Tobacco production control method, device, storage medium and computer program product

Through the steady-state water addition flow prediction model, the water addition flow rate is controlled during the tobacco production process, which solves the problem of unstable water content of tobacco on the material head and improves the quality of tobacco products.

CN118502378BActive Publication Date: 2024-11-15ALIBABA CLOUD FEITIAN (HANGZHOU) CLOUD COMPUTING TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410972072.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-19
Publication Date
2024-11-15
Estimated Expiration
2044-07-19

AI Technical Summary

Technical Problem

The moisture content of the tobacco strips in the tobacco strips is unstable during the tobacco production process, resulting in downgrade or loss of materials and affecting product quality.

Method used

The steady-state water flow prediction model is adopted to predict and generate a water flow control curve based on the current environmental variable data and historical production data to ensure the stability of the water flow during the start-up phase, thereby improving the water content stability of tobacco products.

Benefits of technology

By accurately controlling the water addition flow during the start-up stage, the water content of the tobacco is stable, the product quality of tobacco is improved, and the water addition flow is assisted in the normal operation stage to further improve the product quality.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118502378B_ABST
    Figure CN118502378B_ABST
Patent Text Reader

Abstract

The embodiments of the present application provide a tobacco production control method, device, storage medium and computer program product. In the embodiments of the present application, with the assistance of a steady-state water addition flow prediction model, the water addition flow rate during the startup phase can be accurately controlled, effectively improving the moisture content stability of tobacco products produced during the startup phase, that is, ensuring the stability of the moisture content of the tobacco at the start of the material, and improving the product quality of tobacco production. Further optionally, with the assistance of a tobacco leaf moisture content prediction model, the water addition flow rate of the tobacco production system during the normal operation phase can be accurately controlled, effectively improving the moisture content stability of tobacco products produced during the normal operation phase, and improving the product quality of tobacco production.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of production control technology, and in particular to a tobacco production control method, device, storage medium and computer program product. Background Art

[0002] Tobacco production is a typical multi-link batch production process. The tobacco leaves are processed through multiple production processes such as loosening and rehydration, moistening and adding materials, cutting and drying to obtain the final tobacco products. Usually, each production process is produced in batches, that is, batch production is carried out according to a certain production volume. Since external disturbances such as the moisture content, water absorption, ambient temperature and humidity of the incoming tobacco leaves have a great influence on the quality of tobacco, it is necessary to automatically adjust control variables such as water addition flow rate during the tobacco production process to ensure the moisture stability of the processed tobacco products and ensure the stability of the product quality of tobacco production.

[0003] Taking the leaf moistening and feeding process as an example, there are the following challenges in production control during the tobacco production process: the moisture content of the tobacco leaves produced in the startup phase of the tobacco production system within the same batch fluctuates greatly, that is, the moisture content of the tobacco leaves is unstable, which can easily cause material degradation or loss, affecting the product quality of tobacco production. Summary of the invention

[0004] Multiple aspects of the present application provide a tobacco production control method, device, storage medium and computer program product to solve the technical problem that the moisture content of tobacco is unstable, which easily causes material degradation or loss and affects the product quality of tobacco production.

[0005] The embodiment of the present application provides a tobacco production control method, including: obtaining historical production data of historical batches of tobacco leaves processed by a current production process and its upstream production process, and obtaining current environmental variable data; inputting the current environmental variable data and the historical production data into a steady-state water addition flow prediction model to predict the steady-state water addition flow of the current batch of tobacco leaves processed by the current production process during the normal operation stage of the tobacco production system; generating a water addition flow control curve for the startup stage of the tobacco production system, the water addition flow at the end of the startup stage on the water addition flow control curve being the steady-state water addition flow; and performing water addition control according to the water addition flow control curve during the startup stage.

[0006] An embodiment of the present application provides an electronic device, comprising: a memory and a processor; the memory is used to store a computer program; the processor is coupled to the memory and is used to execute the computer program to execute the steps in the tobacco production control method.

[0007] An embodiment of the present application provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor is enabled to execute the steps in the tobacco production control method.

[0008] An embodiment of the present application provides a computer program product, including a computer program / instruction. When the computer program / instruction is executed by a processor, the processor is enabled to execute the steps in the tobacco production control method.

[0009] The technical solution provided by the embodiment of the present application uses a steady-state water addition flow prediction model based on the current environmental variable data of the production site and the historical production data of the current production process and its previous production process for processing historical batches of tobacco leaves, to predict the steady-state water addition flow of the current production process for processing the current batch of tobacco leaves in the normal operation stage of the tobacco production system; to make the water addition flow of the tobacco production system at the end of the startup stage the steady-state water addition flow as the goal, to generate a water addition flow control curve for the startup stage; to perform water addition control according to the water addition flow control curve during the startup stage. Then, with the assistance of the steady-state water addition flow prediction model, the water addition flow in the startup stage can be accurately controlled, and the water content stability of the tobacco products produced in the startup stage can be effectively improved, that is, the water content of the material head tobacco is stable, and the product quality of tobacco production can be improved. Further, optionally, with the assistance of the tobacco leaf moisture content prediction model, the water addition flow of the tobacco production system in the normal operation stage can be accurately controlled, and the water content stability of the tobacco products produced in the normal operation stage can be effectively improved, and the product quality of tobacco production can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0011] Figure 1 A flowchart of a tobacco production control method provided in an embodiment of the present application;

[0012] Figure 2 A flowchart of another tobacco production control method provided in an embodiment of the present application;

[0013] Figure 3 A flowchart of another tobacco production control method provided in an embodiment of the present application;

[0014] Figure 4 A schematic diagram of the structure of a tobacco production control device provided in an embodiment of the present application;

[0015] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0017] In the embodiments of the present application, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the access relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist at the same time, and B exists alone, where A and B may be singular or plural. In the text description of the present application, the character " / " generally indicates that the previous and next associated objects are in an "or" relationship. In addition, in the embodiments of the present application, "first", "second", "third", etc. are only used to distinguish the contents of different objects and have no other special meanings.

[0018] First, the terms involved in the embodiments of the present application are introduced:

[0019] The water addition flow rate can be understood as the flow rate when the tobacco production system adds water when processing tobacco leaves. The water addition flow rate is, for example, 10 cubic meters per second.

[0020] The moisture content of tobacco leaves, also known as the moisture content of tobacco leaves, refers to the proportion of the mass of water in the tobacco leaves to the total mass of the tobacco leaves, usually expressed as a percentage.

[0021] Tobacco flow rate refers to the instantaneous flow rate of tobacco leaves through specific areas or equipment (such as conveyor belts, pipes, feeders, etc.) during the tobacco processing process. The real-time tobacco flow rate can reflect the total amount of tobacco leaves that need to be processed per unit time.

[0022] The following specific embodiments are used to describe in detail the technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The following is a detailed description of the technical solutions provided by each embodiment of the present application in conjunction with the accompanying drawings.

[0023] Figure 1 This is a flow chart of a tobacco production control method provided in an embodiment of the present application. Figure 1 , the method may include the following steps:

[0024] 101. Obtain historical production data of historical batches of tobacco leaves processed by the current production process and its upstream production process, and obtain current environmental variable data.

[0025] 102. Input the current environmental variable data and historical production data into the steady-state water addition flow prediction model to predict the steady-state water addition flow of the current batch of tobacco leaves in the current production process during the normal operation stage of the tobacco production system.

[0026] 103. Generate a water addition flow control curve for the start-up phase of the tobacco production system, wherein the water addition flow at the end of the start-up phase on the water addition flow control curve is a steady-state water addition flow.

[0027] 104. During the startup phase, water addition control is performed according to the water addition flow control curve.

[0028] Specifically, the tobacco production system is a manufacturing system for producing tobacco, which usually includes multiple mechanical equipment, such as but not limited to: a shredder, a tobacco leaf rehumidification machine, a leaf moistening and feeding equipment, a dryer, etc. In actual applications, the tobacco production process of the tobacco production system may be different. For example, the tobacco production process includes: tobacco leaf unpacking, loosening and rehumidification, leaf moistening and feeding, shredding, shredding, packaging and other production processes, but is not limited to this.

[0029] In this embodiment, a steady-state water addition flow prediction model can be trained. The steady-state water addition flow prediction model is used to predict the steady-state water addition flow. Excluding the special circumstances of the startup phase and shutdown, the steady-state water addition flow can be understood as the average water addition flow after the current batch of tobacco leaves reaches the stable operation phase (that is, the normal operation phase) during the production process such as loosening and rehumidification, leaf moistening and feeding. The steady-state water addition flow can be used as the water addition flow required by the tobacco production system for the production of the current batch of tobacco leaves at the end of the startup phase, so that the tobacco production can quickly enter the stable operation phase and avoid moisture fluctuations during the startup phase.

[0030] In actual applications, different cigarette brands will use different grades of tobacco leaves for manufacturing. The grade can be regarded as the category information of tobacco leaves, and the grade of tobacco leaves can be flexibly set as needed. In actual applications, different grades of tobacco leaves can share the same steady-state water addition flow prediction model. Further, optionally, a corresponding steady-state water addition flow prediction model is trained for each grade of tobacco leaves, that is, different grades of tobacco leaves correspond to different steady-state water addition flow prediction models, and each grade of tobacco leaves has its own exclusive steady-state water addition flow prediction model. This can effectively deal with the problem of "large fluctuations in the moisture content of tobacco produced in different batches, affecting the stability of product quality."

[0031] In practical applications, a corresponding steady-state water addition flow prediction model can be established for any production process, such as loosening and rehydration or moistening and adding materials. In the model training stage of the steady-state water addition flow prediction model, it is necessary to first collect the training data set involved in the model training. The training data set involved in training the steady-state water addition flow prediction model can be referred to as the first training data set. The first training data set includes multiple training data, each of which includes, for example, but is not limited to: historical production data of historical batches of tobacco leaves of the same brand processed by the current production process (for example, moistening and adding materials), historical production data of historical batches of tobacco leaves processed by the previous process before the current production process (for example, loosening and rehydration), and current environmental variable data and annotation results at the production site. It is worth noting that the historical production data involved in model training are production data actually generated in the historical time period. In practical applications, the number of batches of historical batches of tobacco leaves involved in model training is not limited, for example, 1,000 historical batches of tobacco leaves.

[0032] Optionally, the historical production data corresponding to the current production process include, but are not limited to: the water flow rate, tobacco flow rate, steam volume, steam temperature of the same brand of historical batches of tobacco leaves processed by the current production process, and the moisture content of the tobacco leaves of the historical batches of tobacco leaves after being processed by the current production process. The moisture content of the tobacco leaves of the historical batches of tobacco leaves after being processed by the current production process can be understood as the moisture content of the exported tobacco leaves corresponding to the current production process, etc. The moisture content of the exported tobacco leaves corresponding to the current production process is also the moisture content of the tobacco leaves of the historical batches of tobacco leaves that have been processed by the current production process and have flowed out of the current production process.

[0033] Optionally, the historical production data of the historical batches of tobacco leaves processed by the front-end process include, but are not limited to: the water addition flow rate, tobacco leaf flow rate, steam volume, steam temperature, and the moisture content of the historical batches of tobacco leaves after being processed by the front-end process. The moisture content of the tobacco leaves of the historical batches of tobacco leaves after being processed by the front-end process can be understood as the moisture content of the export tobacco leaves corresponding to the front-end process, etc. The moisture content of the export tobacco leaves corresponding to the front-end process is also the moisture content of the tobacco leaves of the historical batches of tobacco leaves that flow out of the front-end process after being processed by the front-end process.

[0034] Optionally, the current environmental variable data of the production site involved in the model training refers to the environmental data when the current production process processes historical batches of tobacco leaves. The current environmental variable data includes, but is not limited to: ambient temperature, ambient humidity, and production date (for example, a certain year and month).

[0035] Optionally, the marking result includes, for example: a steady-state water addition flow rate of the tobacco production system. The steady-state water addition flow rate is the average water addition flow rate of historical batches of tobacco leaves processed by the current production process after reaching a stable operation stage during the production process.

[0036] Optionally, in the training data preparation stage, data preprocessing can be performed on the historical production data collected from the production site, and model training can be performed based on the preprocessed historical production data. Data preprocessing includes, but is not limited to, data cleaning, normalization, feature engineering, etc. Abnormal data processing includes, but is not limited to, outlier removal and missing value filling.

[0037] In practical applications, there is no restriction on the model structure of the steady-state water addition flow prediction model, including but not limited to: Deep Neural Networks (DNNs), Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), Long Short-Term Memory (LSTM) or Transformer network. Transformer network is a neural network architecture with a self-attention mechanism.

[0038] In practical applications, the training data set involved in model training is used for model training to obtain a trained steady-state water addition flow prediction model. After each model training, the steady-state water addition flow prediction model after the model parameters are adjusted is used as the next steady-state water addition flow prediction model to be trained, and the next model training is performed using new training data until the end of training conditions are met, and the steady-state water addition flow prediction model obtained by the last training is used as the final steady-state water addition flow prediction model. Among them, the end of training conditions can be that the model training times reach a specified number of times, or the steady-state water addition flow prediction model converges, and no specific restrictions are made here.

[0039] Specifically, in the model training stage, the historical production data of the historical batches of tobacco leaves processed by the current production process in the training data, the historical production data of the historical batches of tobacco leaves processed by the previous process before the current production process, and the current environmental variable data of the production site can be input into the steady-state water addition flow prediction model to be trained to obtain the predicted steady-state water addition flow; the loss value is determined according to the predicted steady-state water addition flow and the steady-state water addition flow in the labeled results, and the model parameters of the steady-state water addition flow prediction model are adjusted according to the loss value. The above model training is iterated until the training end condition is reached.

[0040] In practical applications, any loss function can be used to determine the loss value. Any loss function includes, but is not limited to, mean square error loss function, L2 loss function (also known as Euclidean distance), L1 loss function (Manhattan distance), cross entropy loss function, softmax loss function, and Focal loss function, which is mainly used to solve the problem of imbalance between difficult and easy samples.

[0041] It is worth noting that when using the loss function to determine the loss value, the steady-state water addition flow rate in the labeled result is input into the loss function as the true value, and the predicted steady-state water addition flow rate is input into the loss function as the predicted value. The loss function calculates the loss value based on the input true value and predicted value. For more information about calculating the loss value using the loss function, see Related Technologies.

[0042] In practical applications, the trained steady-state water addition flow prediction model is published to assist water addition control during the startup phase of the tobacco production system. The following introduces water addition control using the steady-state water addition flow prediction model.

[0043] First, the current environmental variable data of the production site and the historical production data of the same brand of historical batches of tobacco leaves processed by the current production process and its previous production process are obtained. The previous production process may be one or more production processes before the current production process, and there is no restriction on this.

[0044] Optionally, the current environmental variable data of the production site includes, for example, but is not limited to: current ambient temperature, and / or current ambient humidity, production date.

[0045] Optionally, the historical production data corresponding to the current production process include, but are not limited to: the water flow rate, tobacco flow rate, steam volume, steam temperature of historical batches of tobacco leaves of the same brand processed by the current production process during the processing, and the moisture content of the historical batches of tobacco leaves after processing by the current production process.

[0046] Optionally, the historical production data of historical batches of tobacco leaves processed by the previous process include, but are not limited to: the water addition flow rate, tobacco leaf flow rate, steam volume, steam temperature during the process of processing historical batches of tobacco leaves by the previous process, and the moisture content of historical batches of tobacco leaves after processing by the previous process.

[0047] In practical applications, the historical batches of tobacco leaves may be one or more batches of tobacco leaves whose production sequence is before the current batch of tobacco leaves, and the specific quantity is not limited.

[0048] Next, the current environmental variable data of the production site and the historical production data of the current production process and its upstream production process for processing historical batches of tobacco leaves are input into the steady-state water addition flow prediction model for predictive processing to obtain the steady-state water addition flow of the current production process processing the current batch of tobacco leaves during the normal operation stage of the tobacco production system.

[0049] It is worth noting that in the inference stage of the steady-state water addition flow prediction model, there is no restriction on the batch number of historical batches of tobacco leaves, for example, the three most recent batches of tobacco leaves.

[0050] Finally, the steady-state water addition flow predicted by the steady-state water addition flow prediction model is used to control water addition in the startup phase. Specifically, first, a water addition flow control curve for the startup phase is generated. The water addition flow control curve includes the water addition flow at each time point in the startup phase, that is, the water addition flow control curve includes the water addition flow at each time point in the time period from the start of the startup phase to the end of the startup phase. Among them, the water addition flow at the end of the startup phase in the water addition flow control curve is the steady-state water addition flow. Then, water addition control is performed in accordance with the water addition flow control curve during the startup phase.

[0051] In practical applications, there is no restriction on the curve parameters of the water addition flow control curve, and the curve parameters include, but are not limited to, the curve slope and the curve shape. The curve shape describes the overall appearance or form of the curve, which involves the degree of curvature, direction, symmetry, and possible special points (such as extreme points and inflection points) of the curve.

[0052] Optionally, optimal curve parameters can be obtained, including curve slope and / or curve linearity; based on the optimal curve parameters, a water addition flow control curve for the startup phase is generated with the goal of making the water addition flow of the tobacco production system at the end of the startup phase a steady-state water addition flow.

[0053] In practical applications, data mining can be used to obtain optimal curve parameters so that the moisture content of tobacco leaves in the startup stage fluctuates less. Specifically, data mining is performed on the moisture content of tobacco leaves in the startup stage over the past period of time to find the water addition flow control curve corresponding to the smaller fluctuation of tobacco leaf moisture content, and statistical analysis is performed on the curve slope and / or curve linearity of the water addition flow control curve corresponding to the smaller fluctuation of tobacco leaf moisture content to obtain the optimal curve parameters. The optimal curve parameters can be understood as the empirical values ​​obtained by mining a large amount of data. The water addition flow control curve that meets the optimal curve parameters can make the tobacco leaf moisture content in the startup stage fluctuate less, and the tobacco leaf moisture content in the startup stage meets the product quality requirements.

[0054] Optionally, a control instruction may be sent to an actuator responsible for adding water in the tobacco production system, the control instruction instructing to add water according to a water addition flow control curve during the startup phase.

[0055] The technical solution provided by the embodiment of the present application uses a steady-state water addition flow prediction model based on the current environmental variable data of the production site and the historical production data of the current production process and its previous production process for processing historical batches of tobacco leaves, to predict the steady-state water addition flow of the current production process processing the current batch of tobacco leaves in the normal operation stage of the tobacco production system; with the goal of making the water addition flow of the tobacco production system at the end of the startup stage the steady-state water addition flow, a water addition flow control curve for the startup stage is generated; and water addition control is performed according to the water addition flow control curve during the startup stage. In addition, the water addition flow during the current production process processing the current batch of tobacco leaves can be accurately controlled, effectively improving the water content stability of the tobacco products produced during the startup stage, that is, ensuring the stability of the water content of the tobacco in the head, and improving the product quality of tobacco production.

[0056] In some optional embodiments, a tobacco leaf moisture content prediction model can also be trained to assist in water addition control during the normal operation phase of the tobacco production system. The startup phase of the tobacco production system is the starting phase of the entire production process, and the normal operation phase begins after the startup phase.

[0057] In practical applications, different brands of tobacco leaves can share the same tobacco leaf moisture content prediction model. Further, optionally, a corresponding tobacco leaf moisture content prediction model is trained for each brand of tobacco leaves, that is, different brands of tobacco leaves correspond to different tobacco leaf moisture content prediction models, and each brand of tobacco leaves has an exclusive tobacco leaf moisture content prediction model. The tobacco leaf moisture content prediction model can predict the tobacco leaf moisture content in future time periods based on real-time production data before the current moment. According to the predicted tobacco leaf moisture content in the future time period, the water addition flow rate can be adjusted in advance to reduce the moisture fluctuation of the final tobacco product and ensure the product quality of tobacco production.

[0058] In practical applications, a corresponding tobacco leaf moisture content prediction model can be established for any production process, such as loosening and rehumidification or leaf moistening and feeding. In the model training stage of the tobacco leaf moisture content prediction model, it is necessary to first collect the training data set involved in the model training. The training data set involved in training the tobacco leaf moisture content prediction model can be referred to as the second training data set. The second training data set training data set includes multiple training data, each training data includes, for example, but is not limited to: historical production data of historical batches of tobacco leaves of the same brand processed by the current production process (such as leaf moistening and feeding), current environmental variable data and annotation results of the production site. It can be understood that the historical production data collected in the training stage are all production data actually generated in the production process.

[0059] Optionally, the historical production data corresponding to the current production process include, but are not limited to: the time series data of the water addition flow of historical batches of tobacco leaves processed by the current production process, the time series data of the tobacco flow of historical batches of tobacco leaves, and the time series data of the moisture content of historical batches of tobacco leaves before being processed by the current production process. The water addition flow time series data includes the water addition flow at each time point within a period of time, which is a set of time series data; the tobacco flow time series data includes the tobacco flow at each time point within a period of time, which is a set of time series data; the tobacco moisture content time series data includes the tobacco moisture content at each time point within a period of time, which is a set of time series data. The tobacco moisture content in the tobacco moisture content time series data of the historical batches of tobacco leaves before being processed by the current production process can be understood as the inlet tobacco moisture content. The inlet tobacco moisture content refers to the moisture content of the tobacco leaves detected in real time by infrared detectors and other equipment before the tobacco leaves enter the current processing process equipment.

[0060] Optionally, the current environmental variable data of the production site involved in the model training refers to the environmental data when the current production process processes historical batches of tobacco leaves. The current environmental variable data includes, but is not limited to: ambient temperature and ambient humidity.

[0061] Optionally, the labeling result includes, for example, the time series data of the moisture content of tobacco leaves in historical batches after being processed in the current production process. The labeling result includes that the moisture content of tobacco leaves in the time series data of the moisture content of tobacco leaves can be understood as the moisture content of exported tobacco leaves, and the moisture content of exported tobacco leaves can be understood as the moisture content detected in real time by infrared detectors and other equipment at the outlet of the equipment after the tobacco leaves are processed in the current process.

[0062] Optionally, in the training data preparation stage, data preprocessing can be performed on the historical production data collected from the production site, and model training can be performed based on the preprocessed historical production data. Data preprocessing includes, but is not limited to, data cleaning, normalization, feature engineering, etc. Abnormal data processing includes, but is not limited to, outlier removal and missing value filling.

[0063] In practical applications, there is no restriction on the model structure of the tobacco leaf moisture content prediction model, including but not limited to: deep neural network, convolutional neural network, recurrent neural network, long short-term memory network or Transformer network, state space model, finite step response model.

[0064] In practical applications, the training data set involved in model training is used for model training to obtain a trained tobacco leaf moisture content prediction model. After each model training, the tobacco leaf moisture content prediction model after the model parameters are adjusted is used as the next tobacco leaf moisture content prediction model to be trained, and the next model training is performed using new training data until the end of training conditions are met, and the tobacco leaf moisture content prediction model obtained by the last training is used as the final tobacco leaf moisture content prediction model. Among them, the end of training conditions can be that the model training times reach a specified number of times, or the tobacco leaf moisture content prediction model converges, and no specific restrictions are made here.

[0065] Specifically, in the model training stage, the historical production data corresponding to the current production process in the training data and the current environmental variable data of the production site can be input into the tobacco moisture content prediction model to be trained to obtain the tobacco moisture content time series data predicted in the future time period; the loss value is determined according to the predicted tobacco moisture content time series data and the tobacco moisture content time series data in the annotation results, and the model parameters of the tobacco moisture content prediction model are adjusted according to the loss value. The above model training is iterated until the training end condition is reached.

[0066] In practical applications, any loss function can be used to determine the loss value. Any loss function includes, but is not limited to, mean square error loss function, L2 loss function (also known as Euclidean distance), L1 loss function (Manhattan distance), cross entropy loss function, softmax loss function, and Focalloss loss function, which is mainly used to solve the problem of imbalance between difficult and easy samples.

[0067] It is worth noting that when using the loss function to determine the loss value, the tobacco leaf moisture content time series data column in the annotation result is input into the loss function as the true value, and the predicted tobacco leaf moisture content time series data is input into the loss function as the predicted value. The loss function calculates the loss value based on the input true value and predicted value. For more information about the loss function to calculate the loss value, please refer to the related technology.

[0068] Combine the following Figure 2 The control process during normal operation is introduced. Figure 2 This is a flow chart of another tobacco production control method provided in an embodiment of the present application. Figure 2 , the method may include the following steps:

[0069] 201. Obtain historical production data of historical batches of tobacco leaves processed by the current production process and its upstream production process, and obtain current environmental variable data.

[0070] 202. Input the current environmental variable data and historical production data into the steady-state water addition flow prediction model to predict the steady-state water addition flow of the current batch of tobacco leaves in the current production process during the normal operation stage of the tobacco production system.

[0071] 203. Generate a water addition flow control curve for the startup phase of the tobacco production system, wherein the water addition flow at the end of the startup phase on the water addition flow control curve is a steady-state water addition flow.

[0072] 204. During the startup phase, water addition is controlled according to the water addition flow control curve.

[0073] For the implementation of step 201 to step 204, reference may be made to the relevant contents of the aforementioned embodiment, which will not be described in detail here.

[0074] 205. After the tobacco production system enters the normal operation stage, the current environmental variable data and the real-time production data collected before the current moment, as well as the water addition flow at each sampling time point in the future time period, are input into the tobacco leaf moisture content prediction model to predict the tobacco leaf moisture content prediction time series data in the future time period. The tobacco leaf moisture content prediction time series data includes the predicted tobacco leaf moisture content at each sampling time point in the future time period.

[0075] 206. Construct an objective function based on the deviation information between the predicted time series data of tobacco leaf moisture content and the set tobacco leaf moisture content, and the change information between the water addition flow rates at adjacent sampling time points.

[0076] 207. Taking the water addition flow rate at each sampling time point in the future time period as the optimization variable, solving the objective function, and obtaining the optimized water addition flow rate at each sampling time point in the future time period.

[0077] 208. During the period from the current moment to the first sampling time point in the future time period, water addition control is performed according to the optimized water addition flow rate at the first sampling time point.

[0078] 209. If the current moment reaches the first sampling time point, the first sampling time point is used as the new current moment, and the above steps are iteratively executed until the normal operation phase ends.

[0079] In practical applications, the real-time production data collected before the current moment include, but are not limited to: the time series data of the water addition flow rate of the current batch of tobacco leaves in the current production process, the time series data of the tobacco flow rate of the current batch of tobacco leaves, and the time series data of the moisture content of the current batch of tobacco leaves before being processed by the current production process. It can be understood that the real-time production data is the data actually generated during the production process.

[0080] In practical applications, the real-time production data may be the real-time production data within a specified time period before the current moment, and the specified time period may be flexibly set as needed. For example, the real-time production data may be the real-time production data within 60 seconds, 120 seconds, 180 seconds, or 300 seconds before the current moment.

[0081] In this embodiment, after entering the normal operation stage, the current environmental variable data and the real-time production data collected before the current moment, as well as the water addition flow at each sampling time point in the future time period, are input into the tobacco leaf moisture content prediction model to predict the tobacco leaf moisture content prediction time series data in the future time period.

[0082] In practical applications, the future time period is a period of time after the current moment, such as 60 seconds later, 120 seconds later, 180 seconds later, or 300 seconds later, etc. The future time period includes multiple sampling time points, and adjacent sampling time points are separated by a set time length. For example, adjacent sampling time points are separated by 10 seconds, and the future time period is 3 minutes in the future, then the future time period has 18 sampling time points.

[0083] In this embodiment, the water addition flow rate at each sampling time point in the future time period is used as the optimization variable, and the objective function is constructed based on the deviation information between the predicted time series data of tobacco leaf moisture content and the set tobacco leaf moisture content, as well as the change information between the water addition flow rates at adjacent sampling time points.

[0084] In practical applications, the deviation information between the predicted time series data of tobacco leaf moisture content and the set tobacco leaf moisture content is determined by the deviation information between the predicted tobacco leaf moisture content at each sampling time point in the predicted time series data of tobacco leaf moisture content and the set tobacco leaf moisture content. For example, the deviation information between the predicted tobacco leaf moisture content at each sampling time point and the set tobacco leaf moisture content is weighted summed, averaged, or accumulated to obtain the deviation information between the predicted time series data of tobacco leaf moisture content and the set tobacco leaf moisture content.

[0085] In actual applications, the deviation information between the predicted tobacco leaf moisture content at each sampling time point and the set tobacco leaf moisture content reflects the difference information between the predicted tobacco leaf moisture content at each sampling time point and the set tobacco leaf moisture content. The deviation information between the predicted tobacco leaf moisture content at each sampling time point and the set tobacco leaf moisture content includes, but is not limited to: the difference between the predicted tobacco leaf moisture content at each sampling time point and the set tobacco leaf moisture content, the square of the difference, the square difference, the absolute difference, the percentage difference, etc.

[0086] In actual applications, the change information between the water addition flow rates at adjacent sampling time points reflects the difference information between the water addition flow rates at adjacent sampling time points. The change information between the water addition flow rates at adjacent sampling time points includes, but is not limited to: the difference between the water addition flow rates at adjacent sampling time points, the square of the difference, the square difference, the absolute difference, the percentage difference, etc.

[0087] In practical applications, the deviation information between the predicted time series data of tobacco leaf moisture content and the set tobacco leaf moisture content, as well as the change information between the water addition flow rates at adjacent sampling time points can be weighted summed, averaged or accumulated to construct an objective function, but is not limited to this.

[0088] As an example, the objective function is formulated as follows:

[0089] (1)

[0090] Among them, w 1 and w 2 are weight coefficients that can be set flexibly on demand, w 1 and w 2 The sum obtained by adding can be 1; the value of n is the number of sampling time points in the future time period. is the predicted tobacco leaf moisture content at the i-th sampling time point, Flexible setting of tobacco leaf moisture content according to demand; is the water addition flow rate at the i-th sampling time point, is the water addition flow rate at the i-1th sampling time point, and min means minimization;

[0091] in, Represents the deviation information between the predicted time series data of tobacco leaf moisture content and the set tobacco leaf moisture content.

[0092] in, It represents the cumulative sum of the change information between the water addition flow rates at adjacent sampling time points.

[0093] In practical applications, the constraints of the objective function can be set as needed. Optionally, in order to improve the optimization effect, the constraints include, but are not limited to: the predicted time series data of tobacco moisture content meets the first upper and lower limit constraints; the difference between the water addition flow rates at adjacent sampling time points meets the second upper and lower limit constraints; the water addition flow rate at each sampling time point meets the third upper and lower limit constraints.

[0094] Among them, the first upper and lower limit constraints that the tobacco leaf moisture content prediction time series data needs to meet are that the tobacco leaf moisture content at each sampling time point in the tobacco leaf moisture content prediction time series data cannot be less than the lower limit of the tobacco leaf moisture content, nor can it be greater than the upper limit of the tobacco leaf moisture content. The lower limit of the tobacco leaf moisture content is less than the upper limit of the tobacco leaf moisture content, and the lower limit and upper limit of the tobacco leaf moisture content can be flexibly set as needed.

[0095] Among them, the second upper and lower limit constraints that the difference between the water addition flow rates at adjacent sampling time points needs to satisfy are that the difference between the water addition flow rates at adjacent sampling time points cannot be greater than the difference lower limit value, nor greater than the difference upper limit value. The difference lower limit value is less than the difference upper limit value, and the difference lower limit value and the difference upper limit value can be flexibly set as needed.

[0096] Among them, the third upper and lower limit constraints that the water addition flow rate at each sampling time point needs to meet are that the water addition flow rate at the sampling time point cannot be less than the lower limit of the water addition flow rate, nor can it be greater than the upper limit of the water addition flow rate. The lower limit of the water addition flow rate is less than the upper limit of the water addition flow rate, and the lower limit of the water addition flow rate and the upper limit of the water addition flow rate can be flexibly set as needed.

[0097] In practical applications, the minimization of the objective function can be used as the optimization goal, and the objective function can be solved under the constraints of the constraints to obtain the optimized water addition flow rate at each sampling time point in the future time period to obtain the optimized water addition flow rate target value at each time step in the future time period. During the period from the current moment to the first sampling time point in the future time period, water addition control is performed according to the optimized water addition flow rate at the first sampling time point. Optionally, a control instruction can be sent to the actuator responsible for water addition in the tobacco production system, and the control instruction indicates that during the period from the current moment to the first sampling time point in the future time period, water addition control is performed according to the optimized water addition flow rate at the first sampling time point. It can be understood that during the period from the current moment to the first sampling time point, water addition control is performed according to the optimized water addition flow rate at the first sampling time point.

[0098] Further optionally, before the first sampling time point is taken as the new current moment, the tobacco leaf moisture content prediction model can also be calibrated according to the error value between the predicted tobacco leaf moisture content at the first sampling time point and the actual tobacco leaf moisture content. It can be understood that the actual tobacco leaf moisture content at the first sampling time point is obtained through detection.

[0099] In practical applications, the predicted tobacco moisture content at the first sampling time point can be input into the loss function as a predicted value, and the actual tobacco moisture content at the first sampling time point can be input into the loss function as a true value. The loss function calculates the error value based on the input true value and predicted value. With the goal of minimizing the error value, the model parameters of the tobacco moisture content prediction model are adjusted to correct the tobacco moisture content prediction model. It can be understood that correcting the tobacco moisture content prediction model based on the loss value corresponding to the previous sampling time point can enable the tobacco moisture content prediction model to predict a more accurate tobacco moisture content at the next sampling time point.

[0100] The technical solution provided in the embodiment of the present application, with the assistance of the steady-state water addition flow prediction model, can accurately control the water addition flow in the startup phase, effectively improve the moisture content stability of the tobacco products produced in the startup phase, that is, ensure the stability of the moisture content of the tobacco in the head, and improve the product quality of tobacco production. Furthermore, with the assistance of the tobacco leaf moisture content prediction model, the water addition flow of the tobacco production system in the normal operation phase can be accurately controlled, effectively improve the moisture content stability of the tobacco products produced in the normal operation phase, and improve the product quality of tobacco production.

[0101] In order to better understand the technical solution of the embodiment of the present application, Figure 3 A specific scenario implementation example is introduced.

[0102] The intelligent manufacturing platform can integrate artificial intelligence and big data technologies into traditional production lines, helping manufacturing companies to achieve the coordination of data flow, production flow and control flow, improve production efficiency, reduce production costs, and achieve autonomous and controllable intelligent manufacturing in an autonomous and controllable path. The intelligent manufacturing platform can be composed of software and / or hardware. For example, the intelligent manufacturing platform includes but is not limited to: distributed servers, conventional servers, cloud servers, virtual servers, edge servers, high performance computing (HPC) server clusters, etc.

[0103] In the tobacco leaf production scenario, the intelligent manufacturing platform can collect production data in the tobacco leaf production process, perform data preprocessing on the production data, and use the preprocessed production data for model training, and can train a steady-state water addition flow prediction model and a tobacco leaf moisture content prediction model, respectively. The intelligent manufacturing platform publishes the model, so that the steady-state water addition flow prediction model can be used to predict the steady-state water addition flow online, and the tobacco leaf moisture content prediction model can be used to predict the tobacco leaf moisture content online. In the startup phase of the tobacco production system, the control instructions for the startup phase are generated based on the online predicted steady-state water addition flow, and the control instructions for the startup phase are sent to the tobacco production system to control the water addition flow in the startup phase. After entering the normal operation phase from the startup phase, rolling optimization is performed with the sampling time point as the granularity. Specifically, the current environmental variable data and the real-time production data collected before the current moment, as well as the water addition flow at each sampling time point in the future time period, are input into the tobacco leaf moisture content prediction model to predict the tobacco leaf moisture content prediction time series data in the future time period. By establishing and solving the objective function, the optimized water addition flow at each sampling time point in the future time period is obtained. The control instructions for the normal operation phase are issued to the tobacco production system. The control instructions for the normal operation phase include the optimized water addition flow rate at each sampling time point in the future time period to control the water addition flow rate of the tobacco production system from the current moment to the first sampling time point. The tobacco moisture content prediction model is calibrated according to the predicted tobacco leaf moisture content time series data at the first sampling time point and the collected actual tobacco leaf moisture content. After the first sampling time point is reached at the current moment, the first sampling time point is used as the new current moment, and the above steps are iteratively executed until the entire production process is completed.

[0104] Figure 4 This is a schematic diagram of the structure of a tobacco production control device provided in an embodiment of the present application. Figure 4 , the device may include:

[0105] An acquisition module 41 is used to acquire historical production data of historical batches of tobacco leaves processed by the current production process and its previous production process, and to acquire current environmental variable data;

[0106] The steady-state water addition flow prediction module 42 is used to input the current environmental variable data and the historical production data into the steady-state water addition flow prediction model to predict the steady-state water addition flow of the current batch of tobacco leaves in the normal operation stage of the tobacco production system during the current production process;

[0107] A generating module 43 is used to generate a water addition flow control curve of the tobacco production system during the startup phase, wherein the water addition flow control curve at the end of the startup phase is a steady-state water addition flow;

[0108] The water adding control module 44 is used to control the water adding according to the water adding flow control curve during the startup phase.

[0109] Further optionally, the generation module 43 is specifically used to: obtain optimal curve parameters, the optimal curve parameters include curve slope and / or curve linearity; based on the optimal curve parameters, with the goal of making the water addition flow rate of the tobacco production system at the end of the startup phase a steady-state water addition flow rate, generate a water addition flow control curve for the startup phase.

[0110] Further optionally, the historical production data of the current production process includes: the water addition flow rate, tobacco flow rate, steam volume, steam temperature of historical batches of tobacco leaves of the same brand processed by the current production process during the processing, and the moisture content of tobacco leaves of the historical batches after being processed by the current production process;

[0111] and / or,

[0112] The historical production data of the front-end production process includes: the water addition flow rate, tobacco flow rate, steam volume, steam temperature during the front-end process of processing historical batches of tobacco leaves, and the moisture content of tobacco leaves after the historical batches of tobacco leaves have been processed by the front-end process;

[0113] and / or,

[0114] The current environmental variable data includes: current environmental temperature, and / or current environmental humidity.

[0115] Further optionally, the training method of the steady-state water addition flow prediction model includes: obtaining a first training data set, the training data in the first training data set including: historical production data of historical batches of tobacco leaves of the same brand processed by the current production process, historical production data of historical batches of tobacco leaves processed by the previous process, current environmental variable data and annotation results of the production site, the annotation results including the steady-state water addition flow rate of the tobacco production system; using the first training data set to perform model training to obtain a trained steady-state water addition flow rate prediction model.

[0116] Further optionally, the above-mentioned device also includes: a tobacco leaf moisture content prediction module 45, which is used to input the current environmental variable data and the real-time production data collected before the current moment, as well as the water addition flow rate at each sampling time point in the future time period, into the tobacco leaf moisture content prediction model after the tobacco production system enters the normal operation stage, so as to predict the tobacco leaf moisture content prediction time series data in the future time period, and the tobacco leaf moisture content prediction time series data includes the predicted tobacco leaf moisture content at each sampling time point in the future time period;

[0117] The optimization module 46 is used to construct an objective function based on the deviation information between the predicted time series data of tobacco leaf moisture content and the set tobacco leaf moisture content, and the change information between the water addition flow rates at adjacent sampling time points; the water addition flow rate at each sampling time point in the future time period is used as the optimization variable, and the objective function is solved to obtain the optimized water addition flow rate at each sampling time point in the future time period;

[0118] The water addition control module 44 is also used to control the water addition according to the optimized water addition flow rate at the first sampling time point during the period from the current moment to the first sampling time point in the future time period; if the current moment reaches the first sampling time point, the first sampling time point is used as the new current moment, and the above steps are iteratively executed until the normal operation stage ends.

[0119] Further optionally, the optimization module 46 is also used to calibrate the tobacco leaf moisture content prediction model according to the error value between the predicted tobacco leaf moisture content at the first sampling time point and the actual tobacco leaf moisture content.

[0120] Further optionally, when the optimization module 46 solves the objective function, it is specifically used to: solve the objective function under constraints; the constraints include one or more of the following: the predicted time series data of tobacco leaf moisture content satisfies the first upper and lower limit constraints; the difference between the water addition flow rates at adjacent sampling time points satisfies the second upper and lower limit constraints; the water addition flow rate at each sampling time point satisfies the third upper and lower limit constraints.

[0121] Further optionally, the real-time production data collected before the current moment includes: the time series data of the water addition flow of the current batch of tobacco leaves processed by the current production process, the time series data of the tobacco flow of the current batch of tobacco leaves, and the time series data of the moisture content of the current batch of tobacco leaves before being processed by the current production process.

[0122] Further optionally, the training method of the tobacco leaf moisture content prediction model includes: obtaining a second training data set, the training data in the second training data set includes: historical production data of historical batches of tobacco leaves of the same brand processed by the current production process, current environmental variable data of the production site and annotation results, the annotation results include: time series data of the moisture content of historical batches of tobacco leaves after being processed by the current production process; using the second training data set to perform model training to obtain a tobacco leaf moisture content prediction model.

[0123] Figure 4 The device shown can perform Figure 1 or Figure 2 The method shown in the figure, its implementation principle and technical effect are not described in detail. Figure 4 The specific manner in which each module and unit of the device shown performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0124] It should be noted that the execution subject of each step of the method provided in the above embodiment can be the same device, or the method can be executed by different devices. For example, the execution subject of steps 101 to 104 can be device A; for another example, the execution subject of steps 101 and 102 can be device A, and the execution subject of steps 103 and 104 can be device B; and so on.

[0125] In addition, in some of the processes described in the above embodiments and the accompanying drawings, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel, and the sequence numbers of the operations, such as 401, 402, etc., are only used to distinguish between different operations, and the sequence numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0126] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0127] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 5 As shown, the electronic device includes: a memory 51 and a processor 52;

[0128] The memory 51 is used to store computer programs and can be configured to store various other data to support operations on the computing platform. Examples of such data include instructions for any application or method operating on the computing platform, contact data, phone book data, messages, pictures, videos, etc.

[0129] The memory 51 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read only memory (EEPROM), erasable programmable read only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0130] The processor 52 is coupled to the memory 51 and is used to execute the computer program in the memory 51 to execute the steps in the tobacco production control method.

[0131] Further optional, such as Figure 5 As shown, the electronic device also includes: a communication component 53, a display 54, a power component 55, an audio component 56 and other components. Figure 5 Only some components are shown schematically, which does not mean that the electronic device only includes Figure 5 In addition, Figure 5 The components in the dashed box are optional components, not mandatory components, and the specific components depend on the product form of the electronic device. The electronic device of this embodiment can be implemented as a terminal device such as a desktop computer, a laptop computer, a smart phone, or an IOT (Internet of Things) device, or a server device such as a conventional server, a cloud server, or a server array. If the electronic device of this embodiment is implemented as a terminal device such as a desktop computer, a laptop computer, a smart phone, etc., it can include Figure 5 If the electronic device of this embodiment is implemented as a server device such as a conventional server, a cloud server or a server array, it may not include Figure 5 Components within the dashed box.

[0132] For the detailed implementation process of the processor executing each action, please refer to the relevant description in the aforementioned method embodiment or device embodiment, which will not be repeated here.

[0133] Accordingly, an embodiment of the present application further provides a computer-readable storage medium storing a computer program, which, when executed, can implement each step that can be executed by an electronic device in the above method embodiment.

[0134] Accordingly, an embodiment of the present application also provides a computer program product, including a computer program / instruction. When the computer program / instruction is executed by a processor, the processor is enabled to implement each step in the above method embodiment that can be executed by an electronic device.

[0135] The communication component is configured to facilitate wired or wireless communication between the device where the communication component is located and other devices. The device where the communication component is located can access a wireless network based on a communication standard, such as WiFi (Wireless Fidelity), 2G (2 Generation), 3G (3 Generation), 4G (4 Generation) / LTE (long Term Evolution), 5G (5 Generation), and other mobile communication networks, or a combination thereof. In an exemplary embodiment, the communication component receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra wide band (UWB) technology, Bluetooth (BT) technology, and other technologies.

[0136] The display includes a screen, and the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation.

[0137] The power supply assembly provides power to various components of the device where the power supply assembly is located. The power supply assembly may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the device where the power supply assembly is located.

[0138] The above-mentioned audio component can be configured to output and / or input audio signals. For example, the audio component includes a microphone (MIC), and when the device where the audio component is located is in an operating mode, such as a call mode, a recording mode, and a speech recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in a memory or sent via a communication component. In some embodiments, the audio component also includes a speaker for outputting an audio signal.

[0139] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0140] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0141] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0142] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0143] In a typical configuration, a computing device includes one or more processors (Central Processing Unit, CPU), input / output interface, network interface and memory.

[0144] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0145] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, Phase Change RAM (PRAM), Static Random-Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically-Erasable Programmable Read-Only Memory (EEPROM), flash memory or other memory technology, CD-ROM, Digital Versatile Disc (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission medium that can be used to store information that can be accessed by a computing device. According to the definition in this article, computer-readable media does not include temporary computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0146] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0147] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A tobacco production control method, characterized in that: include: Obtain historical production data of historical batches of tobacco leaves processed by the current production process and its previous production process, and obtain current environmental variable data; Inputting the current environmental variable data and the historical production data into a steady-state water addition flow prediction model to predict the steady-state water addition flow of the current batch of tobacco leaves processed by the current production process during the normal operation stage of the tobacco production system; generating a water addition flow control curve for the startup phase of the tobacco production system, wherein the water addition flow at the end of the startup phase on the water addition flow control curve is the steady-state water addition flow; During the startup phase, water addition control is performed according to the water addition flow control curve; After the tobacco production system enters the normal operation stage, the current environmental variable data and the real-time production data collected before the current moment, as well as the water addition flow rate at each sampling time point in the future time period, are input into the tobacco leaf moisture content prediction model to predict the tobacco leaf moisture content prediction time series data in the future time period, wherein the tobacco leaf moisture content prediction time series data includes the predicted tobacco leaf moisture content at each sampling time point in the future time period; Based on the deviation information between the predicted time series data of tobacco leaf moisture content and the set tobacco leaf moisture content, and the change information between the water addition flow rates at adjacent sampling time points, an objective function is constructed; the water addition flow rate at each sampling time point in the future time period is used as the optimization variable to solve the objective function and obtain the optimized water addition flow rate at each sampling time point in the future time period; during the period from the current moment to the first sampling time point in the future time period, water addition control is performed according to the optimized water addition flow rate at the first sampling time point; if the current moment reaches the first sampling time point, the first sampling time point is used as the new current moment, and the above steps are iteratively executed until the normal operation stage ends.

2. The method according to claim 1, characterized in that Generating a water addition flow control curve during the startup phase of the tobacco production system includes: Obtaining optimal curve parameters, wherein the optimal curve parameters include curve slope and / or curve linearity; According to the optimal curve parameters, a water addition flow control curve for the startup phase is generated with the goal of making the water addition flow of the tobacco production system at the end of the startup phase the steady-state water addition flow.

3. The method according to claim 1, characterized in that The historical production data of the current production process includes: the water addition flow rate, tobacco flow rate, steam volume, steam temperature of the historical batch of tobacco leaves of the same brand processed by the current production process during the processing, and the moisture content of the tobacco leaves of the historical batch of tobacco leaves after being processed by the current production process; and / or, The historical production data of the front-end production process includes: the water addition flow rate, tobacco flow rate, steam volume, steam temperature during the process of processing the historical batch of tobacco leaves by the front-end production process, and the moisture content of the tobacco leaves of the historical batch of tobacco leaves after being processed by the front-end production process; and / or, The current environmental variable data includes: current environmental temperature, and / or current environmental humidity.

4. The method according to any one of claims 1 to 3, characterized in that: The training method of the steady-state water addition flow prediction model includes: Acquire a first training data set, wherein the training data in the first training data set includes: historical production data of historical batches of tobacco leaves of the same brand processed by the current production process, historical production data of historical batches of tobacco leaves processed by the previous production process, current environmental variable data of the production site and annotation results, wherein the annotation results include a steady-state water addition flow rate of the tobacco production system; The first training data set is used to perform model training to obtain the trained steady-state water addition flow prediction model.

5. The method according to claim 1, characterized in that Before taking the first sampling time point as the new current time, the method further includes: The tobacco leaf moisture content prediction model is calibrated according to the error value between the predicted tobacco leaf moisture content at the first sampling time point and the actual tobacco leaf moisture content.

6. The method according to claim 1, characterized in that Solving the objective function includes: Solving the objective function under constraints; The constraints include one or more of the following: The tobacco leaf moisture content prediction time series data meets the first upper and lower limit constraints; The difference between the water addition flow rates at adjacent sampling time points satisfies the second upper and lower limit constraints; The water addition flow rate at each sampling time point satisfies the third upper and lower limit constraints.

7. The method according to claim 1, characterized in that The real-time production data collected before the current moment includes: the time series data of the water addition flow of the current batch of tobacco leaves processed by the current production process, the time series data of the tobacco flow of the current batch of tobacco leaves, and the time series data of the moisture content of the current batch of tobacco leaves before being processed by the current production process.

8. The method according to claim 1, characterized in that The training method of the tobacco leaf moisture content prediction model includes: Acquire a second training data set, wherein the training data in the second training data set includes: historical production data of historical batches of tobacco leaves of the same brand processed by the current production process, environmental variable data of the production site, and annotation results, wherein the annotation results include: time series data of moisture content of tobacco leaves of the historical batches after being processed by the current production process; The second training data set is used to perform model training to obtain the tobacco leaf moisture content prediction model.

9. An electronic device, characterized in that: include: Memory and processor; The memory is used to store computer programs; The processor is coupled to the memory and configured to execute the computer program to perform the steps of the method according to any one of claims 1 to 8.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the processor is enabled to implement the steps in the method according to any one of claims 1 to 8.

11. A computer program product, characterized in that The method comprises a computer program / instruction, which, when executed by a processor, enables the processor to implement the steps of the method according to any one of claims 1 to 8.