Tobacco leaf sensory quality information prediction method, device, equipment, medium and product

Abstract

The embodiment of the invention discloses a tobacco leaf sensory quality information prediction method and device, equipment, a medium and a product. The method comprises the following steps: acquiring near infrared spectrum information of tobacco leaves to be detected; inputting the near infrared spectrum information into a pre-trained sensory quality prediction model corresponding to a preset quality index to obtain a quality index characterization value corresponding to the preset quality index in the combustion process of the tobacco leaves to be detected; and on the basis of the quality index characterization value corresponding to the preset quality index, predicting sensory quality information of the to-be-detected tobacco leaves in the combustion process. According to the technical scheme, actual combustion of the tobacco leaves is not needed, the sensory quality information of the tobacco leaves to be detected can be predicted only through the sensory quality prediction model, the efficiency and accuracy of determining the sensory quality information are improved, and the labor cost is effectively reduced.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of information processing technology, and in particular to a method, device, equipment, medium and product for predicting sensory quality information of tobacco leaves. Background Art

[0002] The stability of cigarette products mainly depends on the stability of the quality of tobacco leaf raw materials used by cigarette companies in production.

[0003] In the prior art, the quality of tobacco raw materials is usually determined by manual evaluation by formula personnel. However, the manual evaluation method increases the workload of formula personnel, takes a long time to determine the quality of tobacco raw materials, is inefficient, and increases labor costs; in addition, the manual evaluation method is highly subjective, resulting in poor accuracy of the determination results. Summary of the invention

[0004] The embodiments of the present invention provide a method, device, equipment, medium and product for predicting sensory quality information of tobacco leaves, so as to improve the efficiency and accuracy of determining sensory quality information and effectively reduce labor costs.

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

[0006] Acquire near infrared spectrum information of the tobacco leaves to be tested;

[0007] Inputting the near infrared spectrum information into a pre-trained sensory quality prediction model corresponding to a preset quality index, and obtaining a quality index characterization value corresponding to the preset quality index during the combustion process of the tobacco leaf to be tested;

[0008] Based on the quality indicator characterization value corresponding to the preset quality indicator, the sensory quality information of the tobacco leaf to be tested during the combustion process is predicted.

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

[0010] An information acquisition module is used to obtain near-infrared spectrum information of the tobacco leaves to be tested;

[0011] A prediction value determination module, used for inputting the near infrared spectrum information into a pre-trained sensory quality prediction model corresponding to a preset quality index, and obtaining a quality index characterization value corresponding to the preset quality index during the combustion process of the tobacco leaf to be tested;

[0012] The information determination module is used to predict the sensory quality information of the tobacco leaf to be tested during the combustion process based on the quality indicator characterization value corresponding to the preset quality indicator.

[0013] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0014] at least one processor; and

[0015] a memory communicatively connected to the at least one processor; wherein,

[0016] 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 sensory quality information prediction method described in any embodiment of the present invention.

[0017] 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 sensory quality information prediction method described in any embodiment of the present invention when executed.

[0018] According to another aspect of the present invention, there is provided a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method for predicting tobacco sensory quality information according to any embodiment of the present invention is implemented.

[0019] The technical solution of the embodiment of the present invention obtains the near-infrared spectrum information of the tobacco leaves to be tested, and inputs the near-infrared spectrum information into a pre-trained sensory quality prediction model corresponding to the preset quality index, so as to obtain the quality index characterization value corresponding to the preset quality index of the tobacco leaves to be tested during the combustion process, thereby digitally reflecting the sensory quality of the tobacco leaves to be tested through the quality index characterization value; finally, based on the quality index characterization value corresponding to the preset quality index, the sensory quality information of the tobacco leaves to be tested during the combustion process is predicted. It can be seen that the technical solution of this embodiment does not require human participation, nor does it require the actual burning of tobacco leaves. The sensory quality information of the tobacco leaves to be tested can be predicted only through the sensory quality prediction model, which improves the efficiency and accuracy of determining the sensory quality information and effectively reduces labor costs.

[0020] 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

[0021] 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.

[0022] Figure 1 is a flow chart of a tobacco leaf sensory quality information prediction method provided according to an embodiment of the present invention;

[0023] Figure 2 is a flow chart of another tobacco leaf sensory quality information prediction method provided according to an embodiment of the present invention;

[0024] Figure 3 is a schematic diagram of model training provided according to an embodiment of the present invention;

[0025] Figure 4 is a working schematic diagram of a sensory quality prediction model provided according to an embodiment of the present invention;

[0026] Figure 5 is a schematic diagram of the structure of a tobacco sensory quality information prediction device provided according to an embodiment of the present invention;

[0027] Figure 6 It is a structural schematic diagram of an electronic device for implementing the tobacco leaf sensory quality information prediction method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0028] 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.

[0029] 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 "etc." 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.

[0030] It should be noted that the collection, collection, updating, analysis, processing, use, transmission, storage and other aspects of user personal information involved in the technical solution of this disclosure are in compliance with the provisions of relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. Necessary measures are taken for user personal information to prevent illegal access to user personal information data and maintain the security of user personal information and network security.

[0031] Figure 1 The present invention provides a flowchart of a method for predicting sensory quality information of tobacco leaves according to an embodiment of the present invention. This embodiment is applicable to the case of determining sensory quality information of tobacco leaves. The method can be performed by a device for predicting sensory quality information of tobacco leaves, and the device for predicting sensory quality information of tobacco leaves can be implemented in the form of hardware and / or software.

[0032] like Figure 1 As shown, the method of this embodiment may specifically include:

[0033] S110, obtaining near-infrared spectrum information of the tobacco leaves to be tested.

[0034] The tobacco leaves to be tested are tobacco leaves whose sensory quality needs to be predicted.

[0035] In this embodiment, the near-infrared spectrum information of the tobacco leaves to be tested can be collected by the near-infrared spectrum analysis technology of the near-infrared spectrometer. Among them, the near-infrared spectrum analysis technology is a fast, non-destructive, green and low-cost technology. Specifically, the tobacco leaves to be tested can be ground, and the near-infrared spectrum data of the tobacco leaves can be collected in diffuse reflection mode using a near-infrared spectrometer, and the near-infrared spectrum data can be used as near-infrared spectrum information. The near-infrared spectrum information can reflect the component characteristics of the tobacco leaves to be tested, for example, the component characteristics include the content of moisture, total sugar, salt and alkali, and total nitrogen.

[0036] S120, inputting the near infrared spectrum information into a pre-trained sensory quality prediction model corresponding to a preset quality index, and obtaining a quality index characterization value corresponding to the preset quality index during the combustion process of the tobacco leaf to be tested.

[0037] The preset quality indicators are used to evaluate the quality of the sensory experience produced by tobacco leaves during the combustion process. For example, the preset quality indicators include aroma, impurities, stimulation and aftertaste. Each preset quality indicator corresponds to a sensory quality prediction model, which can be a deep neural network model.

[0038] In a specific implementation, when there are multiple preset quality indicators, a sensory quality prediction model can be trained based on each preset quality indicator. The obtained near-infrared spectral information is respectively input into the sensory quality prediction model corresponding to each preset quality indicator, and based on the output information of each sensory quality prediction model, the quality indicator characterization value corresponding to each preset quality indicator of the tobacco leaves to be tested during the combustion process is obtained. It should be noted that the quality indicator characterization value corresponding to the preset quality indicator can reflect the degree of sensory experience corresponding to the preset quality indicator produced by the tobacco leaves to be tested. For example, when the preset quality indicator is aroma, the higher the quality indicator characterization value corresponding to the aroma, the stronger the aroma perception produced by the tobacco leaves to be tested during the combustion process, which further indicates that the sensory quality of the tobacco leaves to be tested is higher.

[0039] In this embodiment, the preset quality indicators include positive type indicators and negative type indicators; the near-infrared spectral information is input into a pre-trained sensory quality prediction model corresponding to the preset quality indicators to obtain the quality indicator characterization value corresponding to the preset quality indicators during the burning process of the tobacco leaves to be tested, including: the near-infrared spectral information is input into the sensory quality prediction model corresponding to each positive type indicator and the sensory quality prediction model corresponding to each negative type indicator to obtain the positive indicator characterization value corresponding to each positive type indicator and the negative indicator characterization value corresponding to each negative type indicator.

[0040] Among them, each preset quality indicator corresponds to a sensory quality prediction model.

[0041] It should be noted that a positive type indicator is an indicator that has a positive impact on the sensory experience, and a negative type indicator is an indicator that has a negative impact on the sensory experience. For example, a positive type indicator may be an aroma indicator, and the aroma of tobacco leaves can make the sensory experience more pleasant, which is a positive type indicator. A negative type indicator may be a stimulation indicator, and the stimulation brought by tobacco leaves causes discomfort to the senses, which is a negative type indicator.

[0042] Optionally, the positive type indicators include at least one of a light fragrance index, a sweet fragrance index, a caramel fragrance index, a concentration index, a strength index, an aroma quality index and an aroma quantity index; the negative type indicators include at least one of a miscellaneous smell index, a stimulation index and an aftertaste index.

[0043] In order to be able to comprehensively evaluate the quality of the sensory experience produced by the tobacco leaves to be tested, the preset quality indicators include both positive type indicators and negative type indicators. Specifically, the near-infrared spectral information can be input into the sensory quality prediction model corresponding to each positive type indicator to obtain the positive indicator characterization value corresponding to the positive type indicator. The higher the positive indicator characterization value, the higher the sensory quality of the tobacco leaves to be tested; the lower the positive indicator characterization value, the lower the sensory quality of the tobacco leaves to be tested. The near-infrared spectral information is input into the sensory quality prediction model corresponding to each negative type indicator to obtain the negative indicator characterization value corresponding to the negative type indicator. The lower the negative indicator characterization value, the higher the sensory quality of the tobacco leaves to be tested; the higher the negative indicator characterization value, the lower the sensory quality of the tobacco leaves to be tested.

[0044] This embodiment determines the positive index characterization value and the negative index characterization value respectively, so as to perform a multi-index evaluation on the tobacco leaves to be tested from both positive and negative aspects, which can more comprehensively and effectively reflect the sensory quality of the tobacco leaves to be tested.

[0045] S130. Predicting the sensory quality information of the tobacco leaf to be tested during the combustion process based on the quality indicator characterization value corresponding to the preset quality indicator.

[0046] The sensory quality information indicates the quality of the sensory experience generated by the tobacco leaves to be tested during the combustion process. Exemplarily, the sensory quality information includes at least one of excellent quality, medium quality and poor quality.

[0047] In this embodiment, a preset quality range corresponding to the preset quality indicator may be preset. Exemplarily, the preset quality range includes a first range, a second range, and a third range. The first range, the second range, and the third range do not overlap. The minimum value in the first range is greater than the maximum value in the second range; the minimum value in the second range is greater than the maximum value in the third range.

[0048] In the case where there are multiple preset quality indicators, when the quality indicator characterization values ​​corresponding to each preset quality indicator belong to the first range, the sensory quality information of the tobacco leaves to be tested during the combustion process can be determined to be of excellent quality. When there is at least one preset quality indicator whose quality indicator characterization value belongs to the third range, the sensory quality information of the tobacco leaves to be tested during the combustion process can be determined to be of poor quality. When there is at least one preset quality indicator whose quality indicator characterization value belongs to the second range, and there is no quality indicator characterization value belonging to the third range, the sensory quality information of the tobacco leaves to be tested during the combustion process can be determined to be of medium quality.

[0049] Alternatively, based on the quality indicator characterization value corresponding to the preset quality indicator, the implementation method of predicting the sensory quality information of the tobacco leaves to be tested during the burning process may include: determining the average value of each quality indicator characterization value when each quality indicator characterization value corresponding to the preset quality indicator satisfies the preset evaluation condition corresponding to the preset quality indicator; and determining the sensory quality information of the tobacco leaves to be tested during the burning process based on the average value.

[0050] Specifically, different preset evaluation conditions may be set for different preset quality indicators. For example, the preset quality indicators include a sweet aroma indicator, a caramel aroma indicator, and a concentration indicator. The preset evaluation condition corresponding to the sweet aroma indicator is: the quality indicator characterization value is greater than a first threshold. The preset evaluation condition corresponding to the caramel aroma indicator is: the quality indicator characterization value is greater than a second threshold. The preset evaluation condition corresponding to the strong aroma indicator is: the quality indicator characterization value is greater than a third threshold.

[0051] In practical applications, for each preset quality indicator, it can be determined whether the quality indicator characterization value corresponding to the preset quality indicator meets the corresponding preset evaluation condition. If there is a quality indicator characterization value corresponding to the preset quality indicator that does not meet the corresponding preset evaluation condition, the sensory quality information can be directly determined to be poor quality.

[0052] If the quality indicator characterization values ​​corresponding to each preset quality indicator meet the matching preset evaluation conditions, in order to comprehensively consider the experience of the tobacco leaves to be tested for each preset quality indicator, the average value of each quality indicator characterization value can be determined, and based on the average value, the quality of the overall sensory experience produced by the tobacco leaves to be tested during the burning process can be determined.

[0053] For example, the fourth threshold and the fifth threshold may be set based on the requirements for the sensory quality of tobacco leaves. The fourth threshold is greater than the fifth threshold. When the average value is greater than the fourth threshold, the sensory quality information is determined to be of excellent quality. When the average value is greater than the fifth threshold and less than or equal to the fourth threshold, the sensory quality information is determined to be of medium quality. When the average value is less than or equal to the fifth threshold, the sensory quality information is determined to be of poor quality.

[0054] This embodiment determines the sensory quality information by the average value of the quality indicator characterization value, comprehensively considers the sensory experience of the tobacco leaves to be tested for different preset quality indicators, and is conducive to improving the accuracy and effectiveness of the sensory quality information.

[0055] The technical solution of the embodiment of the present invention obtains the near-infrared spectrum information of the tobacco leaves to be tested, and inputs the near-infrared spectrum information into a pre-trained sensory quality prediction model corresponding to the preset quality index, so as to obtain the quality index characterization value corresponding to the preset quality index of the tobacco leaves to be tested during the combustion process, thereby digitally reflecting the sensory quality of the tobacco leaves to be tested through the quality index characterization value; finally, based on the quality index characterization value corresponding to the preset quality index, the sensory quality information of the tobacco leaves to be tested during the combustion process is predicted. It can be seen that the technical solution of this embodiment does not require human participation, nor does it require the actual burning of tobacco leaves. The sensory quality information of the tobacco leaves to be tested can be predicted only through the sensory quality prediction model, which improves the efficiency and accuracy of determining the sensory quality information and effectively reduces labor costs.

[0056] Figure 2 1 is a flowchart of another tobacco sensory quality information prediction method provided according to an embodiment of the present invention. Based on the above embodiment, this embodiment optionally further includes: training the sensory quality prediction model to be trained corresponding to the preset quality index before inputting the near infrared spectrum information into the pre-trained sensory quality prediction model corresponding to the preset quality index. The explanations of the terms that are the same or corresponding to the above embodiments are not repeated here. Figure 2 As shown, the method includes:

[0057] S210, obtaining a training sample set; wherein the training sample set includes a plurality of training samples, each training sample includes sample spectral information of a sample tobacco leaf, and a true sensory quality value for a preset quality index during the burning process of the sample tobacco leaf.

[0058] In a specific implementation, near-infrared spectrum information of sample tobacco leaves can be collected by a near-infrared spectrometer as sample spectrum information. In addition, the evaluation results of the sample tobacco leaves by the evaluation staff are used as the real sensory quality value of the sample tobacco leaves during the combustion process. For example, for different sample tobacco leaves, the evaluation staff evaluates the sample tobacco leaves to obtain the real sensory quality value of the sample tobacco leaves corresponding to different preset quality indicators.

[0059] Figure 3 is a schematic diagram of a model training provided according to an embodiment of the present invention, such as Figure 3 As shown in the figure. The information of tobacco leaf samples is collected by near infrared spectrometer to obtain the sample spectrum information; the same sample tobacco leaves are subjected to manual sensory evaluation to obtain the real sensory quality value. The sample spectrum information and real sensory quality value corresponding to each sample tobacco leaf are used as a training sample. The training sample set is composed of multiple training samples.

[0060] S220. Based on the sample spectral information and the true sensory quality value of each sample tobacco leaf, the to-be-trained sensory quality prediction model corresponding to the preset quality index is trained, and the convergence of the loss function is used as the training goal to obtain a trained sensory quality prediction model corresponding to the preset quality index.

[0061] The sensory quality prediction model to be trained may be a vision transformer model (Vision Transformer).

[0062] In a specific implementation, the sensory quality prediction model to be trained can be trained based on each training sample in the training sample set. Figure 3 As shown, 80% of the training samples in the training sample set can be used as the training set; 20% of the training samples in the training sample set can be used as the validation set. The sensory quality prediction model to be trained is trained by the training samples in the training set, and the convergence of the loss function is used as the training goal, so as to obtain a sensory quality prediction model corresponding to the preset quality index.

[0063] S230, obtaining near-infrared spectrum information of the tobacco leaves to be tested.

[0064] S240, inputting the near infrared spectrum information into a pre-trained sensory quality prediction model corresponding to a preset quality index, and obtaining a quality index characterization value corresponding to the preset quality index during the combustion process of the tobacco leaf to be tested.

[0065] Optionally, the sensory quality prediction model includes a visual converter model; the visual converter model includes a block embedding module, a prediction tag module, a position embedding module, a random inactivation module, a conversion encoder, an extraction prediction tag module and a multi-layer perceptron.

[0066] Figure 4 is a working schematic diagram of a sensory quality prediction model provided according to an embodiment of the present invention. Figure 4 As shown in the figure, the near infrared spectral information can be data with a shape of (16, 1, 2125), where 16 is the batch size, 1 indicates the number of channels, and 2125 is the feature dimension of the spectral data. After the near infrared spectral information is input into the sensory quality prediction model, it first enters the patch embedding module, passes through a one-dimensional convolution layer and a linear layer, and the convolution kernel shape of the convolution layer can be (1, 25), the step size is 25, and the number of output channels is 25. The first feature map with a shape of (16, 85, 25) is output through the patch embedding module.

[0067] Furthermore, the first feature map is input to the prediction tag module (Class Token), which adds a special classification tag to the sequence dimension of the feature map by adding a learnable vector at the beginning or end of the sequence. The output shape is a second feature map of (16, 86, 255).

[0068] Then, the second feature map output by the prediction tag module is processed by the position embedding module (PositionEmbedding), and the third feature map after position embedding is output. The third feature map is passed through the random deactivation module to reduce overfitting and improve the generalization ability of the model. The output information of the random deactivation module is processed by the Transformer Encoder (Transformer Encoder) and the stacked 8 times. Each transformation encoder contains a multi-head self-attention mechanism and a feedforward neural network.

[0069] Finally, the output information of the conversion encoder is input to the Extract ClassToken module to extract the feature vector of the classification token. The shape of the output information is a set of feature vectors of (16, 25), each of which corresponds to the classification token feature of a sample in a batch. In addition, the feature vector set is input to the multi-layer perceptron, which is first normalized and then mapped to the final prediction output, i.e., the quality indicator representation value, through a linear layer.

[0070] S250. Predicting the sensory quality information of the tobacco leaf to be tested during the combustion process based on the quality indicator characterization value corresponding to the preset quality indicator.

[0071] On the one hand, this embodiment reduces the workload of formula testers and reduces the health damage of the testers; on the other hand, it also reduces the difficulty of formula maintenance work to a certain extent and reduces the experience requirements of formula maintenance work for formula staff, so that tobacco resources can be utilized more effectively and the efficiency of cigarette formula maintenance work can be improved.

[0072] Figure 5 1 is a schematic diagram of the structure of a tobacco sensory quality information prediction device provided according to an embodiment of the present invention, and the device is used to execute the tobacco sensory quality information prediction method provided in any of the above embodiments. The device and the tobacco sensory quality information prediction method of the above embodiments belong to the same inventive concept, and the details not described in detail in the embodiments of the tobacco sensory quality information prediction device can refer to the embodiments of the above tobacco sensory quality information prediction method. Figure 5 As shown, the device comprises:

[0073] An information acquisition module 10 is used to acquire near infrared spectrum information of the tobacco leaves to be tested;

[0074] The prediction value determination module 11 is used to input the near infrared spectrum information into a pre-trained sensory quality prediction model corresponding to the preset quality index, and obtain the quality index characterization value corresponding to the preset quality index during the combustion process of the tobacco leaves to be tested;

[0075] The information determination module 12 is used to predict the sensory quality information of the tobacco leaves to be tested during the combustion process based on the quality indicator characterization value corresponding to the preset quality indicator.

[0076] On the basis of any optional technical solution in the embodiment of the present invention, optionally, the preset quality indicator includes a positive type indicator and a negative type indicator; the prediction value determination module 11 includes:

[0077] An information input unit, used to input the near infrared spectrum information into the sensory quality prediction model corresponding to each positive type indicator and the sensory quality prediction model corresponding to each negative type indicator, respectively, to obtain the positive indicator representation value corresponding to each positive type indicator and the negative indicator representation value corresponding to each negative type indicator;

[0078] Among them, each preset quality indicator corresponds to a sensory quality prediction model. The higher the positive indicator representation value, the higher the sensory quality of the tobacco leaves to be tested; the lower the negative indicator representation value, the higher the sensory quality of the tobacco leaves to be tested.

[0079] Based on any optional technical scheme in the embodiments of the present invention, optionally, the positive type index includes at least one of a light fragrance index, a sweet fragrance index, a caramel fragrance index, a concentration index, a strength index, an aroma quality index and an aroma quantity index; the negative type index includes at least one of a miscellaneous smell index, a stimulation index and an aftertaste index.

[0080] Based on any optional technical solution in the embodiment of the present invention, optionally, the information determination module 12 includes:

[0081] an average value determining unit, configured to determine an average value of each quality indicator characterization value when the quality indicator characterization value corresponding to each preset quality indicator satisfies the preset evaluation condition corresponding to the preset quality indicator;

[0082] The information determination unit is used to determine the sensory quality information of the tobacco leaves to be tested during the combustion process based on the average value; wherein the sensory quality information includes at least one of excellent quality, medium quality and poor quality.

[0083] Based on any optional technical solution in the embodiments of the present invention, optionally, the method further includes:

[0084] A training sample set acquisition module is used to acquire a training sample set before inputting the near infrared spectrum information into a pre-trained sensory quality prediction model corresponding to a preset quality index; wherein the training sample set includes a plurality of training samples, each training sample includes sample spectrum information of a sample tobacco leaf, and a true sensory quality value for a preset quality index during the combustion process of the sample tobacco leaf;

[0085] The model training module is used to train the sensory quality prediction model to be trained corresponding to the preset quality index based on the sample spectral information and the true sensory quality value of each sample tobacco leaf, take the convergence of the loss function as the training goal, and obtain the trained sensory quality prediction model corresponding to the preset quality index.

[0086] Based on any optional technical solution in the embodiments of the present invention, optionally, the sensory quality prediction model includes a visual converter model; the visual converter model includes a block embedding module, a prediction tag module, a position embedding module, a random inactivation module, a conversion encoder, an extraction prediction tag module and a multi-layer perceptron.

[0087] The technical solution of the embodiment of the present invention obtains the near-infrared spectrum information of the tobacco leaves to be tested, and inputs the near-infrared spectrum information into a pre-trained sensory quality prediction model corresponding to the preset quality index, so as to obtain the quality index characterization value corresponding to the preset quality index of the tobacco leaves to be tested during the combustion process, thereby digitally reflecting the sensory quality of the tobacco leaves to be tested through the quality index characterization value; finally, based on the quality index characterization value corresponding to the preset quality index, the sensory quality information of the tobacco leaves to be tested during the combustion process is predicted. It can be seen that the technical solution of this embodiment does not require human participation, nor does it require the actual burning of tobacco leaves. The sensory quality information of the tobacco leaves to be tested can be predicted only through the sensory quality prediction model, which improves the efficiency and accuracy of determining the sensory quality information and effectively reduces labor costs.

[0088] It is worth noting that in the embodiment of the above-mentioned tobacco sensory quality information prediction device, the various units and modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0089] Figure 6It is a structural schematic diagram of an electronic device for implementing the tobacco sensory quality information prediction method of an embodiment of the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, 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 only examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0090] like Figure 6 As shown, the electronic device 20 includes at least one processor 21, and a memory connected to the at least one processor 21, such as a read-only memory (ROM) 22, a random access memory (RAM) 23, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 21 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 22 or the computer program loaded from the storage unit 28 to the random access memory (RAM) 23. In RAM23, various programs and data required for the operation of the electronic device 20 can also be stored. The processor 21, ROM22 and RAM23 are connected to each other through a bus 24. An input / output (I / O) interface 25 is also connected to the bus 24.

[0091] A number of components in the electronic device 20 are connected to the I / O interface 25, including: an input unit 26, such as a keyboard, a mouse, etc.; an output unit 27, such as various types of displays, speakers, etc.; a storage unit 28, such as a disk, an optical disk, etc.; and a communication unit 29, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 29 allows the electronic device 20 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0092] The processor 21 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 21 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 21 executes the various methods and processes described above, such as the tobacco sensory quality information prediction method.

[0093] In some embodiments, the method for predicting tobacco sensory quality information may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 28. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 20 via the ROM 22 and / or the communication unit 29. When the computer program is loaded into the RAM 23 and executed by the processor 21, one or more steps of the method for predicting tobacco sensory quality information described above may be performed. Alternatively, in other embodiments, the processor 21 may be configured to execute the method for predicting tobacco sensory quality information in any other appropriate manner (e.g., by means of firmware).

[0094] 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.

[0095] 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.

[0096] 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.

[0097] 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).

[0098] 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 blockchain network, and the Internet.

[0099] 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.

[0100] This embodiment also provides a computer program product, including a computer program, which, when executed by a processor, implements the tobacco sensory quality information prediction method provided in any embodiment of the present application.

[0101] In the process of implementation, the computer program product can be written in one or more programming languages ​​or a combination thereof to perform the computer program code of the present invention, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0102] 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.

[0103] 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 sensory quality information, characterized in that: include: Acquire near infrared spectrum information of the tobacco leaves to be tested; Inputting the near infrared spectrum information into a pre-trained sensory quality prediction model corresponding to a preset quality index, and obtaining a quality index characterization value corresponding to the preset quality index during the combustion process of the tobacco leaf to be tested; Based on the quality indicator characterization value corresponding to the preset quality indicator, the sensory quality information of the tobacco leaf to be tested during the combustion process is predicted.

2. The method according to claim 1, characterized in that The preset quality indicators include positive type indicators and negative type indicators; The step of inputting the near infrared spectrum information into a pre-trained sensory quality prediction model corresponding to a preset quality index to obtain a quality index characterization value corresponding to the preset quality index during the combustion of the tobacco leaf to be tested comprises: Inputting the near-infrared spectrum information into the sensory quality prediction model corresponding to each of the positive type indicators and the sensory quality prediction model corresponding to each of the negative type indicators, respectively, to obtain the positive indicator characterization value corresponding to each of the positive type indicators and the negative indicator characterization value corresponding to each of the negative type indicators; Among them, each preset quality indicator corresponds to a sensory quality prediction model, and the higher the positive indicator representation value, the higher the sensory quality of the tobacco leaves to be tested; the lower the negative indicator representation value, the higher the sensory quality of the tobacco leaves to be tested.

3. The method according to claim 2, characterized in that The positive type indicators include at least one of a light fragrance index, a sweet fragrance index, a caramel fragrance index, a concentration index, a strength index, an aroma quality index and an aroma quantity index; the negative type indicators include at least one of a miscellaneous smell index, a stimulation index and an aftertaste index.

4. The method according to claim 1, characterized in that: The predicting of the sensory quality information of the tobacco leaf to be tested during the combustion process based on the quality indicator characterization value corresponding to the preset quality indicator includes: When each of the quality indicator characterization values ​​corresponding to the preset quality indicator satisfies the preset evaluation condition corresponding to the preset quality indicator, determining an average value of each of the quality indicator characterization values; Based on the average value, the sensory quality information of the tobacco leaf to be tested during the combustion process is determined; wherein the sensory quality information includes at least one of excellent quality, medium quality and poor quality.

5. The method according to claim 1, characterized in that: Before inputting the near infrared spectrum information into a pre-trained sensory quality prediction model corresponding to a preset quality index, the method further includes: Acquire a training sample set; wherein the training sample set includes a plurality of training samples, each of the training samples includes sample spectral information of a sample tobacco leaf, and a true sensory quality value for the preset quality index during the combustion process of the sample tobacco leaf; Based on the sample spectral information and the true sensory quality value of each sample tobacco leaf, the to-be-trained sensory quality prediction model corresponding to the preset quality index is trained, and the convergence of the loss function is taken as the training goal to obtain a trained sensory quality prediction model corresponding to the preset quality index.

6. The method according to claim 1, characterized in that The sensory quality prediction model includes a visual converter model; the visual converter model includes a block embedding module, a prediction tag module, a position embedding module, a random inactivation module, a conversion encoder, an extraction prediction tag module and a multi-layer perceptron.

7. A tobacco leaf sensory quality information prediction device, characterized in that: include: An information acquisition module is used to obtain near-infrared spectrum information of the tobacco leaves to be tested; A prediction value determination module, used for inputting the near infrared spectrum information into a pre-trained sensory quality prediction model corresponding to a preset quality index, and obtaining a quality index characterization value corresponding to the preset quality index during the combustion process of the tobacco leaf to be tested; The information determination module is used to predict the sensory quality information of the tobacco leaf to be tested during the combustion process based on the quality indicator characterization value corresponding to the preset quality indicator.

8. 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 sensory quality information prediction method according to any one of claims 1 to 6.

9. 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 sensory quality information prediction method according to any one of claims 1 to 6 when executed.

10. A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method for predicting tobacco sensory quality information according to any one of claims 1 to 6 is implemented.

Images