (en) | The water feeding method of loosening and gaining moisture process and the device for predicting the process amount of water | |
| CN112327960B (en) | Intelligent control system for loosening and dampening equipment | |
| CN105996107A (en) | Cut tobacco quality control method and system based on sheet cut tobacco drying | |
| CN105468850B (en) | Electronic product degradation trend prediction technique based on more residual error regression forecasting algorithms | |
| CN111103854A (en) | System and method for improving production stability of tobacco cut-tobacco drier | |
| CN105996106A (en) | Cut tobacco quality control method and system based on airflow drying | |
| CN114115393A (en) | Method for controlling moisture and temperature at outlet of cut tobacco dryer for sheet cut tobacco making line | |
| CN113885310B (en) | Intelligent control system for vacuum dry pump test | |
| CN101995845B (en) | Field programmable gate array (FPGA)-based auto-leveling control system and method | |
| CN113303489A (en) | Method for accurately controlling moisture of tobacco leaves in tobacco shred making process | |
| CN101275811A (en) | Intelligent control method of clinker grid type cooling machine cooling procedure | |
| CN106072743B (en) | A kind of stem method of quality control and system based on stem drying | |
| CN116757354A (en) | Tobacco redrying section key parameter screening method based on multilayer perceptron | |
| CN115356995A (en) | Tobacco drying process heat energy analysis system based on industrial big data | |
| CN115933360A (en) | Intelligent frequency conversion module temperature shunt control system and method based on big data | |
| CN116019248A (en) | Loose leaf-wetting outlet material temperature control method | |
| CN115167552B (en) | Automatic control method for optimizing charging circulating air temperature based on response surface method |
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 |