Tobacco smoking quality prediction method, device, equipment, medium and product

Abstract

The invention discloses a tobacco smoke panel test quality prediction method, device and equipment, a medium and a product. The method comprises the following steps: acquiring a historical climate data matrix corresponding to a to-be-predicted tobacco leaf producing area in a historical duration; wherein the historical climate data matrix comprises historical climate data corresponding to a plurality of target climate characteristics; and processing the historical climate data matrix according to a smoking quality prediction model obtained by pre-training, and predicting to obtain predicted smoking quality data corresponding to tobacco leaves produced by the tobacco leaf producing area to be predicted within the historical duration. According to the technical scheme provided by the embodiment of the invention, an effect of predicting the smoking quality of the tobacco leaves only according to the neural network model and the historical climate data of the climate characteristics without obtaining an actual sample of the tobacco leaves is achieved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method, device, equipment, medium and product for predicting tobacco leaf smoking quality. Background Art

[0002] As a leaf crop, tobacco is one of the most important cash crops. The quality of tobacco leaves has a decisive influence on the quality of cigarettes, and climate characteristics have a key impact on the characteristic style and quality of tobacco leaves. The characteristic style and quality of tobacco leaves are directly reflected in their smoking quality. Therefore, how to predict the smoking quality of tobacco leaves based on climate characteristics is of great value, which can provide important decision-making support for the production of flue-cured tobacco.

[0003] In the related art, the smoking quality of tobacco leaves is usually obtained after professionals have evaluated and scored various indicators. This method of obtaining the smoking quality of tobacco leaves may have problems such as low efficiency and accuracy, and consumes manpower and material costs, and reduces the timeliness of tobacco leaf smoking quality prediction work. Summary of the invention

[0004] The present invention provides a tobacco leaf smoking quality prediction method, device, equipment, medium and product, so as to achieve the effect of predicting the smoking quality of tobacco leaves based on a neural network model and historical climate data of climate characteristics without obtaining actual tobacco leaf samples.

[0005] According to one aspect of the present invention, a method for predicting tobacco leaf smoking quality is provided, the method comprising:

[0006] Obtaining a historical climate data matrix corresponding to the tobacco production area to be predicted within a historical period; wherein the historical climate data matrix includes historical climate data corresponding to a plurality of target climate characteristics;

[0007] The historical climate data matrix is ​​processed according to the pre-trained smoking quality prediction model to predict the smoking quality data corresponding to the tobacco produced in the tobacco production area to be predicted within the historical period;

[0008] Among them, the smoking quality prediction model is used to predict the smoking quality data of tobacco produced in the tobacco producing area within the historical period according to the historical climate data of the tobacco producing area within the historical period; the smoking quality prediction model is based on the sample climate data matrix of the sample tobacco producing area within the historical period and the real smoking quality data of the tobacco produced in the sample tobacco producing area within the historical period. The pre-constructed deep learning model is trained.

[0009] According to another aspect of the present invention, a tobacco leaf smoking quality prediction device is provided, the device comprising:

[0010] A data acquisition module is used to acquire a historical climate data matrix corresponding to the tobacco production area to be predicted within a historical period; wherein the historical climate data matrix includes historical climate data corresponding to multiple target climate characteristics;

[0011] The smoking quality data prediction module is used to process the historical climate data matrix according to the pre-trained smoking quality prediction model, and predict the predicted smoking quality data corresponding to the tobacco produced in the predicted tobacco producing area within the historical period; wherein the smoking quality prediction model is used to predict the smoking quality data of the tobacco produced in the tobacco producing area within the historical period according to the historical climate data of the tobacco producing area within the historical period; the smoking quality prediction model is trained on a pre-constructed deep learning model based on the sample climate data matrix of the sample tobacco producing area within the historical period and the real smoking quality data of the tobacco produced in the sample tobacco producing area within the historical period.

[0012] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0013] at least one processor; and

[0014] a memory communicatively connected to the at least one processor; wherein,

[0015] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the tobacco leaf smoking quality prediction method described in any embodiment of the present invention.

[0016] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the tobacco leaf smoking quality prediction method described in any embodiment of the present invention when executed.

[0017] According to another aspect of the present invention, a computer program product is provided, the computer program product comprising a computer program, and the computer program, when executed by a processor, implements the tobacco leaf smoking quality prediction method described in any embodiment of the present invention.

[0018] The technical solution of the embodiment of the present invention obtains the historical climate data matrix corresponding to the tobacco production area to be predicted within the historical period; wherein the historical climate data matrix includes historical climate data corresponding to multiple target climate characteristics, which provides a data basis for subsequent smoking quality prediction, and the constructed data matrix can be helpful in characterizing the relationship between climate changes and seasons and dates, which helps to improve the prediction accuracy of smoking quality; further, the historical climate data matrix is ​​processed according to the pre-trained smoking quality prediction model to predict the predicted smoking quality data corresponding to the tobacco produced in the tobacco production area to be predicted within the historical period, which solves the problems of low efficiency, low accuracy, and labor and material costs in related technologies, and achieves the effect of predicting the smoking quality of tobacco leaves based on the neural network model and historical climate data of climate characteristics without obtaining actual tobacco leaf samples, greatly improving the timeliness of tobacco leaf smoking quality prediction, improving the prediction efficiency and prediction accuracy of tobacco leaf smoking quality, and providing corresponding guidance for early decision-making in the tobacco industry to a certain extent.

[0019] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0021] Figure 1 is a flow chart of a tobacco leaf smoking quality prediction method provided in accordance with the first embodiment of the present invention;

[0022] Figure 2 is a flow chart of a tobacco leaf smoking quality prediction method provided in accordance with the second embodiment of the present invention;

[0023] Figure 3 This is a flow chart of a method for predicting tobacco leaf smoking quality provided in Embodiment 2 of the present invention;

[0024] Figure 4 is a flow chart of a tobacco leaf smoking quality prediction method provided in accordance with Embodiment 3 of the present invention;

[0025] Figure 5 2 is a schematic diagram of the structure of a tobacco leaf smoking quality prediction device provided according to Embodiment 4 of the present invention;

[0026] Figure 6 It is a structural schematic diagram of an electronic device for implementing the tobacco leaf smoking quality prediction method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0027] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0028] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0029] Embodiment 1

[0030] Figure 1 This is a flowchart of a tobacco leaf smoking quality prediction method provided by the first embodiment of the present invention. This embodiment is applicable to the case of predicting the smoking quality of tobacco leaves produced in a tobacco leaf producing area. The method can be executed by a tobacco leaf smoking quality prediction device. The tobacco leaf smoking quality prediction device can be implemented in the form of hardware and / or software. The tobacco leaf smoking quality prediction device can be configured in a terminal and / or a server. Figure 1 As shown, the method includes:

[0031] S110, obtaining a historical climate data matrix corresponding to the tobacco production area to be predicted within a historical period; wherein the historical climate data matrix includes historical climate data corresponding to multiple target climate characteristics.

[0032] Among them, the tobacco production area to be predicted can be the tobacco production area for which the tobacco smoking quality prediction is to be performed. The tobacco production area generally refers to the area where tobacco is planted and produced, and such an area usually has climatic conditions, soil environment and planting techniques suitable for tobacco growth. The historical duration can be any duration before the current moment. In this embodiment, the historical duration can be a time interval determined according to the growth law of tobacco leaves. Exemplarily, the historical duration can be a time interval including every day of May, June, July, August and September in a year, for example, 153 days. The historical climate data matrix can be a matrix constructed based on historical climate data. The historical climate data matrix includes historical climate data corresponding to multiple climate characteristics. The target climate characteristics can be climate characteristics that have a greater impact on the growth process of tobacco leaves. In general, some climate characteristics will have a greater impact on the growth process of tobacco leaves, and may further affect the tobacco leaf evaluation quality, tobacco leaf chemical component content and other tobacco leaf quality indicators, and these climate characteristics can be used as target climate characteristics. It should be noted that the introduction of inappropriate climate characteristics may introduce too much interference, which may directly affect the prediction results of the smoking quality. Therefore, screening out the target climate characteristics that have a greater impact on the tobacco leaf growth process can improve the prediction accuracy of tobacco leaf smoking quality. Optionally, multiple target climate characteristics include daily maximum temperature, daily average temperature, daily average atmospheric pressure, daily total precipitation, daily average net sunshine intensity and daily average wind speed. It should be noted that the reason for selecting daily maximum temperature and daily average temperature is that the daily maximum temperature reflects the peak heat in a day, which has a significant impact on the physiological response of crops, and may affect the photosynthesis efficiency together with the sunshine intensity, thereby affecting the growth and smoking quality level of crops; the daily average temperature provides a comprehensive description of the heat throughout the day and is a good indicator for measuring the growth environment. The reason for selecting the daily average atmospheric pressure is that different regions have unique climate characteristics due to different altitudes, and there is a complex invisible connection between climate and altitude. In addition, since atmospheric pressure is a direct reflection of altitude characteristics, the daily average atmospheric pressure is additionally introduced into the target climate characteristics to improve the richness of the data. In addition, the introduction of the daily average atmospheric pressure climate feature to characterize the characteristics of different tobacco leaf producing areas enhances the cross-regional applicability and generalization ability of the model. The reason for choosing the daily average wind speed is that for tobacco leaves, the wind speed will change the temperature gradient on the leaf surface, affect the water vapor exchange rate between the leaf and the air, stimulate the opening and closing of stomata, affect the absorption of carbon dioxide, and further affect the rate of transpiration and photosynthesis. Therefore, the daily average wind speed is also introduced into the target climate characteristics to improve the richness of the data. The historical climate data can be the climate data corresponding to the target climate characteristics in the historical period of the tobacco production area to be predicted.

[0033] In this embodiment, the historical climate data matrix can be constructed based on the historical climate data corresponding to multiple target climate characteristics. Furthermore, the historical climate data corresponding to multiple target climate characteristics corresponding to the tobacco production area to be predicted within the historical period can be first obtained. Further, the historical climate data matrix can be constructed based on the historical climate data corresponding to multiple target climate characteristics.

[0034] Optionally, a historical climate data matrix corresponding to the tobacco leaf producing area to be predicted within the historical time period is obtained, including: obtaining historical climate data corresponding to multiple target climate characteristics of the tobacco leaf producing area to be predicted within the historical time period; for multiple target climate characteristics, arranging the historical climate data corresponding to the target climate characteristics in chronological order to obtain historical climate data vectors corresponding to the target climate characteristics; splicing the historical climate data vectors corresponding to multiple target climate characteristics to obtain a historical climate data matrix.

[0035] The historical climate data vector may be a one-dimensional matrix consisting of a row of historical climate data or a column of historical climate data. For example, assuming that the target climate feature is the daily average temperature, the historical duration is 153 days, and the number of historical climate data corresponding to the daily average temperature is 153. In this case, the historical climate data vector corresponding to the daily average temperature is a column of historical climate data composed of the 153 historical climate data arranged in the chronological order of the 153 days, that is, a matrix with 153 rows and 1 column.

[0036] As an optional implementation mode of this embodiment, when multiple target climate characteristics are determined from the climate characteristics and the historical duration associated with the tobacco leaf growth law is determined, the historical climate data corresponding to the multiple target climate characteristics of the tobacco leaf production area to be predicted within the historical duration can be obtained. Further, for multiple target climate characteristics, the historical climate data corresponding to the target climate characteristics can be arranged in order of the historical duration, and the vector obtained after the arrangement is used as the historical climate data vector corresponding to the target climate characteristics. Further, the historical climate data vectors corresponding to the multiple target climate characteristics are spliced ​​together to obtain a historical climate data matrix. Exemplarily, assuming that the number of target climate characteristics is 6, the historical duration is 153 days, and the historical climate data vector corresponding to one target climate characteristic is a 153×1 matrix, and then, the historical climate data matrix is ​​a 153×6 matrix, 153 represents the number of days in the historical duration, and 6 represents the number of target climate characteristics.

[0037] It should be noted that the historical climate data vectors can be spliced ​​randomly, that is, the arrangement order of multiple target climate features is not limited.

[0038] It should also be noted that the historical climate data structure based on chronological order has rich time series information, which can better fit the climate change relationship in the real environment. In the process of prediction using models that are good at capturing time series information, it can establish a more complete climate change relationship and improve the prediction accuracy of tobacco leaf smoking quality.

[0039] S120. Process the historical climate data matrix according to the pre-trained smoking quality prediction model to predict the smoking quality data corresponding to the tobacco produced in the tobacco production area to be predicted within a historical period of time.

[0040] Among them, the smoking quality prediction model can be a neural network model that takes the historical climate data of the tobacco leaf producing area to be predicted as an input object to predict the smoking quality of the tobacco leaves produced in the tobacco leaf producing area to be predicted based on the input object. The smoking quality prediction model is used to predict the smoking quality data of the tobacco leaves produced in the tobacco leaf producing area within the historical length of time according to the historical climate data of the tobacco leaf producing area within the historical length of time. The smoking quality prediction model can be a deep learning model of any model structure. The model structure of the smoking quality prediction model is not limited here, and can be customized according to actual needs. Optionally, the smoking quality prediction model includes a convolutional neural network, a long short-term memory network, and a fully connected layer. Convolutional neural networks are good at extracting local features of input data, especially in multidimensional time series data. Capturing important feature information. Long short-term memory networks can maintain information over a long period of time, thereby effectively processing long-term dependencies in time series data. By combining the advantages of these two neural networks, the smoking quality prediction model can capture the complex spatiotemporal interactions in climate characteristics, providing a more accurate model for predicting tobacco leaf smoking quality. The predicted smoking quality data can be tobacco smoking quality data predicted based on a neural network model. Optionally, the predicted smoking quality data includes quality parameters corresponding to multiple smoking quality indicators to be predicted. The smoking quality indicators to be predicted are usually key elements for measuring the intrinsic quality of tobacco leaves. Multiple smoking quality indicators to be predicted include fragrance, sweet fragrance, burnt fragrance, concentration, aroma quality, aroma volume, miscellaneous smell, stimulation, aftertaste and total score. It can be understood that fragrance, sweet fragrance and burnt fragrance are the aroma types of tobacco leaves. Fragrance has a prominent fragrance, comfortable taste, lighter smoke concentration, lighter local miscellaneous smell, and soft to moderate strength; sweet fragrance is usually related to certain sugar substances and other aroma components in tobacco leaves, and there is a feeling of pleasure and comfort; burnt fragrance is usually related to the burning process of tobacco leaves. It is a unique aroma produced by tobacco leaves at high temperatures. The smoke of burnt-flavor tobacco leaves is strong, the taste is comfortable, the local miscellaneous smell is very light to heavy, and the strength is moderate. Concentration: The concentration of cigarette smoke refers to the degree of strength of the smoke after it enters the mouth. Aroma quality: It reflects the quality and type of tobacco aroma and is an important indicator for evaluating the aroma of tobacco leaves. Aroma quantity: It indicates the amount or concentration of tobacco aroma and is an important indicator for evaluating the richness of tobacco aroma. Miscellaneous smells: It refers to the unpleasant smells that should not be present in the smoke. Irritation: It describes the degree of irritation of smoke to the mouth, tongue, throat, nasal cavity and other parts. Aftertaste: It refers to the feeling of smoke remaining in the mouth after exhaled. The total score is an evaluation of the comprehensive quality of tobacco leaves.

[0041] As an optional implementation of the first embodiment of this invention, when the historical climate data matrix corresponding to the tobacco production area to be predicted within the historical period is obtained, the historical climate data matrix can be input into the pre-trained smoking quality prediction model. Furthermore, the historical climate data matrix can be processed in sequence based on the convolutional neural network, long short-term memory network and fully connected layer in the smoking quality prediction model, and the predicted smoking quality data corresponding to the tobacco produced in the tobacco production area to be predicted within the historical period is output.

[0042] In this embodiment, in order to clearly and intuitively observe the correlation between the climate data of the target climate characteristics and the tobacco leaf smoking quality data, and provide a data basis for improving the tobacco leaf smoking quality, after obtaining the predicted smoking quality data, a correlation diagram can also be generated based on the predicted smoking quality data and the historical climate data matrix.

[0043] Optionally, after predicting the predicted smoking quality data corresponding to the tobacco produced in the tobacco producing area to be predicted within the historical period, it also includes: generating a predicted smoking quality curve based on the predicted smoking quality data; generating a climate smoking quality correlation diagram corresponding to the tobacco producing area to be predicted within the historical period based on the predicted smoking quality curve and the historical climate data matrix.

[0044] Among them, the predicted smoking quality curve can be a curve that characterizes the distribution law of the predicted smoking quality data. The predicted smoking quality curve can be a curve with time as the horizontal axis and the various smoking quality indicators to be predicted as the vertical axis. The climate smoking quality correlation diagram can be an image that characterizes the correlation between tobacco leaf smoking quality and historical climate data. The climate smoking quality correlation diagram can be any graph that can characterize the correlation between two types of data. Optionally, the graphic type of the climate smoking quality correlation diagram can be at least one of a scatter plot, a bubble plot, a heat map, a network diagram, and a parallel coordinates diagram.

[0045] As an optional implementation of this embodiment, after obtaining the predicted smoking quality data, data analysis is performed on it and it is converted from data form to a graphical form to obtain a predicted smoking quality curve. Further, data analysis can be performed on the predicted smoking quality curve and the historical climate data matrix to obtain a climate smoking quality association diagram corresponding to the predicted tobacco production area within the historical period.

[0046] It should be noted that the advantage of first generating the predicted smoking quality curve and then generating the climate smoking quality correlation diagram is that the changing trend of the quality parameters of the various smoking quality indicators to be predicted of tobacco leaves can be clearly and intuitively observed based on the predicted smoking quality curve. Furthermore, the correlation between the smoking quality of tobacco leaves and historical climate data can be clearly and intuitively observed based on the climate smoking quality correlation diagram.

[0047] The technical solution of the embodiment of the present invention obtains the historical climate data matrix corresponding to the tobacco production area to be predicted within the historical period; wherein the historical climate data matrix includes historical climate data corresponding to multiple target climate characteristics, which provides a data basis for subsequent smoking quality prediction, and the constructed data matrix can be helpful in characterizing the relationship between climate changes and seasons and dates, which helps to improve the prediction accuracy of smoking quality; further, the historical climate data matrix is ​​processed according to the pre-trained smoking quality prediction model to predict the predicted smoking quality data corresponding to the tobacco produced in the tobacco production area to be predicted within the historical period, which solves the problems of low efficiency, low accuracy, and labor and material costs in related technologies, and achieves the effect of predicting the smoking quality of tobacco leaves based on the neural network model and historical climate data of climate characteristics without obtaining actual tobacco leaf samples, greatly improving the timeliness of tobacco leaf smoking quality prediction, improving the prediction efficiency and prediction accuracy of tobacco leaf smoking quality, and providing corresponding guidance for early decision-making in the tobacco industry to a certain extent.

[0048] Embodiment 2

[0049] Figure 2 It is a flow chart of a tobacco leaf smoking quality prediction method provided in the second embodiment of the present invention. On the basis of the above-mentioned embodiment, the process of determining the predicted smoking quality data is further refined. Optionally, the smoking quality prediction model includes a convolutional neural network and a long short-term memory network; the historical climate data matrix is ​​processed according to the pre-trained smoking quality prediction model to predict the predicted smoking quality data corresponding to the tobacco produced in the predicted tobacco production area within the historical period, including: extracting features from the historical climate data matrix according to the convolutional neural network to obtain a historical climate feature matrix; processing the historical climate feature matrix according to the long short-term memory network to obtain the predicted smoking quality data corresponding to the tobacco produced in the predicted tobacco production area within the historical period. The specific implementation method can be referred to the technical solution of this embodiment. Among them, the technical terms that are the same or similar to those in the above-mentioned embodiments are not repeated here.

[0050] like Figure 2 As shown, the method includes:

[0051] S210, obtaining a historical climate data matrix corresponding to the tobacco production area to be predicted within a historical period; wherein the historical climate data matrix includes historical climate data corresponding to multiple target climate characteristics.

[0052] S220. Extract features of the historical climate data matrix according to the convolutional neural network to obtain a historical climate feature matrix.

[0053] Among them, the convolutional neural network (CNN) is a type of feedforward neural network that includes convolution calculations and has a deep structure. It is one of the representative algorithms of deep learning. At the same time, CNN has the ability to represent learning and can perform translation-invariant classification of input information according to its hierarchical structure. In this embodiment, the convolutional neural network may include at least one convolutional layer, and the convolution kernel size of the convolutional layer may be any value, optionally 1×6. It should be noted that setting the convolution kernel size to 1×6 is associated with the number of target climate features. Based on the description in the above embodiment, the number of target climate features is 6, and then, the convolution kernel size of the convolutional layer in the convolutional neural network can be set to 1×6, which allows the convolutional neural network to extract the climate data features of the 6 target climate features as one data feature. The historical climate data matrix is ​​input into the convolutional neural network, and the output historical climate feature matrix is ​​the local data feature of the historical climate data matrix. Compared with the global data feature, the local data feature can be understood as the local expression of the data feature, reflecting the local features on the historical climate data matrix. The feature data included in the historical climate feature matrix is ​​important climate information in the historical climate data matrix. It should be noted that the advantage of feature extraction of the historical climate data matrix is ​​that it can speed up the model calculation and improve the prediction accuracy of tobacco leaf smoking quality data.

[0054] As an optional implementation of this embodiment, after obtaining the historical climate data matrix, the historical climate data matrix can be input into the inhalation quality prediction model. Further, the historical climate data matrix can be feature extracted based on the convolutional neural network in the inhalation quality prediction model to obtain the historical climate feature matrix.

[0055] For example, assuming that the historical climate data matrix is ​​a 153×6 matrix, the convolution kernel size of the convolution layer in the convolutional neural network is 1×6. Then, the historical climate data matrix is ​​input into the convolutional neural network, and the output historical climate feature matrix is ​​a 153×1 feature matrix.

[0056] S230. The historical climate characteristic matrix is ​​processed according to the long short-term memory network to predict and obtain the prediction evaluation quality characteristics.

[0057] Among them, the Long Short-Term Memory (LSTM) network can be used to capture the temporal dependencies of climate characteristics. The LSTM network can usually understand the temporal dynamics of data well and make predictions based on time series data. The predicted air quality characteristics can be obtained by analyzing the historical climate characteristic matrix through the LSTM network and predicting the characteristic values ​​corresponding to each air quality indicator to be predicted.

[0058] As an optional implementation of the first embodiment of this invention, after obtaining the historical climate feature matrix, the historical climate feature matrix can be input into the long short-term memory network. Further, the historical climate feature matrix can be processed based on the long short-term memory network to output the eigenvalues ​​corresponding to each of the inhalation quality indicators to be predicted, and the eigenvalues ​​corresponding to the outputted multiple inhalation quality indicators to be predicted are used as the predicted inhalation quality features.

[0059] S240. Process the predicted smoking quality features according to the fully connected layer to obtain the predicted smoking quality data corresponding to the tobacco produced in the predicted tobacco producing area within a historical period of time.

[0060] Among them, the fully connected layer is used to map the learned features to specific output types or values ​​to complete tasks such as classification, regression, and numerical prediction. At the same time, the fully connected layer usually performs nonlinear transformations on the results of linear combinations under the action of activation functions, introduces nonlinear properties, and enables the evaluation quality prediction model to fit more complex nonlinear patterns. In addition, the fully connected layer can also be used to reduce the dimension of the data for subsequent processing.

[0061] As an optional implementation of this embodiment, after obtaining the predicted smoking quality features, the predicted smoking quality features can be input into the fully connected layer. Further, the predicted smoking quality features can be transformed based on the fully connected layer to obtain the predicted smoking quality data corresponding to the tobacco produced in the predicted tobacco production area within the historical period.

[0062] For example, Figure 3 FIG. 1 is a flow chart of a method for predicting tobacco leaf smoking quality provided by an embodiment of the present invention. Figure 3 As shown, the historical climate data matrix to be input into the inhalation quality prediction model is a 153×6 matrix. Further, the historical climate data matrix is ​​input into the inhalation quality prediction model. Then, the historical climate data matrix is ​​processed based on the convolutional neural network in the inhalation quality prediction model, and the historical climate feature matrix is ​​output. The convolution kernel size of the convolution layer in the convolutional neural network is 1×6, and the matrix dimension of the historical climate feature matrix is ​​153×1. Further, the historical climate feature matrix can be input into the long short-term memory network, the historical climate feature matrix is ​​processed based on the long short-term memory network, and the predicted inhalation quality features are output. Further, the predicted inhalation quality features can be input into the fully connected layer, the predicted inhalation quality features are processed based on the fully connected layer, and the predicted inhalation quality data are output.

[0063] The technical solution of the embodiment of the present invention is to extract features from the historical climate data matrix according to a convolutional neural network to obtain a historical climate feature matrix; further, the historical climate feature matrix is ​​processed according to a long short-term memory network to predict predicted smoking quality features; further, the predicted smoking quality features are processed according to a fully connected layer to obtain predicted smoking quality data corresponding to tobacco produced in the tobacco production area to be predicted within a historical period of time, thereby achieving the effect of predicting the smoking quality of tobacco leaves based on the historical climate data of the neural network model and climate characteristics, and the deep learning model combining the convolutional neural network and the long short-term memory network can fully extract the time series characteristics of the climate data and capture the complex nonlinear relationship between climate characteristics and the smoking quality of tobacco leaves.

[0064] Embodiment 3

[0065] Figure 4 This is a flowchart of a tobacco leaf smoking quality prediction method provided in the third embodiment of the present invention. On the basis of the above-mentioned embodiment, a smoking quality prediction model can be obtained based on training samples, and then, the smoking quality data of tobacco leaves produced in the tobacco producing area within a historical period can be predicted based on the smoking quality prediction model. The specific implementation method can refer to the technical solution of this embodiment. Among them, the technical terms that are the same or similar to the above-mentioned embodiments are not repeated here.

[0066] like Figure 4 As shown, the method includes:

[0067] S310, training to obtain a smoking quality prediction model.

[0068] It should be noted that before applying the smoking quality prediction model provided in this embodiment, the pre-constructed deep learning model can be trained in a supervised or unsupervised manner. Before training the deep learning model, multiple training samples can be constructed to train the model based on multiple training samples. In order to improve the prediction accuracy of the smoking quality prediction model, as many and rich training samples as possible can be constructed.

[0069] Optionally, training to obtain a smoking quality prediction model includes: obtaining multiple training samples, wherein the training samples include a sample climate data matrix corresponding to the sample tobacco leaf producing area within a historical period and actual smoking quality data of tobacco leaves produced in the sample tobacco leaf producing area within a historical period; training a pre-constructed deep learning model based on the multiple training samples to obtain a smoking quality prediction model.

[0070] The sample climate data matrix may include sample climate data corresponding to multiple target climate features. The sample climate data may be pre-stored data retrieved from the climate data storage space, or may be data generated by a data generation model, etc. The real smoking quality data may include real quality parameters corresponding to multiple smoking quality indicators to be predicted. The real smoking quality data may characterize the real quality score of tobacco leaves on each smoking quality indicator to be predicted.

[0071] In this embodiment, in order to construct more and richer training samples, tobacco leaves produced in multiple sample tobacco leaf producing areas within a historical period of time can be obtained. Furthermore, the tobacco leaves produced in each sample tobacco leaf producing area are respectively tested for smoking quality, and the real smoking quality data corresponding to the tobacco leaves are obtained. Furthermore, the sample climate data matrix of multiple sample tobacco leaf producing areas within a historical period of time can be obtained. Afterwards, for multiple sample tobacco leaf producing areas, training samples are constructed based on the sample climate data matrix corresponding to the sample tobacco leaf producing areas and the real smoking quality data. Then, multiple training samples can be obtained. Furthermore, a pre-constructed deep learning model can be trained based on multiple training samples to obtain a trained deep learning model. Furthermore, a smoking quality prediction model can be determined based on the trained deep learning model.

[0072] Optionally, a pre-constructed deep learning model is trained based on multiple training samples to obtain an inhalation quality prediction model, including: for multiple training samples, inputting the sample climate data matrix in the training samples into the pre-constructed deep learning model to obtain output inhalation quality data corresponding to the training samples; determining a loss value based on the output inhalation quality data and the actual inhalation quality data in the training samples, correcting the model parameters in the deep learning model based on the loss value, taking the convergence of the loss function in the deep learning model as the training goal, and obtaining at least one model to be verified; verifying at least one model to be verified based on the test samples, and taking the model to be verified whose verification result meets preset conditions as the inhalation quality prediction model.

[0073] Among them, the deep learning model constructed for prediction can be a neural network model with model parameters as default values ​​or initial values. The output quality evaluation data can be the predicted quality evaluation data output after the sample climate data matrix is ​​predicted by the deep learning model. The loss value can be a numerical value that characterizes the degree of difference between the model output and the true output. The loss function can be a function determined based on the loss value and used to characterize the degree of difference between the predicted output and the actual output. The model to be verified can be a model obtained by correcting the model parameters in the training model according to the loss value. The number of models to be verified can be consistent with the number of training iterations (or the number of model parameter corrections). The test sample can be a part of the sample in the training sample, or it can be a sample data reconstructed based on the construction method of the training sample. The preset conditions can include at least one of the mean square error (MSE) reaching the first error threshold, the mean absolute error (MAE) reaching the second error threshold, and the mean absolute percentage error (MAPE) reaching the third error threshold. Among them, the mean square error is usually used to measure the average square difference between the predicted value and the true value. Mean absolute error is often used to measure the average absolute difference between the predicted values ​​and the true values. Mean absolute percentage error is often used to measure the average absolute percentage by which the predicted values ​​deviate from the true values.

[0074] As an optional implementation of this embodiment, after obtaining multiple training samples, for multiple training samples, the sample climate data matrix in the training samples can be input into a pre-built deep learning model to process the sample climate data matrix based on the deep learning model, and output the output evaluation quality data corresponding to the training samples. Further, the output evaluation quality data can be compared with the real evaluation quality data in the training samples, and the loss value can be obtained. Further, in the case of obtaining the loss value, the model parameters in the deep learning model can be corrected based on the loss value, the corrected model can be used as the first model to be processed, and the model to be processed can be backed up and the backup model can be saved. Further, the training sample processing can be repeated to determine the loss value. Then, the model parameters in the model to be processed can be corrected based on the obtained loss value, the corrected model can be used as the second model to be processed, and the model to be processed can be backed up and the backup model can be saved. Further, when the loss function converges, such as when the training error of the loss function is less than the preset error, or the error change trend tends to be stable, the iterative training can be stopped at this time, and the model obtained at this time can be used as the model to be verified. Furthermore, the saved backup model can be retrieved, and both the retrieved backup model and the model to be verified obtained at this time can be used as the model to be verified, so as to obtain at least one model to be verified. Furthermore, in order to evaluate the model performance of at least one model to be verified, a test sample can be obtained, and for at least one model to be verified, the model to be verified can be tested and verified based on the test sample, and a verification result corresponding to the model to be verified can be obtained. Furthermore, it can be determined whether the verification result of at least one model to be verified meets the preset conditions. Furthermore, the model to be verified whose verification result meets the preset conditions can be used as an evaluation and suction quality prediction model.

[0075] S320, obtaining a historical climate data matrix corresponding to the tobacco production area to be predicted within a historical period; wherein the historical climate data matrix includes historical climate data corresponding to multiple target climate characteristics.

[0076] S330, processing the historical climate data matrix according to the pre-trained smoking quality prediction model, and predicting the predicted smoking quality data corresponding to the tobacco produced in the predicted tobacco producing area within the historical period.

[0077] The technical solution of the embodiment of the present invention, by acquiring multiple training samples, further, trains a pre-constructed deep learning model based on the multiple training samples to obtain a smoking quality prediction model, thereby achieving the effect of training a smoking quality prediction model that can accurately predict the smoking quality of tobacco leaves based on the constructed training samples, effectively improving the accuracy and robustness of the model, and effectively improving the model performance.

[0078] Embodiment 4

[0079] Figure 5 Schematic diagram of a tobacco leaf smoking quality prediction device provided by the fourth embodiment of the present invention. Figure 5 As shown, the device includes: a data acquisition module 410 and a smoking quality evaluation data prediction module 420.

[0080] Among them, the data acquisition module 410 is used to obtain the historical climate data matrix corresponding to the tobacco leaf producing area to be predicted within the historical time period; wherein the historical climate data matrix includes historical climate data corresponding to multiple target climate characteristics; the smoking quality data prediction module 420 is used to process the historical climate data matrix according to the pre-trained smoking quality prediction model, and predict the predicted smoking quality data corresponding to the tobacco leaves produced in the tobacco leaf producing area to be predicted within the historical time period; wherein the smoking quality prediction model is used to predict the smoking quality data of the tobacco leaves produced in the tobacco leaf producing area within the historical time period based on the historical climate data of the tobacco leaf producing area within the historical time period; the smoking quality prediction model is obtained by training a pre-constructed deep learning model based on the sample climate data matrix of the sample tobacco leaf producing area within the historical time period and the real smoking quality data of the tobacco leaves produced in the sample tobacco leaf producing area within the historical time period.

[0081] The technical solution of the embodiment of the present invention obtains the historical climate data matrix corresponding to the tobacco production area to be predicted within the historical period; wherein the historical climate data matrix includes historical climate data corresponding to multiple target climate characteristics, which provides a data basis for subsequent smoking quality prediction, and the constructed data matrix can be helpful in characterizing the relationship between climate changes and seasons and dates, which helps to improve the prediction accuracy of smoking quality; further, the historical climate data matrix is ​​processed according to the pre-trained smoking quality prediction model to predict the predicted smoking quality data corresponding to the tobacco produced in the tobacco production area to be predicted within the historical period, which solves the problems of low efficiency, low accuracy, and labor and material costs in related technologies, and achieves the effect of predicting the smoking quality of tobacco leaves based on the neural network model and historical climate data of climate characteristics without obtaining actual tobacco leaf samples, greatly improving the timeliness of tobacco leaf smoking quality prediction, improving the prediction efficiency and prediction accuracy of tobacco leaf smoking quality, and providing corresponding guidance for early decision-making in the tobacco industry to a certain extent.

[0082] Optionally, the multiple target climate characteristics include daily maximum temperature, daily average temperature, daily average atmospheric pressure, daily total precipitation, daily average net sunshine intensity and daily average wind speed; the predicted smoking quality data include quality parameters corresponding to multiple smoking quality indicators to be predicted; the multiple smoking quality indicators to be predicted include fragrance, sweet fragrance, caramel fragrance, concentration, aroma quality, aroma quantity, miscellaneous smell, stimulation, aftertaste and total score.

[0083] Optionally, the data acquisition module 410 includes: a climate data acquisition unit, a vector construction unit and a matrix construction unit.

[0084] A climate data acquisition unit, used to acquire historical climate data corresponding to multiple target climate characteristics of the tobacco production area to be predicted within a historical period;

[0085] A vector construction unit, for arranging the historical climate data corresponding to the target climate characteristics in chronological order for a plurality of the target climate characteristics, to obtain a historical climate data vector corresponding to the target climate characteristics;

[0086] The matrix construction unit is used to concatenate the historical climate data vectors corresponding to the target climate characteristics to obtain a historical climate data matrix.

[0087] Optionally, the smoking quality prediction model includes a convolutional neural network, a long short-term memory network and a fully connected layer; the smoking quality data prediction module 420 includes: a feature extraction unit, a smoking quality feature prediction unit and a smoking quality data determination unit.

[0088] A feature extraction unit is used to extract features from the historical climate data matrix according to the convolutional neural network to obtain a historical climate feature matrix; a smoking quality feature prediction unit is used to process the historical climate feature matrix according to the long short-term memory network to predict predicted smoking quality features; a smoking quality data determination unit is used to process the predicted smoking quality features according to the fully connected layer to obtain predicted smoking quality data corresponding to the tobacco produced in the tobacco producing area to be predicted within the historical period.

[0089] Optionally, the device further includes: a curve generating module and a correlation graph generating module.

[0090] A curve generating module, used for generating a predicted smoking quality curve according to the predicted smoking quality data;

[0091] The association diagram generation module is used to generate a climate smoking quality association diagram corresponding to the predicted tobacco leaf producing area within the historical period according to the predicted smoking quality curve and the historical climate data matrix.

[0092] Optionally, the device also includes: a model training module.

[0093] Model training module, used to train and obtain the smoking quality prediction model;

[0094] The model training module includes: a training sample acquisition unit and a model training unit.

[0095] A model output data determination unit, a model to be verified determination unit and a model verification unit.

[0096] A training sample acquisition unit, used to acquire a plurality of training samples, wherein the training samples include a sample climate data matrix corresponding to a sample tobacco leaf producing area within a historical period and real smoking quality data of tobacco leaves produced in the sample tobacco leaf producing area within the historical period;

[0097] The model training unit is used to train the pre-built deep learning model based on the multiple training samples to obtain a smoking quality prediction model.

[0098] Optionally, the model training unit includes: a model output data determination subunit, a model to be verified determination subunit and a model verification subunit.

[0099] A model output data determination subunit is used for inputting the sample climate data matrix in the training samples into a pre-built deep learning model for a plurality of the training samples to obtain output evaluation quality data corresponding to the training samples;

[0100] A subunit for determining a model to be verified is used to determine a loss value according to the output inhalation quality evaluation data and the real inhalation quality evaluation data in the training sample, and to modify the model parameters in the deep learning model according to the loss value, and to take the convergence of the loss function in the deep learning model as a training target to obtain at least one model to be verified;

[0101] The model verification subunit is used to verify at least one of the models to be verified based on the test samples, and use the model to be verified whose verification result meets the preset conditions as the smoking quality prediction model.

[0102] The tobacco leaf smoking quality prediction device provided in the embodiment of the present invention can execute the tobacco leaf smoking quality prediction method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0103] Embodiment 5

[0104] Figure 6A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0105] like Figure 6 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0106] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0107] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the tobacco leaf smoking quality prediction method.

[0108] In some embodiments, the tobacco leaf smoking quality prediction method can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the tobacco leaf smoking quality prediction method described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to execute the tobacco leaf smoking quality prediction method in any other appropriate manner (for example, by means of firmware).

[0109] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0110] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0111] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0112] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0113] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a target blockchain network, and the Internet.

[0114] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0115] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0116] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for predicting tobacco leaf smoking quality, characterized in that: include: Obtaining a historical climate data matrix corresponding to the tobacco production area to be predicted within a historical period; wherein the historical climate data matrix includes historical climate data corresponding to a plurality of target climate characteristics; The historical climate data matrix is ​​processed according to the pre-trained smoking quality prediction model to predict the smoking quality data corresponding to the tobacco produced in the tobacco production area to be predicted within the historical period; Among them, the smoking quality prediction model is used to predict the smoking quality data of tobacco produced in the tobacco producing area within the historical period according to the historical climate data of the tobacco producing area within the historical period; the smoking quality prediction model is based on the sample climate data matrix of the sample tobacco producing area within the historical period and the real smoking quality data of the tobacco produced in the sample tobacco producing area within the historical period. The pre-constructed deep learning model is trained.

2. The tobacco leaf smoking quality prediction method according to claim 1, characterized in that: The multiple target climate characteristics include daily maximum temperature, daily average temperature, daily average atmospheric pressure, daily total precipitation, daily average net sunshine intensity and daily average wind speed; the predicted smoking quality evaluation data include quality parameters corresponding to multiple smoking quality evaluation indicators to be predicted; the multiple smoking quality evaluation indicators to be predicted include fragrance, sweet fragrance, burnt fragrance, concentration, aroma quality, aroma quantity, miscellaneous smell, stimulation, aftertaste and total score.

3. The method for predicting tobacco leaf smoking quality according to claim 1, characterized in that: The step of obtaining the historical climate data matrix corresponding to the tobacco production area to be predicted within the historical period includes: Obtain historical climate data corresponding to multiple target climate characteristics of the tobacco production area to be predicted within a historical period; For a plurality of the target climate characteristics, the historical climate data corresponding to the target climate characteristics are arranged in chronological order to obtain a historical climate data vector corresponding to the target climate characteristics; The historical climate data vectors corresponding to the plurality of target climate characteristics are concatenated to obtain a historical climate data matrix.

4. The method for predicting tobacco leaf smoking quality according to claim 1, characterized in that: The smoking quality prediction model includes a convolutional neural network, a long short-term memory network and a fully connected layer; the smoking quality prediction model obtained by pre-training processes the historical climate data matrix to predict the smoking quality data corresponding to the tobacco produced in the predicted tobacco production area within the historical period, including: Extracting features from the historical climate data matrix according to the convolutional neural network to obtain a historical climate feature matrix; Processing the historical climate feature matrix according to the long short-term memory network to predict and obtain the predicted evaluation quality features; The predicted smoking quality features are processed according to the fully connected layer to obtain predicted smoking quality data corresponding to the tobacco produced in the predicted tobacco producing area within the historical period.

5. The method for predicting tobacco leaf smoking quality according to claim 1, characterized in that: Also includes: Generating a predicted smoking quality curve according to the predicted smoking quality data; According to the predicted smoking quality curve and the historical climate data matrix, a climate smoking quality association diagram corresponding to the predicted tobacco production area within the historical period is generated.

6. The method for predicting tobacco leaf smoking quality according to claim 1, characterized in that: Also includes: The training results in a prediction model of the smoking quality; The training obtains a smoking quality prediction model, including: Acquire multiple training samples, wherein the training samples include a sample climate data matrix corresponding to a sample tobacco leaf producing area within a historical period and real smoking quality data of tobacco leaves produced in the sample tobacco leaf producing area within the historical period; The pre-built deep learning model is trained based on the plurality of training samples to obtain a smoking quality prediction model.

7. The method for predicting tobacco leaf smoking quality according to claim 6, characterized in that: The pre-built deep learning model is trained based on the plurality of training samples to obtain a smoking quality prediction model, including: For a plurality of the training samples, the sample climate data matrix in the training samples is input into a pre-built deep learning model to obtain output suction quality data corresponding to the training samples; Determine a loss value according to the output inhalation quality data and the real inhalation quality data in the training sample, modify the model parameters in the deep learning model according to the loss value, take the convergence of the loss function in the deep learning model as the training goal, and obtain at least one model to be verified; At least one of the models to be verified is verified based on the test sample, and the model to be verified whose verification result meets the preset conditions is used as the smoking quality prediction model.

8. A tobacco leaf smoking quality prediction device, characterized in that: include: A data acquisition module is used to acquire a historical climate data matrix corresponding to the tobacco production area to be predicted within a historical period; wherein the historical climate data matrix includes historical climate data corresponding to multiple target climate characteristics; The smoking quality data prediction module is used to process the historical climate data matrix according to the pre-trained smoking quality prediction model, and predict the predicted smoking quality data corresponding to the tobacco produced in the predicted tobacco producing area within the historical period; wherein the smoking quality prediction model is used to predict the smoking quality data of the tobacco produced in the tobacco producing area within the historical period according to the historical climate data of the tobacco producing area within the historical period; the smoking quality prediction model is trained on a pre-constructed deep learning model based on the sample climate data matrix of the sample tobacco producing area within the historical period and the real smoking quality data of the tobacco produced in the sample tobacco producing area within the historical period.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the tobacco leaf smoking quality prediction method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the tobacco leaf smoking quality prediction method according to any one of claims 1 to 7 when executed.

11. A computer program product, characterized in that The computer program product comprises a computer program, which, when executed by a processor, implements the tobacco leaf smoking quality prediction method according to claims 1-7.

Images