Intelligent tea drying equipment and tea drying method

Abstract

The present application belongs to the field of drying technology, and in particular, relates to an intelligent tea drying device and a method for drying tea. The existing tea drying devices have relatively simple functions. The present application provides an intelligent tea drying device, including a shell, in which an inlet, a drying component, a sorting and processing component and a storage component are arranged in sequence, and an intelligent recognition module is arranged on the drying component, the intelligent recognition module is connected to a controller, the drying component is connected to the controller, and the sorting and processing component is connected to the controller; after the intelligent recognition module identifies the type and moisture degree of the tea leaves, the data is transmitted to the controller, the controller calculates the drying and frying time, and controls the drying component to dry the tea leaves, and the dried tea leaves are sorted and stored, and the intelligent recognition module includes a convolutional neural network for tea posture and color recognition. Not only can the interactive process of drying tea leaves be observed intuitively, but also the tea leaves can be processed secondary.

Description

Technical Field

[0001] The present application belongs to the field of drying technology, and in particular relates to an intelligent tea drying device and a tea drying method. Background Art

[0002] As tea is becoming more and more popular among the people, the original intention of this equipment is to let people get closer to the tea making process, generate interactive psychology, let people have a deeper understanding of tea drinking culture, and slow down the impetuous life. There are three ways to dry tea leaves: drying, frying and sun drying. People use drying equipment to process the already prepared tea leaves, dry the opened tea leaves, and grind the tea leaves that are too damp.

[0003] The tea drying machine relies on flowing hot air to dry the tea leaves, but the existing tea drying equipment has relatively single functions. Summary of the invention

[0004] 1. Technical issues to be solved

[0005] Based on the problem that the existing tea drying equipment has relatively single functions, the present application provides an intelligent tea drying equipment and a tea drying method.

[0006] 2. Technical solution

[0007] In order to achieve the above-mentioned purpose, the present application provides an intelligent tea drying device, including a shell, in which an inlet, a drying component, a sorting and processing component and a storage component connected in sequence are arranged, and an intelligent recognition module is arranged on the drying component, and the intelligent recognition module is connected to a controller, and the drying component is connected to the controller, and the sorting and processing component is connected to the controller; after the intelligent recognition module identifies the type of tea and the degree of moisture, the data is transmitted to the controller, and the controller calculates the drying and frying time, and controls the drying component to dry the tea, and stores the dried tea after sorting. The intelligent recognition module includes a convolutional neural network for tea posture and color recognition.

[0008] Another embodiment provided by the present application is: the drying component includes a ceramic heating ball, a ceramic heating roller is arranged inside the ceramic heating ball, and the intelligent recognition module is arranged on the ceramic heating ball.

[0009] Another embodiment provided by the present application is: the sorting and processing component includes a tea sorter, one end of the tea sorter is connected to the ceramic heating ball, the other end of the tea sorter is connected to the first tea channel, and the other end of the tea sorter is connected to the second tea channel. The first tea channel is arranged above the tea sorting funnel, and the second tea channel is arranged above the tea grinder.

[0010] Another embodiment provided by the present application is: the storage assembly includes a first tea storage box and a second tea storage box, the first tea storage box is arranged below the tea sorting funnel, and the second tea storage box is arranged below the tea grinder.

[0011] Another embodiment provided by the present application is: a first sorting baffle is provided between the first tea storage box and the tea sorting funnel, and a second sorting baffle is provided between the second tea storage box and the tea grinder.

[0012] Another implementation manner provided by the present application is: a pop-up spring piece is provided on the first tea storage box, and a pop-up spring piece is provided on the second tea storage box.

[0013] Another implementation manner provided by the present application is: the number of the intelligent recognition modules is 3, and the 3 intelligent recognition modules are distributed on the ceramic heating ball in a 120-degree array.

[0014] Another implementation manner provided by the present application is that the intelligent recognition module is connected to a display.

[0015] The present application also provides a method for drying tea leaves using the intelligent tea drying device, wherein a convolutional neural network for tea leaf posture and color recognition is used to identify tea leaves, and the degree of moisture of the tea leaves is determined based on the difference in the curled posture and color of the tea leaves from tea leaves of good quality. The heating time and heating temperature are set based on the degree of moisture, so that the tea leaves reach the quality of good quality after drying.

[0016] Another embodiment provided by the present application is: the convolutional neural network training method for tea posture and color recognition includes forward propagation and backward propagation, and the forward propagation includes the following steps:

[0017] 1) Select a tea posture color dataset and divide it into a tea posture color training set, a tea posture color test set, and a tea posture color verification set;

[0018] 2) Randomly initialize the weights, bias terms, and sample error values ​​in the convolutional neural network for tea posture and color recognition, and manually set the initial learning rate;

[0019] 3) Inputting data vectors of samples of the tea posture and color training set into the network, the data vectors are continuously transformed through the processing of each layer from the input layer to the output layer;

[0020] The back propagation includes the following steps:

[0021] a) Using the method of minimizing the error, the error threshold is calculated based on the actual output value and the predicted output value of the sample, and then the error value is back-propagated layer by layer to adjust the parameters of each layer according to the error value;

[0022] b) Compare the adjusted total error value with the error threshold value. If the former is greater than the latter, the network has not achieved the task goal, and then returns to 3) of the forward propagation. Otherwise, proceed to the next step.

[0023] c) The training and learning task is completed, a trained convolutional neural network is obtained and the parameters of the network are saved;

[0024] After the training is completed, the trained model can be used for prediction. At the same time, the trained model can be loaded into the testing phase, and damp tea leaves can be collected as input. According to the known collection labels, the prediction results of the network can be compared to know the prediction effect of the model, and then the actual test accuracy of the model can be obtained.

[0025] 3. Beneficial effects

[0026] Compared with the prior art, the intelligent tea drying device and tea drying method provided by the present application have the following beneficial effects:

[0027] The intelligent tea drying equipment provided in this application involves intelligent identification and ceramic heating, and is particularly capable of identifying the quality and type of tea.

[0028] The intelligent tea drying equipment provided by the present application can be used to dry damp tea leaves. It determines the drying time required for different types of tea through intelligent recognition technology and has the function of grinding tea powder. Ceramic heating technology is applied therein. A resistance wire is inserted into a ceramic with a high degree of insulation and fire resistance, and an alloy wire is cast in a semiconductor made of quartz glass as raw material. The entire drying process can be intuitively felt through the isolation glass, thereby increasing the user's sense of interaction. Waste tea leaves can be turned into powder and put into sachets, etc., turning waste into treasure.

[0029] The intelligent tea drying equipment provided in the present application can not only visually observe the interactive process of drying tea leaves, but also perform secondary processing on the tea leaves. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a schematic diagram of the intelligent tea drying equipment of the present application;

[0031] Figure 2 It is a schematic diagram of the structure of the intelligent tea drying equipment of the present application;

[0032] Figure 3 It is a partial structural schematic diagram of the intelligent tea drying equipment of the present application;

[0033] Figure 4 This is a schematic diagram of the second partial structure of the intelligent tea drying device of the present application;

[0034] Figure 5 It is a schematic diagram of the drying process of the intelligent tea drying equipment of the present application;

[0035] Figure 6 It is a schematic diagram of the display interface of the display of this application. DETAILED DESCRIPTION

[0036] Hereinafter, the specific embodiments of the present application will be described in detail with reference to the accompanying drawings. According to these detailed descriptions, those skilled in the art can clearly understand the present application and can implement the present application. Without violating the principles of the present application, the features in the various embodiments can be combined to obtain new implementations, or certain features in certain embodiments can be replaced to obtain other preferred implementations.

[0037] See also Figures 1 to 6 The present application provides an intelligent tea drying device, including a shell 101, wherein the shell 101 is provided with an entrance 201, a drying component, a sorting and processing component and a storage component which are connected in sequence, wherein the drying component is provided with an intelligent recognition module 210, wherein the intelligent recognition module 210 is connected to a controller, wherein the drying component is connected to the controller, and wherein the sorting and processing component is connected to the controller; after the intelligent recognition module 210 recognizes the type of tea and the degree of moisture, the data is transmitted to the controller, wherein the controller calculates the drying and frying time, and controls the drying component to dry the tea, and then stores the dried tea after sorting.

[0038] The controller here can be a smart terminal, including a mobile phone, computer, tablet computer, etc.; the controller can also be a small program or APP, or a physical button.

[0039] Intelligent recognition technology is used to identify tea varieties and corresponding drying and frying time. In addition, the intelligent recognition module is combined with the tea sorter to select new and old teas and automatically produce tea.

[0040] The storage component stores tea leaves processed by the tea sorter or ground tea powder.

[0041] The intelligent tea drying device of the present application provides a device for identifying the type of tea and automatically setting the drying temperature and time. Power on the device, open the entrance 201 and put the tea leaves in. The intelligent identification module 210 automatically identifies the type of tea leaves and the degree of moisture. After the drying component completes the drying process, the tea leaves are sorted by the sorting component. After the tea leaves that have been completely dampened are dried, they are processed by the sorting component for secondary use.

[0042] An observation window 102 may be provided on the housing 101. During the frying process, the process of the ceramic heating roller frying the tea leaves may be seen through the observation window 102, thereby increasing the sense of interaction of the user.

[0043] Further, the drying component includes a ceramic heating ball 209, a ceramic heating roller 208 is disposed inside the ceramic heating ball 209, and the intelligent identification module 210 is disposed on the ceramic heating ball 209. The internal ceramic heating roller 208 and the ceramic heating ball 209 can be seen through the observation window 102.

[0044] The ceramic heating ball 209 and the ceramic heating roller 208 use ceramic heating (MCH heating element) technology, with a high degree of insulation and fire resistance of ceramics with resistance wire, and the alloy wire is cast in a semiconductor made of quartz glass to form a drying component. The ceramic heating ball 209 remains stationary, and the ceramic heating roller 208 rotates clockwise to dry the tea leaves. The rotation of the ceramic heating roller 208 is achieved through gear transmission.

[0045] Furthermore, the sorting and processing component includes a tea sorter 104, one end of the tea sorter 104 is connected to the ceramic heating ball 209, the other end of the tea sorter 104 is connected to the first tea channel, and the other end of the tea sorter 104 is connected to the second tea channel. The first tea channel is arranged above the tea sorting funnel 204, and the second tea channel is arranged above the tea grinder 205.

[0046] Furthermore, the storage assembly includes a first tea storage box 103 and a second tea storage box, the first tea storage box is arranged below the tea sorting funnel 204, and the second tea storage box is arranged below the tea grinder 205.

[0047] Furthermore, a first sorting baffle 206 is provided between the first tea storage box 103 and the tea sorting hopper, and a second sorting baffle is provided between the second tea storage box and the tea grinder.

[0048] Furthermore, a pop-up spring piece 207 is provided on the first tea storage box 103 , and a pop-up spring piece 207 is provided on the second tea storage box.

[0049] Furthermore, there are three intelligent recognition modules 210 , and the three intelligent recognition modules 210 are distributed on the ceramic heating ball 209 in a 120-degree array.

[0050] Furthermore, the intelligent recognition module 210 is connected to a display. The display here is provided with a display interface, and the display can be provided on the drying device or can be provided separately.

[0051] The user puts tea leaves into the tea drying equipment, which can automatically identify and dry the tea leaves, visually observe the tea drying process, and perform secondary processing on different types of tea leaves.

[0052] The present application also provides a method for drying tea leaves using the intelligent tea drying device. The intelligent tea drying device has a built-in convolutional neural network for tea leaf posture and color recognition. By collecting images of damp tea leaves and normal and intact tea leaves, interference from irrelevant backgrounds is avoided and the focus is concentrated on obvious features.

[0053] The convolutional neural network for tea posture and color recognition uses a CNN-based image recognition solution, and its network parameters are:

[0054] a. Deep learning framework: pytorch

[0055] b. Network structure: resent50

[0056] c.Learning rate: 0.01

[0057] d.Batch size:32

[0058] e.Optimize:SGD

[0059] f. Loss function: cross-entropy

[0060] g.Epoch:40

[0061] The convolutional neural network training method for tea posture and color recognition includes forward propagation and backward propagation. The specific method is as follows:

[0062] Forward propagation:

[0063] The first step is to select an appropriate tea posture color dataset and divide it into a tea posture color training set, a tea posture color test set, and a tea posture color verification set.

[0064] In the second step, the weights, bias items and sample error values ​​in the convolutional neural network for tea posture and color recognition are randomly initialized, and the initial learning rate is manually set. The characteristic is that the learning rate in the network can affect the degree of increase or decrease in the process of weight adjustment. When the learning rate is too large, it is possible to miss the optimal value in the process of weight adjustment; conversely, it may cause the network to only obtain the local optimum. The range of manually set initial learning rate is generally selected in the interval [0.1, 0.8].

[0065] The third step is to input the data vector of the sample of the tea posture color training set into the network. The vector is continuously transformed through the processing of each layer from the input layer to the output layer. The weights of each layer in the network are different. When the vector passes through a layer, it is multiplied with the weight matrix, and then added with the bias term to output the changed vector matrix. The operation is repeated through each layer of the network until the final output.

[0066] Backward propagation:

[0067] In the first step, the error threshold is calculated based on the actual output value and predicted output value of the sample using the error minimization method. The error value is then back-propagated layer by layer to adjust the parameters of each layer according to the error value.

[0068] In the second step, the adjusted total error value is compared with the error threshold value. If the former is greater than the latter, the network has not achieved the task goal, and then returns to the third step of forward propagation. Otherwise, the next step can be continued.

[0069] The third step is to complete the training and learning task, obtain a trained convolutional neural network and save the parameters of the network.

[0070] After the training is completed, the trained model can be used for prediction. After the training is completed, we can load the trained model into the testing phase, collect damp tea leaves as input, and compare the network's prediction results based on the known collection labels to know the model's prediction effect, and then obtain the model's actual test accuracy and other parameters.

[0071] Furthermore, based on the convolutional neural network for tea leaf posture and color recognition, the damp tea leaves placed in the tea drying equipment are identified, and their degree of moisture is determined based on their curled posture and color, which are different from those of good quality tea leaves. The specific heating time and heating temperature are set in the controller, so that the tea leaves reach the quality of good quality tea after drying.

[0072] Example

[0073] S100, power on the tea dryer, open the entrance 201 and put tea leaves in, and start on the display interface. The intelligent identification module 210 on the ceramic heating ball 209 automatically identifies the type of tea leaves and the difference between the moisture level and normal tea leaves according to the curling degree and color depth of the tea leaves put in, and transmits the information to the controller, which sets the drying temperature for drying.

[0074] S200, after the controller is finished working, the motor drives the transmission gear 203 to make the ceramic heating roller 208 start to fry clockwise, and the completely damp and undrinkable tea leaves and the ordinary damp tea leaves are completely dried for subsequent processing; it is assumed here that the tea leaves entering the intelligent tea drying equipment are all in the same state.

[0075] S300, after the ceramic heating roller 208 completes the drying process, the tea leaves are sorted by the tea sorter 104, the tea leaves that can still be drunk directly enter the first tea storage box 103 through the tea sorting funnel 204, and the tea leaves that cannot be drunk enter the second tea storage box after being crushed by the tea grinder 205, so as to facilitate the subsequent secondary utilization.

[0076] S400, when the tea leaves enter the tea sorter 104, the ceramic heating roller 208 rotates counterclockwise and clockwise in sequence to remove the residual tea leaves in the ceramic heating ball 209 and the ceramic heating roller 208 to ensure the next use of the device;

[0077] S500, before the end of use, all channels remain open, and the ceramic heating roller 208 and the tea grinder 205 operate for 30 seconds so that there is no tea residue in the device.

[0078] exist Figure 1 , Figure 3 In the present intelligent tea drying device, the inlet 201 is at the top of the intelligent tea drying device, and the shell 101 is at the outermost side. The shell 101 supports the tea contact mechanism, and the shell 101 ensures the operation of the tea contact module.

[0079] exist Figure 6 The display interface of the entrance 201 shown in the figure shows the real-time temperature display during the drying process in the upper left corner of the interface, and the right side corresponds to the type of tea identified by the intelligent recognition module 210 and the corresponding recommended drying temperature for the tea, as well as the temperature of the tea at that time after identification. There are 5 buttons below to control the intelligent tea drying equipment, namely: pause, identify, grind, idle cleaning, and lock. The operation process of the intelligent tea drying equipment can be paused, re-identified, re-ground, idle cleaning of residual tea leaves in the equipment, and locked to prevent accidental touch later. The midpoint of the upper edge of the entrance 201 is the center of the circle, and tea leaves can be placed by unscrewing it.

[0080] exist Figure 1 The shell 101 shown has vertical textures for easy gripping during normal use. The shell 101 has two parts, front and back, and the bottom tea storage box 103 is located in the shell 101. At the same time, the observation window 102 is embedded in the upper middle part of the shell 101, and is concentric with the internal ceramic heating ball 209 and the ceramic heating roller 208.

[0081] exist Figure 3 , Figure 4 The internal ceramic heating roller 208 and ceramic heating ball 209 shown in the figure are characterized in that three intelligent recognition modules 210 are arranged in an array on the ceramic heating ball 209. The intelligent recognition module 210 recognizes the corresponding type of tea and the color depth change to determine its moisture content, sets the working temperature of the intelligent tea drying equipment and appropriately controls the gas circulation to ensure that the humidity reaches a predetermined value; the ceramic heating roller 208 is built into the ceramic heating ball 209, and the ceramic heating roller rotating gear 208 is driven and rotated by the rotating gear 203 controlled by the motor. The ceramic heating ball 209 and the ceramic heating roller 208 use ceramic heating (MCH heating element) technology. The ceramic with high insulation and fire resistance is threaded with resistance wire, and the alloy wire is cast in quartz glass. In the semiconductor of the raw material, the ceramic heating roller 208 is built into the ceramic heating ball 209 and connected to the outside world through the upper entrance 201, and is connected to the tea sorter 104 at the bottom. After drying, the switch below the ceramic heating ball 209 is turned on, and the ordinary damp tea leaves after drying are placed into the tea sorting funnel 204 below through the tea sorter 104, and finally the qualified tea leaves are placed in the first tea storage box 103; and the completely damp and inedible tea leaves are made into tea powder by the tea grinder 205 below the tea sorter 104 and can be taken out from the second tea storage box. When fully loaded, the first sorting baffle 206 controls the amount of tea leaves in time, and the first tea storage box 103 pops out the spring sheet 207 to control the taking out and putting in of the first tea storage box 103.

[0082] Working principle: After the intelligent tea drying equipment is powered on, the intelligent recognition module 210 identifies the corresponding type of tea and the color depth change to determine its moisture content, sets the working temperature of the intelligent tea drying equipment and appropriately controls the gas circulation to ensure that the humidity reaches the predetermined value. When the intelligent recognition module 210 identifies the tea, the ceramic heating ball 209 and the ceramic heating roller 208 start to heat, and the ceramic heating roller rotating gear 208 controlled by the motor drives the ceramic heating roller 208 to start rotating clockwise to dry the tea. The ceramic heating (MCH heating element) technology is applied, and the ceramic inner with high insulation and fire resistance is electrically connected. The wire is blocked, and the alloy wire is cast in a semiconductor made of quartz glass as raw material. At the end of drying, the channel below the ceramic heating ball 209 is opened, and the ceramic heating roller 208 continues to work. According to the moisture degree of the tea leaves, the ordinary damp tea leaves are placed in the tea sorting funnel 204 below through the tea sorter 104 after drying, and finally the qualified tea leaves are placed in the first tea storage box 103; and the completely damp and inedible tea leaves are made into tea powder by the tea grinder 205 below the tea sorter 104 and can be taken out from the second tea storage box. The sorting baffle controls the amount of tea leaves in time, and the tea storage box pops out the spring sheet 207 to control the taking out and putting in of the tea storage box.

[0083] Although the present application is described above with reference to specific embodiments, it should be understood by those skilled in the art that many modifications may be made to the configurations and details disclosed in the present application within the principles and scope disclosed in the present application. The scope of protection of the present application is determined by the appended claims, and the claims are intended to cover all modifications included in the equivalent literal meaning or scope of the technical features in the claims.

Claims

1. An intelligent tea drying device, comprising a housing, Features: The shell is provided with an entrance, a drying component, a sorting and processing component and a storage component which are connected in sequence. The drying component is provided with an intelligent recognition module, which is connected to a controller, the drying component is connected to the controller, and the sorting and processing component is connected to the controller; after the intelligent recognition module recognizes the type and moisture degree of the tea leaves, the data is transmitted to the controller, the controller calculates the drying and frying time, and controls the drying component to dry the tea leaves, and stores the dried tea leaves after sorting. The intelligent recognition module includes a convolutional neural network for tea leaf posture and color recognition; the entrance is opened to put in the tea leaves, and the intelligent recognition module automatically recognizes the type and moisture degree of the tea leaves; after the drying component completes the drying process, the tea leaves are sorted by the sorting and processing component; after the tea leaves that have been completely dampened are dried, they are processed by the sorting and processing component for secondary use; The drying component includes a ceramic heating ball, a ceramic heating roller is arranged inside the ceramic heating ball, and the intelligent identification module is arranged on the ceramic heating ball; The sorting and processing assembly comprises a tea sorter, one end of which is connected to the ceramic heating ball, and the other end of which is connected to a first tea passage and a second tea passage respectively, wherein the first tea passage is arranged above the tea sorting hopper, and the second tea passage is arranged above the tea grinder; The storage assembly includes a first tea storage box and a second tea storage box, wherein the first tea storage box is arranged below the tea sorting hopper, and the second tea storage box is arranged below the tea grinder.

2. The intelligent tea drying device according to claim 1, Features: A first sorting baffle is arranged between the first tea storage box and the tea sorting hopper, and a second sorting baffle is arranged between the second tea storage box and the tea grinder.

3. The intelligent tea drying device according to claim 2, Features: The first tea storage box is provided with a pop-up spring piece, and the second tea storage box is provided with a pop-up spring piece.

4. The intelligent tea drying device according to claim 1, Features: There are three intelligent recognition modules, and the three intelligent recognition modules are distributed on the ceramic heating ball in a 120-degree array.

5. The intelligent tea drying device according to claim 3, Features: The intelligent identification module is connected to the display.

6. A method for drying tea leaves using the intelligent tea drying device according to any one of claims 1 to 5, Features: A convolutional neural network for tea leaf posture and color recognition is used to identify tea leaves. The degree of moisture of the tea leaves is determined based on the difference in the curled posture and color of the tea leaves compared to tea leaves of good quality. The heating time and heating temperature are set based on the degree of moisture, so that the tea leaves reach the quality of good quality after drying.

7. The method for drying tea leaves using the intelligent tea drying device according to claim 6, Features: The convolutional neural network training method for tea posture and color recognition includes forward propagation and backward propagation, and the forward propagation includes the following steps: 1) Select the tea posture color dataset and divide it into the tea posture color training set, tea posture color test set and tea posture color verification set; 2) Randomly initialize the weights, bias terms, and sample error values ​​in the convolutional neural network for tea posture and color recognition, and manually set the initial learning rate; 3) Inputting data vectors of samples of the tea posture and color training set into the network, the data vectors are continuously transformed through the processing of each layer from the input layer to the output layer; The back propagation includes the following steps: a) Using the method of minimizing the error, the error threshold is calculated based on the actual output value and the predicted output value of the sample, and then the error value is back-propagated layer by layer to adjust the parameters of each layer according to the error value; b) Compare the adjusted total error value with the error threshold value. If the former is greater than the latter, the network has not achieved the task goal, and then returns to step 3 of the forward propagation. Otherwise, proceed to the next step. c) The training and learning task is completed, a trained convolutional neural network is obtained and the parameters of the network are saved; After the training is completed, the trained model can be used for prediction. At the same time, the trained model can be loaded into the testing phase, and damp tea leaves can be collected as input. According to the known collection labels, the prediction results of the network can be compared to know the prediction effect of the model, and then the actual test accuracy of the model can be obtained.

Images