Intelligent control system and method for automatic production of black tea
Through intelligent control systems and deep neural network technology, the fermentation temperature of black tea is monitored and controlled in real time, which solves the problem of lack of real-time perception and response in traditional black tea production, and improves production efficiency and quality.
Patent Information
- Application Number
- CN202411555646.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2023-12-04
- Filing Date
- 2024-11-04
- Publication Date
- 2024-11-29
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional black tea production control solutions lack real-time perception and response capabilities, and cannot flexibly adjust operating strategies, resulting in a decrease in production efficiency and consistency, and rely on manual observation and empirical judgment, which lacks accuracy and real-timeness.
The intelligent control system is adopted to pick tea through robots or drones, and the tea fermentation status and temperature is monitored and controlled in real time by using deep neural networks and one-dimensional convolutional layers to achieve adaptive adjustment of fermentation temperature.
It improves the quality and efficiency of black tea production, ensures the accuracy and consistency of the fermentation process, and reduces the influence of human factors.
Smart Images

Figure CN119045397A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of intelligent control, and more specifically, to an intelligent control system and method for automated production of black tea. Background Art
[0002] Black tea is a kind of tea made through the processes of withering, rolling, fermentation and drying. It has the characteristics of ruddy color, rich aroma and mellow taste. In the production process of black tea, fermentation is the most critical step, which determines the quality and flavor of black tea. The fermentation temperature is an important factor affecting the fermentation effect. If the temperature is too high or too low, the color, aroma, taste and shape of the tea will change, reducing the quality of black tea. Therefore, it is very important to monitor and control the fermentation temperature in real time during the fermentation process.
[0003] However, traditional black tea production control solutions are usually operated based on preset rules and parameters, lacking the ability to perceive and respond to the fermentation environment and tea status in real time, which results in the system's poor performance in dealing with complex production environments and tea changes, and inability to flexibly adjust operating strategies. In addition, traditional control solutions require manual operation and monitoring, and have limited capabilities in data processing and analysis. For example, during the fermentation process, the monitoring and analysis of tea fermentation status usually relies on manual observation and empirical judgment, lacks accuracy and real-time performance, and is easily affected by human factors, resulting in a decrease in production efficiency and consistency.
[0004] Therefore, an intelligent control solution for automated production of black tea is desired. Summary of the invention
[0005] In view of this, the present application proposes an intelligent control system and method for automated production of black tea, which can obtain high-quality black tea.
[0006] According to one aspect of the present application, an intelligent control method for automated production of black tea is provided, comprising: picking fresh tea leaves in a tea garden using a robot or a drone, and classifying and distributing the tea leaves according to their variety, grade and maturity to obtain classified tea leaves; sending the classified tea leaves into a withering machine, and subjecting the classified tea leaves to heat treatment with high-temperature steam or hot air to obtain withered tea leaves; sending the withered tea leaves into a rolling machine, and subjecting the withered tea leaves to squeezing, kneading and twisting operations with the mechanical force of the rolling machine to obtain rolled tea leaves; sending the rolled tea leaves into a fermentation machine, and controlling the temperature of the fermentation machine in real time The polyphenols in the rolled tea leaves undergo an oxidation reaction to form red pigments and aroma substances to obtain fermented tea leaves, wherein the fermented tea leaves are red; the fermented tea leaves are sent to a dryer to dry the fermented tea leaves by hot air or microwaves to reduce the moisture content of the fermented tea leaves to fix the color, aroma, taste and shape of the tea leaves to obtain dried tea leaves; the dried tea leaves are sent to a screening machine to grade the dried tea leaves by screens of different sizes to obtain screened tea leaves; and the screened tea leaves are sent to a packaging machine to seal and package the screened tea leaves by vacuum or nitrogen filling to obtain a finished black tea product.
[0007] According to another aspect of the present application, an intelligent control system for automated production of black tea is provided, which includes: a picking and sorting module, for picking fresh tea leaves in a tea garden using a robot or an unmanned aerial vehicle, and sorting and distributing the tea leaves according to their variety, grade and maturity to obtain sorted tea leaves; a heat treatment module, for feeding the sorted tea leaves into a withering machine, and heat-treating the sorted tea leaves by high-temperature steam or hot air to obtain withered tea leaves; a rolling module, for feeding the withered tea leaves into a rolling machine, and squeezing, kneading and twisting the withered tea leaves by the mechanical force of the rolling machine to obtain rolled tea leaves; a fermentation module, for feeding the rolled tea leaves into a fermentation machine, and controlling the fermentation machine in real time to obtain rolled tea leaves. The temperature of the fermentation machine is used to cause the polyphenols in the rolled tea leaves to undergo an oxidation reaction to form red pigments and aroma substances to obtain fermented tea leaves, wherein the fermented tea leaves are red; a drying module is used to send the fermented tea leaves into a dryer, dry the fermented tea leaves by hot air or microwaves to reduce the moisture of the fermented tea leaves to fix the color, aroma, taste and shape of the tea leaves to obtain dried tea leaves; a grading module is used to send the dried tea leaves into a screening machine, grade the dried tea leaves through sieves of different sizes to obtain screened tea leaves; and a sealing and packaging module is used to send the screened tea leaves into a packaging machine, seal and package the screened tea leaves by vacuum or nitrogen filling to obtain a finished black tea product.
[0008] According to the embodiment of the present application, the tea leaves are first classified and distributed according to the variety, grade and maturity of the tea leaves to obtain classified tea leaves, then the classified tea leaves are heat treated to obtain the tea leaves after killing green, then the tea leaves after killing green are squeezed, rubbed and twisted to obtain the tea leaves after twisting, then the temperature is controlled in real time to make the polyphenols in the tea leaves after twisting undergo oxidation reaction to form red pigments and aroma substances to obtain fermented tea leaves, then the fermented tea leaves are dried to obtain dried tea leaves, then the dried tea leaves are graded to obtain sieved tea leaves, and finally the sieved tea leaves are sealed and packaged to obtain finished black tea. In this way, high-quality black tea can be obtained.
[0009] Other features and aspects of the present application will become apparent from the following detailed description of exemplary embodiments with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0010] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate exemplary embodiments, features, and aspects of the present application and, together with the description, serve to explain the principles of the present application.
[0011] Figure 1 A flow chart showing an intelligent control method for automated production of black tea according to an embodiment of the present application.
[0012] Figure 2 A flowchart of sub-step S140 of the intelligent control method for automated production of black tea according to an embodiment of the present application is shown.
[0013] Figure 3 A schematic diagram of the architecture of sub-step S140 of the intelligent control method for automated production of black tea according to an embodiment of the present application is shown.
[0014] Figure 4 A block diagram of an intelligent control system for automated production of black tea according to an embodiment of the present application is shown.
[0015] Figure 5 An application scenario diagram of an intelligent control method for automated production of black tea according to an embodiment of the present application is shown. DETAILED DESCRIPTION
[0016] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work also fall within the scope of protection of the present application.
[0017] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.
[0018] Various exemplary embodiments, features and aspects of the present application will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.
[0019] In addition, in order to better illustrate the present application, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present application can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the subject matter of the present application.
[0020] In view of the above technical problems, in the technical solution of this application, an intelligent control method for automated production of black tea is proposed. Figure 1 The flowchart of the intelligent control method for automatic production of black tea according to the embodiment of the present application is shown. Figure 1 As shown, according to the embodiment of the present application, the intelligent control method for automated production of black tea comprises the following steps: S110, picking fresh tea leaves in a tea garden using a robot or an unmanned aerial vehicle, and classifying and distributing the tea leaves according to their variety, grade and maturity to obtain classified tea leaves; S120, sending the classified tea leaves into a fixing machine, and heat-treating the classified tea leaves by high-temperature steam or hot air to obtain fixed tea leaves; S130, sending the fixed tea leaves into a rolling machine, and squeezing, kneading and twisting the fixed tea leaves by the mechanical force of the rolling machine to obtain rolled tea leaves; S140, sending the rolled tea leaves into a fermentation machine, and controlling the temperature of the fermentation machine in real time. degree to cause the polyphenols in the rolled tea leaves to undergo an oxidation reaction to form red pigments and aroma substances to obtain fermented tea leaves, wherein the fermented tea leaves are red; S150, sending the fermented tea leaves into a dryer, drying the fermented tea leaves by hot air or microwaves to reduce the moisture content of the fermented tea leaves to fix the color, aroma, taste and shape of the tea leaves to obtain dried tea leaves; S160, sending the dried tea leaves into a screening machine, grading the dried tea leaves through screens of different sizes to obtain screened tea leaves; and, S170, sending the screened tea leaves into a packaging machine, sealing and packaging the screened tea leaves by vacuum or nitrogen filling to obtain a finished black tea product.
[0021] Accordingly, considering that fermentation is the most critical step in the production process of black tea, it determines the quality and flavor of black tea. The fermentation temperature is an important factor affecting the fermentation effect. If the temperature is too high or too low, the color, aroma, taste and shape of the tea will change, reducing the quality of black tea. Therefore, it is very important to monitor and control the fermentation temperature in real time during the fermentation process.
[0022] Based on this, the technical concept of this application is to collect tea fermentation status monitoring video and fermentation temperature value in real time, and introduce data processing and analysis algorithms in the back end to perform time series analysis of tea fermentation status monitoring video and fermentation temperature value, so as to realize real-time intelligent adjustment of fermentation temperature. In this way, the adaptive change of fermentation temperature value can be controlled in real time based on the tea fermentation status, so as to optimize the fermentation process and improve the quality and efficiency of black tea production.
[0023] Figure 2 A flowchart of sub-step S140 of the intelligent control method for automated production of black tea according to an embodiment of the present application is shown. Figure 3 FIG. 1 is a schematic diagram showing a sub-step S140 of the intelligent control method for automated production of black tea according to an embodiment of the present application. Figure 2 and Figure 3As shown, according to the intelligent control method for automated production of black tea in the embodiment of the present application, the rolled tea leaves are fed into a fermentation machine, and the temperature of the fermentation machine is controlled in real time to cause the polyphenols in the rolled tea leaves to undergo an oxidation reaction to form red pigments and aroma substances to obtain fermented tea leaves, including: S141, obtaining a tea fermentation status monitoring video of a predetermined time period captured by a camera and obtaining fermentation temperature values at a plurality of predetermined time points within the predetermined time period; S142, extracting image frames corresponding to the plurality of predetermined time points from the tea fermentation status monitoring video to obtain a sequence of tea fermentation status monitoring key frames; S143, performing feature extraction on the sequence of tea fermentation status monitoring key frames respectively through a tea fermentation status feature extractor based on a deep neural network model to obtain tea leaves. S144, performing time series correlation analysis on the sequence of tea fermentation state feature vectors to obtain a tea fermentation state time series change feature vector; S145, arranging the fermentation temperature values of the multiple predetermined time points into a fermentation temperature time series input vector according to the time dimension, and then passing the fermentation temperature time series input vector through a fermentation temperature time series feature extractor based on a one-dimensional convolutional layer to obtain a fermentation temperature time series feature vector; S146, calculating the responsiveness estimate of the tea fermentation state time series change feature vector relative to the fermentation temperature time series feature vector to obtain a fermentation state-fermentation temperature responsiveness transfer feature matrix; and, S147, based on the fermentation state-fermentation temperature responsiveness transfer feature matrix, determining whether the fermentation temperature at the current time point should be increased, decreased or maintained.
[0024] Specifically, in the technical solution of the present application, first, a monitoring video of the fermentation status of tea leaves in a predetermined time period captured by a camera is obtained, and the fermentation temperature values at multiple predetermined time points within the predetermined time period are obtained. Next, considering that the process of tea fermentation can be recorded in real time through monitoring video, the video data usually contains a large amount of redundant information and continuous frame images. In order to reduce the amount of data and extract key information, it is necessary to select some representative image frames from the video. Therefore, in order to be able to monitor the fermentation process and the fermentation status of tea leaves in real time, and to perform corresponding analysis on the temporal changes of the fermentation temperature, in the technical solution of the present application, the image frames corresponding to the multiple predetermined time points are further extracted from the monitoring video of the fermentation status of tea leaves to obtain a sequence of key frames for monitoring the fermentation status of tea leaves.
[0025] Then, the sequence of the tea fermentation status monitoring key frames is subjected to feature mining in a tea fermentation status feature extractor based on a convolutional neural network model to extract the feature distribution information about the tea fermentation status in each of the tea fermentation status monitoring key frames, thereby obtaining a sequence of tea fermentation status feature vectors.
[0026] Correspondingly, in step S143, the deep neural network model is a convolutional neural network model. It is worth mentioning that the convolutional neural network (CNN) is a deep learning model mainly used to process data with a grid structure. The core idea of the convolutional neural network is to extract features from the input data through convolution operations. The convolution operation uses a set of learnable filters (also called convolution kernels or feature detectors) to perform sliding window calculations on the input data and generate feature maps (also called convolution features). These feature maps capture the local spatial relationship of the input data, so that the edges, textures and other local features in the image can be captured. The convolutional neural network also includes a pooling layer (Pooling Layer) for reducing the spatial size of the feature map and retaining the main feature information. Commonly used pooling operations include maximum pooling (Max Pooling) and average pooling (Average Pooling). After the convolution and pooling layers, a fully connected layer (Fully Connected Layer) is usually added to map the extracted features to specific output categories. The output of the fully connected layer can be calculated by the softmax function for classification probability. Convolutional neural networks can also build deeper network structures by stacking multiple convolutional layers, pooling layers, and fully connected layers, such as VGG, ResNet, and Inception. Through end-to-end training, convolutional neural networks can automatically learn the feature representation of input data without manually designing feature extractors. This makes convolutional neural networks have strong expressiveness and generalization capabilities in image processing tasks.
[0027] It should be understood that during the fermentation of tea leaves, the fermentation state of tea leaves will change over time, and the change process of the fermentation state of tea leaves may be relatively subtle and difficult to detect and capture features. Therefore, in the technical solution of the present application, in order to further capture the subtle temporal changes and trends of the fermentation state of tea leaves, it is necessary to further calculate the correlation between each adjacent two tea fermentation state feature vectors in the sequence of the tea fermentation state feature vector to obtain a tea fermentation state temporal change feature vector composed of multiple correlations. By calculating the correlation between the feature vectors, the subtle temporal changes of the tea fermentation state can be captured. In other words, the correlation reflects the similarity and correlation degree between each feature vector, which can be used to measure the continuity and change trend of the tea fermentation state. In addition, the tea fermentation state temporal change feature vector can be used to analyze the dynamic change characteristics in the tea fermentation process, such as the fermentation speed and fermentation stability of tea leaves. These features can provide a reference for subsequent fermentation temperature control, help the control system to more accurately judge the current fermentation state of tea leaves, and make corresponding adjustments.
[0028] Correspondingly, in step S144, a time series correlation analysis is performed on the sequence of the tea fermentation state feature vectors to obtain a time series change feature vector of the tea fermentation state, including: calculating the correlation between every two adjacent tea fermentation state feature vectors in the sequence of the tea fermentation state feature vectors to obtain a time series change feature vector of the tea fermentation state composed of multiple correlations.
[0029] Specifically, calculating the correlation between each two adjacent tea fermentation state feature vectors in the sequence of the tea fermentation state feature vectors to obtain the tea fermentation state time series change feature vector composed of multiple correlations, including: calculating the correlation between each two adjacent tea fermentation state feature vectors in the sequence of the tea fermentation state feature vectors using the following correlation formula to obtain the multiple correlations; wherein the correlation formula is: ;in, is the previous tea fermentation state feature vector of every two adjacent tea fermentation state feature vectors in the sequence of tea fermentation state feature vectors, is the next tea fermentation state feature vector between every two adjacent tea fermentation state feature vectors in the sequence of tea fermentation state feature vectors, and are two different linear transformations, represents the transpose operation, is the correlation between the previous tea fermentation state feature vector and the next tea fermentation state feature vector; and the plurality of correlations are arranged to obtain the tea fermentation state time series change feature vector.
[0030] Next, for the fermentation temperature, it has a dynamic change rule of time series in the time dimension, that is, the fermentation temperature values at the multiple predetermined time points have dynamic correlation characteristic information of time series. Therefore, in order to be able to analyze and characterize the temporal changes of the fermentation temperature values, in the technical solution of the present application, after the fermentation temperature values at the multiple predetermined time points are arranged according to the time dimension as a fermentation temperature time series input vector, the fermentation temperature time series input vector is subjected to feature mining in a fermentation temperature time series feature extractor based on a one-dimensional convolutional layer to extract the temporal correlation characteristic information of the fermentation temperature values in the time dimension, thereby obtaining a fermentation temperature time series feature vector.
[0031] In the production process of black tea, fermentation temperature is a key parameter, which has an important impact on the fermentation process and quality of tea. By monitoring and controlling the fermentation temperature, the fermentation speed and degree of tea can be controlled, thereby affecting the aroma, taste and quality of black tea. In the actual process of controlling the fermentation temperature of tea, it is necessary to comprehensively consider the changes and trends of the fermentation state of tea. Accordingly, in the technical solution of the present application, the tea fermentation state time series change feature vector reflects the dynamic changes of the tea fermentation state, and the fermentation temperature time series feature vector reflects the time series changes of temperature during the fermentation process. Therefore, in order to be able to evaluate the sensitivity of the tea fermentation state to the fermentation temperature, so as to more accurately control the fermentation temperature in real time based on the changes in the fermentation state of tea, in the technical solution of the present application, the responsiveness of the tea fermentation state time series change feature vector relative to the fermentation temperature time series feature vector is further calculated to obtain the fermentation state-fermentation temperature responsiveness transfer feature matrix. By establishing the responsiveness transfer feature matrix between the fermentation state and the fermentation temperature, the changes in the tea fermentation state can be time-series associated with the changes in the fermentation temperature, which is beneficial to the decision-making and regulation of the control system.
[0032] Then, the fermentation state-fermentation temperature responsiveness transfer feature matrix is passed through a classifier to obtain a classification result, and the classification result is used to indicate whether the fermentation temperature at the current time point should be increased, decreased, or maintained. In other words, the responsiveness transfer associated feature information between the temporal change characteristics of the fermentation state of the tea leaves and the temporal change characteristics of the temperature value is used for classification processing, so as to achieve real-time intelligent adjustment of the fermentation temperature. In this way, the adaptive change of the fermentation temperature value can be controlled in real time based on the fermentation state of the tea leaves, so as to optimize the fermentation process and improve the quality and efficiency of black tea production.
[0033] Accordingly, in step S147, based on the fermentation state-fermentation temperature responsiveness transfer feature matrix, it is determined whether the fermentation temperature at the current time point should be increased, decreased or maintained, including: passing the fermentation state-fermentation temperature responsiveness transfer feature matrix through a classifier to obtain a classification result, and the classification result is used to indicate that the fermentation temperature at the current time point should be increased, decreased or maintained.
[0034] Specifically, the fermentation state-fermentation temperature responsiveness transfer feature matrix is passed through a classifier to obtain a classification result, and the classification result is used to indicate whether the fermentation temperature at the current time point should be increased, decreased or maintained, including: expanding the fermentation state-fermentation temperature responsiveness transfer feature matrix into a classification feature vector according to a row vector or a column vector; using the fully connected layer of the classifier to fully connect encode the classification feature vector to obtain an encoded classification feature vector; and inputting the encoded classification feature vector into the Softmax classification function of the classifier to obtain the classification result.
[0035] It should be understood that the role of the classifier is to use the given categories and known training data to learn classification rules and classifiers, and then classify (or predict) unknown data. Logistic regression, SVM, etc. are often used to solve binary classification problems. For multi-class classification problems, logistic regression or SVM can also be used, but multiple binary classifications are required to form a multi-classification, but this is prone to errors and inefficient. Commonly used multi-classification methods include the Softmax classification function.
[0036] In the above technical scheme, each tea fermentation state feature vector in the sequence of tea fermentation state feature vectors expresses the image semantic features of the corresponding tea fermentation state monitoring key frame. Therefore, the tea fermentation state temporal change feature vector expresses the image semantic temporal correlation domain transfer features in adjacent time domains, and the fermentation temperature temporal feature vector expresses the local temporal correlation features of the fermentation temperature value. In this way, when calculating the responsiveness estimate of the tea fermentation state temporal change feature vector relative to the fermentation temperature temporal feature vector, considering the distribution difference between the tea fermentation state temporal change feature vector and the fermentation temperature temporal feature vector in the temporal dimension, and the randomness introduced in the Gaussian discretization process, the overall feature distribution of the fermentation state-fermentation temperature responsiveness transfer feature matrix will have more significant inconsistency and instability, and thus it is necessary to improve the logical dependence of the temporal feature distribution of the fermentation state-fermentation temperature responsiveness transfer feature matrix.
[0037] In a preferred example, the fermentation state-fermentation temperature responsiveness transfer feature matrix is passed through a classifier to obtain a classification result, including: firstly, the fermentation state-fermentation temperature responsiveness transfer feature matrix is expanded into a fermentation state-fermentation temperature responsiveness transfer feature vector, and then the fermentation state-fermentation temperature responsiveness transfer feature vector is optimized, specifically including: determining a first weight matrix and a second weight matrix, wherein the first weight matrix and the second weight matrix are The eigenvalues are the first The eigenvalue and The square root of the correlation value and the L2 distance value of the eigenvalue: ;in, and respectively represent the first The eigenvalue and The square root of the correlation value and the L2 distance value of the eigenvalue, represents the first weight matrix Eigenvalues, represents the second weight matrix Eigenvalue.
[0038] The fermentation state-fermentation temperature responsiveness transfer feature vector is matrix-multiplied with the first weight matrix to obtain a first intermediate vector, the fermentation state-fermentation temperature responsiveness transfer feature vector is matrix-multiplied with the second weight matrix to obtain a second intermediate vector, and the first weight matrix is dot-multiplied with the second weight matrix and then matrix-multiplied with the fermentation state-fermentation temperature responsiveness transfer feature vector to obtain a third intermediate vector; the first intermediate vector, the second intermediate vector and the third intermediate vector are dot-multiplied to obtain an optimized fermentation state-fermentation temperature responsiveness transfer feature vector; the optimized fermentation state-fermentation temperature responsiveness transfer feature vector is passed through a classifier to obtain a classification result.
[0039] For example, the fermentation state-fermentation temperature responsiveness transfer characteristic vector is a row vector, then the optimized fermentation state-fermentation temperature responsiveness transfer feature vector It is expressed as: ;in, represents the fermentation state-fermentation temperature responsiveness transfer feature vector, represents the first weight matrix, represents the second weight matrix, Indicates point addition, represents matrix multiplication, represents dot product, , and represent the first intermediate vector, the second intermediate vector and the third intermediate vector respectively, Represents the optimized fermentation state-fermentation temperature responsiveness transfer feature vector.
[0040] That is, by taking the fermentation state-fermentation temperature responsiveness transfer feature vector as its statistical reference-free distribution response framework based on the auto-association matrix and the self-distance matrix, and constructing it based on the retrieval-enhanced reverse response of the fermentation state-fermentation temperature responsiveness transfer feature vector to avoid simple repetition of feature distribution, and by ensuring the intrinsic mapping logic of the fermentation state-fermentation temperature responsiveness transfer feature vector based on the retrieval-response context to avoid its surface combinatorial mapping, so as to achieve the semantic logical dependency mapping of the fermentation state-fermentation temperature responsiveness transfer feature vector while maintaining the intuitive response correlation, thereby improving the accuracy of the classification results.
[0041] Furthermore, in the technical solution of the present application, the intelligent control method for the automated production of black tea also includes a training step: for training the tea fermentation state feature extractor based on the convolutional neural network model, the fermentation temperature time series feature extractor based on the one-dimensional convolutional layer, and the classifier.
[0042] In one example, the training step includes: obtaining training data, the training data including a training tea fermentation state monitoring video of a predetermined time period collected by a camera, training fermentation temperature values at multiple predetermined time points within the predetermined time period, and a real value of the fermentation temperature at the current time point that should be increased, decreased, or maintained; extracting training image frames corresponding to the multiple predetermined time points from the training tea fermentation state monitoring video to obtain a sequence of training tea fermentation state monitoring key frames; performing feature extraction on the sequence of training tea fermentation state monitoring key frames by the tea fermentation state feature extractor based on the convolutional neural network model to obtain a sequence of training tea fermentation state feature vectors; performing time series association analysis on the sequence of training tea fermentation state feature vectors to obtain a training tea fermentation state time series change feature vector; arranging the training fermentation temperature values at the multiple predetermined time points according to the time dimension as a training fermentation temperature time series input vector , passing the training fermentation temperature time series input vector through the fermentation temperature time series feature extractor based on the one-dimensional convolution layer to obtain the training fermentation temperature time series feature vector; calculating the responsiveness estimate of the training tea fermentation state time series change feature vector relative to the training fermentation temperature time series feature vector to obtain the training fermentation state-fermentation temperature responsiveness transfer feature matrix; passing the training fermentation state-fermentation temperature responsiveness transfer feature matrix through the classifier to obtain the classification loss function value; and, based on the classification loss function value and through the direction propagation of gradient descent, training the tea fermentation state feature extractor based on the convolutional neural network model, the fermentation temperature time series feature extractor based on the one-dimensional convolution layer and the classifier, wherein, at each weight matrix iteration of the training, the training fermentation state-fermentation temperature responsiveness transfer feature vector obtained after the training fermentation state-fermentation temperature responsiveness transfer feature matrix is expanded is trained and optimized.
[0043] In particular, in the above technical scheme, each training tea fermentation state feature vector in the sequence of training tea fermentation state feature vectors expresses the image semantic features of the corresponding training tea fermentation state monitoring key frame, thereby, the training tea fermentation state time series change feature vector expresses the image semantic time series related domain transfer features in adjacent time domains, and the training fermentation temperature time series feature vector expresses the local time series correlation features of the training fermentation temperature value. In this way, when calculating the responsiveness estimate of the training tea fermentation state time series change feature vector relative to the training fermentation temperature time series feature vector, considering the distribution difference between the training tea fermentation state time series change feature vector and the training fermentation temperature time series feature vector in the time series dimension, and the randomness introduced in the Gaussian discretization process, the overall feature distribution of the training fermentation state-fermentation temperature responsiveness transfer feature matrix will have more significant inconsistency and instability, thereby affecting its stability in classification training through the classifier.
[0044] Based on this, the applicant of the present application performs training optimization on the training fermentation state-fermentation temperature responsiveness transfer feature vector obtained after expanding the training fermentation state-fermentation temperature responsiveness transfer feature matrix through a classifier for classification training.
[0045] Accordingly, in one example, at each weight matrix iteration of the training, the training fermentation state-fermentation temperature responsiveness transfer feature vector obtained after the training fermentation state-fermentation temperature responsiveness transfer feature matrix is expanded is trained and optimized, including: training and optimizing the training fermentation state-fermentation temperature responsiveness transfer feature vector obtained after the training fermentation state-fermentation temperature responsiveness transfer feature matrix is expanded using the following optimization formula to obtain an optimized training fermentation state-fermentation temperature responsiveness transfer feature matrix; wherein the optimization formula is: ;in, is the training fermentation state-fermentation temperature responsiveness transfer feature vector obtained after the training fermentation state-fermentation temperature responsiveness transfer feature matrix is expanded, is the training fermentation state-fermentation temperature responsiveness transfer feature vector The characteristic value of and are respectively the training fermentation state-fermentation temperature responsiveness transfer feature vectors The 1-norm and 2-norm of is the training fermentation state-fermentation temperature responsiveness transfer feature vector The length of is with The relevant weight hyperparameters, represents an exponential operation of a value, wherein the exponential operation of the value represents calculating the value of a natural exponential function with the value as a power, It is the eigenvalue of the optimized training fermentation state-fermentation temperature responsiveness transfer feature vector obtained after the optimized training fermentation state-fermentation temperature responsiveness transfer feature matrix is expanded.
[0046] Here, the training fermentation state-fermentation temperature responsiveness transfer feature vector The overall feature distribution of is represented by the structural consistency and stability under the rigid structure of absolute distance and the non-rigid structure of spatial distance, so that the training fermentation state-fermentation temperature responsiveness transfer feature vector The global feature distribution of the local pattern change has a certain repeatability, so as to transfer the feature vector in the training fermentation state-fermentation temperature responsiveness When the classifier is used for classification, the global feature distribution is robust to the scale and rotation changes of the weight matrix of the classifier, which improves the stability of classification training. In this way, the adaptive change of the fermentation temperature value can be controlled in real time based on the fermentation state of the tea leaves, so as to optimize the fermentation process and improve the quality and efficiency of black tea production.
[0047] In summary, the intelligent control method for automated production of black tea based on the embodiment of the present application can adaptively control the fermentation temperature value in real time based on the fermentation state of the tea leaves, thereby optimizing the fermentation process and improving the quality and efficiency of black tea production.
[0048] Figure 4 FIG. 1 is a block diagram of an intelligent control system 100 for automated production of black tea according to an embodiment of the present application. Figure 4As shown, according to the embodiment of the present application, the intelligent control system 100 for the automated production of black tea includes: a picking and sorting module 110, for picking fresh tea leaves in a tea garden using a robot or an unmanned aerial vehicle, and sorting and distributing the tea leaves according to their variety, grade and maturity to obtain sorted tea leaves; a heat treatment module 120, for sending the sorted tea leaves into a fixing machine, and heat-treating the sorted tea leaves by high-temperature steam or hot air to obtain fixed tea leaves; a rolling module 130, for sending the fixed tea leaves into a rolling machine, and squeezing, kneading and twisting the fixed tea leaves by the mechanical force of the rolling machine to obtain rolled tea leaves; a fermentation module 140, for sending the rolled tea leaves into a fermentation machine, and controlling the fermentation machine in real time to obtain the rolled tea leaves. The temperature of the fermentation machine is used to cause the polyphenols in the rolled tea leaves to undergo an oxidation reaction to form red pigments and aroma substances to obtain fermented tea leaves, wherein the fermented tea leaves are red; a drying module 150 is used to send the fermented tea leaves into a dryer, dry the fermented tea leaves by hot air or microwaves to reduce the moisture of the fermented tea leaves to fix the color, aroma, taste and shape of the tea leaves to obtain dried tea leaves; a grading module 160 is used to send the dried tea leaves into a screening machine, grade the dried tea leaves through sieves of different sizes to obtain screened tea leaves; and a sealing and packaging module 170 is used to send the screened tea leaves into a packaging machine, seal and package the screened tea leaves by vacuum or nitrogen filling to obtain a finished black tea product.
[0049] In a possible implementation, the fermentation module 140 includes: a data acquisition unit, which is used to acquire a tea fermentation status monitoring video of a predetermined time period collected by a camera and to acquire fermentation temperature values at a plurality of predetermined time points within the predetermined time period; an image frame extraction unit, which is used to extract image frames corresponding to the plurality of predetermined time points from the tea fermentation status monitoring video to obtain a sequence of tea fermentation status monitoring key frames; a tea fermentation status feature extraction unit, which is used to perform feature extraction on the sequence of tea fermentation status monitoring key frames respectively through a tea fermentation status feature extractor based on a deep neural network model to obtain a sequence of tea fermentation status feature vectors; a time series correlation analysis unit, which is used to perform time series correlation analysis on the sequence of tea fermentation status feature vectors; A sequential correlation analysis is performed to obtain a time-series variation feature vector of the fermentation state of tea leaves; a fermentation temperature time-series feature extraction unit is used to arrange the fermentation temperature values of the multiple predetermined time points into a fermentation temperature time-series input vector according to the time dimension, and then pass the fermentation temperature time-series input vector through a fermentation temperature time-series feature extractor based on a one-dimensional convolutional layer to obtain a fermentation temperature time-series feature vector; a responsiveness estimation unit is used to calculate the responsiveness estimation of the time-series variation feature vector of the fermentation state of tea leaves relative to the fermentation temperature time-series feature vector to obtain a fermentation state-fermentation temperature responsiveness transfer feature matrix; and a fermentation temperature control unit is used to determine whether the fermentation temperature at the current time point should be increased, decreased or maintained based on the fermentation state-fermentation temperature responsiveness transfer feature matrix.
[0050] Here, those skilled in the art will appreciate that the specific functions and operations of the various units and modules in the intelligent control system 100 for automated production of black tea have been described in detail above. Figures 1 to 3 The invention has been described in detail in the description of the intelligent control method for automated production of black tea, and therefore, its repeated description will be omitted.
[0051] As described above, the intelligent control system 100 for automated production of black tea according to the embodiment of the present application can be implemented in various wireless terminals, such as a server with an intelligent control algorithm for automated production of black tea. In a possible implementation, the intelligent control system 100 for automated production of black tea according to the embodiment of the present application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the intelligent control system 100 for automated production of black tea can be a software module in the operating system of the wireless terminal, or can be an application developed for the wireless terminal; of course, the intelligent control system 100 for automated production of black tea can also be one of the many hardware modules of the wireless terminal.
[0052] Alternatively, in another example, the intelligent control system 100 for automated production of black tea and the wireless terminal may be separate devices, and the intelligent control system 100 for automated production of black tea may be connected to the wireless terminal via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.
[0053] Figure 5 The following is a diagram showing an application scenario of an intelligent control method for automated production of black tea according to an embodiment of the present application. Figure 5 As shown, in this application scenario, first, a monitoring video of tea fermentation status collected by a camera for a predetermined time period is obtained (for example, Figure 5 D1) and the fermentation temperature values at multiple predetermined time points within the predetermined time period (for example, Figure 5 Then, the tea fermentation status monitoring video and the fermentation temperature values at the plurality of predetermined time points are input to a server (for example, Figure 5 S) as shown in , wherein the server is able to use the intelligent control algorithm for automated production of black tea to process the tea fermentation status monitoring video and the fermentation temperature values at the multiple predetermined time points to obtain a classification result indicating whether the fermentation temperature at the current time point should be increased, decreased or maintained.
[0054] The flow chart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to multiple embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a part of a module, program segment or instruction, and a part of the module, program segment or instruction includes one or more executable instructions for realizing the logical function of the specification. In some alternative implementations, the function marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two continuous square boxes can actually be executed substantially in parallel, and they can also be executed in the opposite order sometimes, depending on the function involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be realized by a dedicated hardware-based system that performs the function or action of the specification, or can be realized by a combination of special-purpose hardware and computer instructions.
[0055] The embodiments of the present application have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.
Claims
1. An intelligent control method for automated production of black tea, characterized in that: include: Use robots or drones to pick fresh tea leaves in a tea garden, and classify and distribute them according to the variety, grade and maturity of the tea leaves to obtain classified tea leaves; send the classified tea leaves to a fixing machine, and heat-treat the classified tea leaves with high-temperature steam or hot air to obtain fixed tea leaves; send the fixed tea leaves to a rolling machine, and squeeze, rub and twist the fixed tea leaves with the mechanical force of the rolling machine to obtain rolled tea leaves; send the rolled tea leaves to a fermentation machine, and control the temperature of the fermentation machine in real time to allow polyphenols in the rolled tea leaves to undergo decomposition. The oxidation reaction forms red pigment and aroma substances to obtain fermented tea leaves, wherein the fermented tea leaves are red; the fermented tea leaves are sent to a dryer, and the fermented tea leaves are dried by hot air or microwaves to reduce the moisture content of the fermented tea leaves to fix the color, aroma, taste and shape of the tea leaves to obtain dried tea leaves; the dried tea leaves are sent to a screening machine, and the dried tea leaves are graded by screens of different sizes to obtain screened tea leaves; and the screened tea leaves are sent to a packaging machine, and the screened tea leaves are sealed and packaged by vacuum or nitrogen filling to obtain a finished black tea product.
2. The intelligent control method for automated production of black tea according to claim 1, characterized in that: The rolled tea leaves are fed into a fermentation machine, and the temperature of the fermentation machine is controlled in real time to cause the polyphenols in the rolled tea leaves to undergo an oxidation reaction to form red pigments and aroma substances to obtain fermented tea leaves, including: obtaining a tea fermentation status monitoring video of a predetermined time period collected by a camera and obtaining fermentation temperature values at multiple predetermined time points within the predetermined time period; extracting image frames corresponding to the multiple predetermined time points from the tea fermentation status monitoring video to obtain a sequence of tea fermentation status monitoring key frames; and extracting features from the sequence of tea fermentation status monitoring key frames using a tea fermentation status feature extractor based on a deep neural network model to obtain a sequence of tea fermentation status feature vectors. ; Perform time series correlation analysis on the sequence of the tea fermentation state feature vectors to obtain the tea fermentation state time series change feature vector; After arranging the fermentation temperature values of the multiple predetermined time points according to the time dimension as a fermentation temperature time series input vector, the fermentation temperature time series input vector is passed through a fermentation temperature time series feature extractor based on a one-dimensional convolutional layer to obtain a fermentation temperature time series feature vector; Calculate the responsiveness estimate of the tea fermentation state time series change feature vector relative to the fermentation temperature time series feature vector to obtain a fermentation state-fermentation temperature responsiveness transfer feature matrix; and based on the fermentation state-fermentation temperature responsiveness transfer feature matrix, determine whether the fermentation temperature at the current time point should be increased, decreased or maintained.
3. The intelligent control method for automated production of black tea according to claim 2, characterized in that: The deep neural network model is a convolutional neural network model.
4. The intelligent control method for automated production of black tea according to claim 3, characterized in that: Performing a time series correlation analysis on the sequence of the tea fermentation state feature vectors to obtain the tea fermentation state time series change feature vector, including: calculating the correlation between every two adjacent tea fermentation state feature vectors in the sequence of the tea fermentation state feature vectors to obtain the tea fermentation state time series change feature vector composed of multiple correlations.
5. The intelligent control method for automated production of black tea according to claim 4, characterized in that: Calculating the correlation between each two adjacent tea fermentation state feature vectors in the sequence of the tea fermentation state feature vectors to obtain the tea fermentation state time series change feature vector composed of multiple correlations, comprising: calculating the correlation between each two adjacent tea fermentation state feature vectors in the sequence of the tea fermentation state feature vectors using the following correlation formula to obtain the multiple correlations; wherein the correlation formula is: ;in, is the previous tea fermentation state feature vector of every two adjacent tea fermentation state feature vectors in the sequence of tea fermentation state feature vectors, is the next tea fermentation state feature vector between every two adjacent tea fermentation state feature vectors in the sequence of tea fermentation state feature vectors, and are two different linear transformations, represents the transpose operation, is the correlation between the previous tea fermentation state feature vector and the next tea fermentation state feature vector; and the multiple correlations are arranged to obtain the tea fermentation state time series change feature vector.
6. The intelligent control method for automated production of black tea according to claim 5, characterized in that: Based on the fermentation state-fermentation temperature responsiveness transfer feature matrix, determine whether the fermentation temperature at the current time point should be increased, decreased or maintained, including: passing the fermentation state-fermentation temperature responsiveness transfer feature matrix through a classifier to obtain a classification result, and the classification result is used to indicate whether the fermentation temperature at the current time point should be increased, decreased or maintained.
7. The intelligent control method for automated production of black tea according to claim 6, characterized in that: It also includes a training step: for training the tea fermentation state feature extractor based on the convolutional neural network model, the fermentation temperature time series feature extractor based on the one-dimensional convolutional layer, and the classifier.
8. The intelligent control method for automated production of black tea according to claim 7, characterized in that: The training step comprises: obtaining training data, wherein the training data comprises a training tea fermentation state monitoring video of a predetermined time period collected by a camera, training fermentation temperature values at a plurality of predetermined time points within the predetermined time period, and a real value of the fermentation temperature at the current time point that should be increased, decreased or maintained; extracting training image frames corresponding to the plurality of predetermined time points from the training tea fermentation state monitoring video to obtain a sequence of training tea fermentation state monitoring key frames; performing feature extraction on the sequence of training tea fermentation state monitoring key frames by the tea fermentation state feature extractor based on the convolutional neural network model to obtain a sequence of training tea fermentation state feature vectors; performing time series association analysis on the sequence of training tea fermentation state feature vectors to obtain a training tea fermentation state time series change feature vector; arranging the training fermentation temperature values at the plurality of predetermined time points into a training fermentation temperature time series input vector according to the time dimension, and then The training fermentation temperature time series input vector is passed through the fermentation temperature time series feature extractor based on the one-dimensional convolution layer to obtain the training fermentation temperature time series feature vector; the responsiveness estimate of the training tea fermentation state time series change feature vector relative to the training fermentation temperature time series feature vector is calculated to obtain the training fermentation state-fermentation temperature responsiveness transfer feature matrix; the training fermentation state-fermentation temperature responsiveness transfer feature matrix is passed through the classifier to obtain the classification loss function value; and based on the classification loss function value and through the direction propagation of gradient descent, the tea fermentation state feature extractor based on the convolutional neural network model, the fermentation temperature time series feature extractor based on the one-dimensional convolution layer and the classifier are trained, wherein, at each weight matrix iteration of the training, the training fermentation state-fermentation temperature responsiveness transfer feature vector obtained after the training fermentation state-fermentation temperature responsiveness transfer feature matrix is expanded is trained and optimized.
9. An intelligent control system for automated production of black tea, characterized in that: include: A picking and sorting module, used for picking fresh tea leaves in a tea garden using a robot or a drone, and sorting and distributing the tea leaves according to their variety, grade and maturity to obtain sorted tea leaves; a heat treatment module for feeding the classified tea leaves into a fixing machine, and heat-treating the classified tea leaves by high-temperature steam or hot air to obtain fixing tea leaves; a rolling module for feeding the fixing tea leaves into a rolling machine, and squeezing, kneading and twisting the fixing tea leaves by the mechanical force of the rolling machine to obtain rolling tea leaves; a fermentation module for feeding the rolling tea leaves into a fermentation machine, and causing polyphenols in the rolling tea leaves to undergo oxidation reaction to form red pigments and aroma substances by real-time control of the temperature of the fermentation machine to obtain fermented tea leaves, wherein the fermented tea leaves are red; a drying module for feeding the fermented tea leaves into a drying machine, and drying the fermented tea leaves by hot air or microwaves to reduce the moisture content of the fermented tea leaves to fix the color, aroma, taste and shape of the tea leaves to obtain dried tea leaves; a grading module for feeding the dried tea leaves into a screening machine, and grading the dried tea leaves by sieves of different sizes to obtain sieved tea leaves; And a sealing packaging module is used for feeding the screened tea leaves into a packaging machine, and sealing and packaging the screened tea leaves by vacuum or nitrogen filling to obtain a finished black tea product.
10. The intelligent control system for automated production of black tea according to claim 9, characterized in that: The fermentation module includes: a data acquisition unit, which is used to acquire a tea fermentation status monitoring video of a predetermined time period collected by a camera and to acquire fermentation temperature values at a plurality of predetermined time points within the predetermined time period; an image frame extraction unit, which is used to extract image frames corresponding to the plurality of predetermined time points from the tea fermentation status monitoring video to obtain a sequence of tea fermentation status monitoring key frames; a tea fermentation status feature extraction unit, which is used to perform feature extraction on the sequence of tea fermentation status monitoring key frames respectively through a tea fermentation status feature extractor based on a deep neural network model to obtain a sequence of tea fermentation status feature vectors; a time series correlation analysis unit, which is used to perform time series correlation analysis on the sequence of tea fermentation status feature vectors to obtain to the time series change feature vector of the tea fermentation state; a fermentation temperature time series feature extraction unit, which is used to arrange the fermentation temperature values of the multiple predetermined time points into a fermentation temperature time series input vector according to the time dimension, and then pass the fermentation temperature time series input vector through a fermentation temperature time series feature extractor based on a one-dimensional convolutional layer to obtain a fermentation temperature time series feature vector; a responsiveness estimation unit, which is used to calculate the responsiveness estimate of the time series change feature vector of the tea fermentation state relative to the fermentation temperature time series feature vector to obtain a fermentation state-fermentation temperature responsiveness transfer feature matrix; and a fermentation temperature control unit, which is used to determine whether the fermentation temperature at the current time point should be increased, decreased or maintained based on the fermentation state-fermentation temperature responsiveness transfer feature matrix.
Citation Information
Patent Citations
Automatic black tea fermenting and producing device
CN102742685A
Method for making multiple types of black tea through multifunctional production line
CN105211349A
Rose and red date tea and preparing method thereof
CN106306178A
Red tea processing technology
CN106359686A
Preparation method of zymolytic black tea
CN108935786A
Cited By
Method for production control of cassia tea, system and production process thereof
CN120181637B
Method and system for withering schedule prediction of tea leaves
EP4625365A1
