A licun red intelligent withering machine and a digital withering method thereof

Abstract

The present application belongs to the technical field of tea processing, and particularly relates to a Lichuan red intelligent withering machine and a digital withering method thereof, comprising a withering groove, a heat pump unit, a hot air unit, a vibration leaf turning mechanism, an adjustable inclination angle conveying mechanism, an online moisture content detection device and a main control system. The tea is subjected to hot air withering and yellow light withering from top and bottom. The online moisture content detection device obtains the real-time overall average moisture content of the tea. The main control system is used for judging the moisture content difference degree of the tea in the regions corresponding to the photos taken by different cameras, and judging the moisture content difference degree of the tea in the regions corresponding to the photos and the overall average moisture content calculated based on the weight determined by the weight sensor. The color, texture and shape information of the tea photos in the withering operation region collected by each camera are compared to judge the uniformity of the tea withering process. The vibration leaf turning mechanism and the adjustable inclination angle conveying mechanism cooperate to realize the continuity of the tea withering operation.

Description

Technical Field

[0001] This invention belongs to the field of tea processing technology, and in particular relates to a Lichuan Red intelligent withering machine and its digital withering method. Background Technology

[0002] Lichuan Red Tea, distributed in Lichuan City, Hubei Province, is a fully fermented tea. Due to its unique fermentation process, the tea leaves undergo non-enzymatic auto-oxidation under the action of water and heat, resulting in a tightly rolled and straight appearance, a dark and lustrous color, golden tips, a bright red liquor and leaf base, a rich aroma, and a sweet and mellow taste. It has anti-cancer, antibacterial, and antioxidant functions, so its sales have been increasing in recent years.

[0003] Withering is the first step in processing Lichuan black tea. It involves spreading out freshly picked leaves to allow them to lose their grassy smell, reduce moisture content, concentrate cell sap, alter the physicochemical properties of proteins, and cause enzymes to transition from bound to free states. This process enhances the activity of enzymes such as oxidases and hydrolases, and initiates the hydrolysis of large molecules into simpler substances, laying the foundation for the formation of flavor, aroma, and color compounds in tea. Currently, withering equipment on the market includes trough-type and multi-layer circulating conveyor withering machines. Trough-type withering machines are prone to uneven withering along the length, width, and thickness of the trough, as well as unevenness between different parts of the buds and leaves, and require manual turning. Multi-layer circulating conveyor withering machines have large spaces, making it difficult to control temperature and humidity during the withering process. The continuous circulation of tea leaves during withering can easily cause damage, and the large size of the equipment makes it unsuitable for small-batch production. In addition, judging the degree of withering in production mainly relies on manual experience, which is somewhat subjective. Moisture content, as an important indicator for evaluating the degree of withering, shows that Lichuan black tea processed with a moisture content of around 58% at the end of withering has better quality. With the increasing aging population and the continuous decline in rural labor force, the problem of manpower resources is becoming increasingly prominent. Addressing the problems of high labor intensity, low withering uniformity, and difficulty in turning leaves in the current traditional withering process, there is an urgent need to develop an intelligent withering trough and withering method to solve the processing problems of Lichuan black tea in my country and improve the economic benefits of my country's tea industry. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the purpose of this invention is to provide a Lichuan Red intelligent withering machine and a digital withering method.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] A Lichuan Red intelligent withering machine includes a withering trough, a heat pump unit, a hot air unit, a vibrating leaf turning mechanism, an adjustable tilt conveying mechanism, an online moisture content detection device, and a main control system.

[0007] The wilting trough includes a chassis, air duct, stainless steel mesh, wilting work area, and yellow fluorescent tubes. The frame is divided into three layers from top to bottom. The upper layer is the wilting work area; the upper half of the middle layer contains a vibrating leaf-turning mechanism, and the lower half contains an adjustable tilt conveyor mechanism. The top of the lower layer features an array of yellow fluorescent tubes. Inside the chassis, the wilting work area includes a conveyor trough. The four side walls of the wilting work area are equipped with arrays of yellow fluorescent tubes, and the bottom is a stainless steel mesh fixed to the chassis. Two temperature and humidity sensors are fixedly installed on the stainless steel mesh. The space above the stainless steel mesh is the conveyor trough. A main hot air duct and a branch hot air duct are located in the middle of the upper layer. A camera is located at each of the four corners at the top of the upper layer (i.e., the four corners inside the chassis). In the lower layer, the two ends of the yellow fluorescent tubes are fixed to the chassis by brackets, and the tubes are linearly and equidistantly distributed. The top surface of the chassis has a fixed cover and a movable cover. The stainless steel mesh is used to spread the tea leaves, and yellow light tubes installed on the upper side wall and the lower layer of the stainless steel mesh provide yellow light irradiation; the hot air unit and the heat pump unit wither the tea leaves from the upper and lower parts, respectively.

[0008] Yellow light withering is beneficial for increasing the amino acid content and forming aldehydes in tea leaves, significantly enhancing the aroma of the tea. Temperature and humidity sensors on the stainless steel mesh are used to measure the temperature and humidity of the upper withering area. Suitable temperature and humidity can facilitate the transformation of substances within the tea leaves, thereby improving the quality of the tea.

[0009] The online moisture content detection device includes a weight sensor, a rapid moisture analyzer, and cameras. The weight sensor is installed at the bottom of the casing of the entire withering trough; four cameras are installed at the four corners of the top of the casing of the withering trough, fixed to the frame by brackets, and distributed diagonally; the rapid moisture analyzer is used to obtain the initial moisture content of the tea leaves; the weight sensor is used to obtain the real-time weight of the tea leaves during the withering process, and the real-time overall average moisture content of the tea leaves during the withering process is obtained based on the real-time weight of the tea leaves and the initial moisture content.

[0010] The main control system includes a control box for digital control of the withering process, determining whether the real-time overall average moisture content of the tea leaves reaches the set target moisture content during the withering process; the control box has a built-in main control board, which includes a moisture content monitoring unit. The moisture content monitoring unit is used to determine the degree of difference in the moisture content of the tea leaves in the areas corresponding to photos taken by different cameras, and to determine the degree of difference between the moisture content of the tea leaves in the areas corresponding to the photos and the overall average moisture content calculated based on the weight measured by a weight sensor.

[0011] Preferably, such as Figure 12The heat pump unit includes an evaporator, a gas-liquid separator, a compressor, a condenser, a circulating fan I, and an exhaust fan I. The heat pump unit is located on the left side of the withering tank. A duct connecting the heat pump unit and the withering tank is formed in the lower part between the two. A circulating fan II is installed on one side of the withering tank within the duct, and an exhaust fan II is installed in the upper right corner directly in front of the withering tank. The heat pump unit absorbs heat from the surrounding environment and outputs hot air. The hot air enters the duct and is then sent to the lower layer of the casing by the circulating fan II. The hot air then rises and passes through the stainless steel mesh to wither the lower part of the tea leaves inside. By using the heat absorbed from the surrounding environment for the withering of the tea leaves on the stainless steel mesh, the heat pump unit saves energy and reduces emissions of pollutants such as CO2, achieving energy conservation and emission reduction.

[0012] Preferably, the hot air unit includes a hot air blower, a main hot air duct, and branch hot air ducts. The hot air blower is mounted on a bracket between the chassis and the heat pump unit. The outlet of the hot air blower is connected to the inlet of the main hot air duct. Branch hot air ducts are equidistantly connected to the main hot air duct, and small holes with a diameter of 1 cm are evenly distributed on the branch hot air ducts. The main hot air duct is mounted above a stainless steel mesh and fixed to the chassis by a bracket. The hot air unit is used to deliver hot air to wither the tea leaves above the tea leaves on the stainless steel mesh.

[0013] Existing withering tanks are all ventilated and heated from the bottom, which can easily cause uneven withering of tea leaves between the upper and lower parts. This invention uses a hot air unit and a heat pump unit to wither the tea leaves from the upper and lower parts respectively, which can significantly improve the withering quality of tea leaves and promote a more uniform transformation of substances inside the tea leaves.

[0014] Preferably, the vibrating leaf-turning mechanism includes an excitation motor and vibration springs. The four vibration springs are located below the four corners of the stainless steel mesh, and the excitation motor is located in the middle between every two springs. The vibrating leaf-turning mechanism makes the tea leaves in the stainless steel mesh more evenly dispersed, thus making the withering more uniform.

[0015] Preferably, the adjustable tilt conveyor mechanism includes a bushing I, a right-angle lifter, a drive shaft, a geared motor, a rotary bearing I, a slider, a guide rail, a support, a bushing II, and a rotary bearing II. The adjustable tilt conveyor mechanism is located below the vibrating leaf-turning mechanism. On the left side of the adjustable tilt conveyor mechanism are two right-angle lifters, and on the right side are two supports. The two right-angle lifters and two supports are respectively located below a vibrating spring. The bottom end of the right-angle lifter is installed on the corner of the bottom side of one side of the frame of the middle layer of the machine housing, and the top end is the lifting end connected to the bushing I. One end of the rotary bearing I is connected to the bushing I via a rotating shaft, and the other end is connected to the slider. The slider is connected to the guide rail, and the guide rail is installed below the vibrating spring. The bottom end of the support is installed on the corner of the bottom side of the other side of the frame of the middle layer of the machine housing, and the top end of the support is installed with the bushing II. One end of the rotary bearing II is connected to the bushing II via a connecting shaft, and the other end is installed below the vibrating spring. The geared motor is installed in the middle between the two right-angle lifters and is connected to the right-angle lifters via a drive shaft. Different teas have different flow characteristics. The upper leaf conveying mechanism transports the tea to the stainless steel mesh, and then the movable cover is closed. The adjustable tilt conveying mechanism tilts the stainless steel mesh by 5°, and at the same time, the vibrating leaf turning mechanism is activated to vibrate the stainless steel mesh, so that the tea is evenly spread on the stainless steel mesh.

[0016] Preferably, the main control system includes a control box for digital control of the withering process, determining whether the real-time overall average moisture content of the tea leaves reaches the set target moisture content during withering. A touchscreen is mounted on the surface of the control box, which houses the main control board. The target moisture content, upper temperature limit, and process program are input on the touchscreen. The main control board includes a power supply module, a communication module, a WiFi communication module, a microcontroller main control module, an I / O control module, and a data storage module. The microcontroller main control module includes a moisture content monitoring unit, which in turn includes a data preprocessing module and a data monitoring module.

[0017] The output terminal of the power supply module is electrically connected to the input terminals of the communication module, WiFi communication module, microcontroller main control module, IO control module, and data storage module, and is used to supply power to the communication module, microcontroller main control module, IO control module, and data storage module; the power supply module also serves as the power source for external devices, including electromagnetic relays, touch screens, rapid moisture detectors, weight sensors, temperature and humidity sensors, and alarm lights.

[0018] The communication module is connected to the weight sensor, temperature and humidity sensor, rapid moisture detector, and microcontroller main control module for communication.

[0019] The WiFi communication module is connected to the microcontroller main control module to enable wireless communication with the host computer.

[0020] The output of the IO control module is connected to the input of the electromagnetic relay. Specifically, the output of the first electromagnetic relay is connected to the alarm, the output of the second electromagnetic relay is connected to the hot air blower, the output of the third electromagnetic relay is connected to the heat pump unit, the output of the fourth electromagnetic relay is connected to the vibrating leaf-turning mechanism, the output of the fifth electromagnetic relay is connected to the adjustable tilt conveyor mechanism, the output of the sixth electromagnetic relay is connected to the circulating fan, and the output of the seventh electromagnetic relay is connected to the dehumidifying fan. This module is used to control the start and stop of the hot air blower, the circulating fan, the dehumidifying fan, the heat pump unit, the alarm, the vibrating leaf-turning mechanism, and the adjustable tilt conveyor mechanism.

[0021] The communication module has RS485-1, RS232-1, and UART interfaces. These interfaces provide data exchange channels for external devices, ensuring the normal operation of the system. Specifically, the RS485-1 interface communicates with the weighing sensor and temperature / humidity sensor, while the RS232-1 interface receives data from the rapid moisture analyzer. The UART interfaces include UART-1, UART-2, UART-3, and UART-4. UART-1 transmits the level-converted RS485-1 signal into the microcontroller main control module, and UART-2 transmits the level-converted RS232-1 signal into the microcontroller main control module. UART-4 and UART-3 are directly led out from the microcontroller main control module and are used for communication with the WiFi communication module and the touchscreen, respectively.

[0022] Preferably, the microcontroller main control module is the core of the entire main control system, and is responsible for the acquisition, processing, and transmission of sensor data as well as the control of the withering process.

[0023] Preferably, the data storage module is connected to the microcontroller main control module and is used to store various data. The data storage module is connected to the microcontroller main control module via an IIC-1 interface. After receiving various data, the microcontroller main control module stores them in the data storage module, including initial moisture content, real-time weight, and target moisture content data. During the withering process, the microcontroller main control module transmits the current time and the current weight, current moisture content, current temperature and humidity, and feature parameter data obtained from preprocessed camera photos to the data storage module according to a set time interval. The initial moisture content, initial weight, and target moisture content are displayed on the touchscreen, and various data from the withering process can be output to the host computer.

[0024] Preferably, the working process of the moisture content monitoring unit is as follows: During the withering process, the camera takes pictures of the tea leaves in the withering frame every 20 minutes and transmits the pictures to the data preprocessing module. The data preprocessing module preprocesses the tea leaf pictures to obtain color images and grayscale images. Then, it extracts features from the color images and grayscale images to obtain the morphological feature parameters, color feature parameters, and texture feature parameters of the tea leaves. The data monitoring module is loaded into the tea leaf moisture content monitoring model. The tea leaf moisture content monitoring model outputs the real-time moisture content of the tea leaves based on the real-time morphological feature parameters, color feature parameters, and texture feature parameters of the tea leaves. This allows the model to determine the degree of difference in the moisture content of the tea leaves in the areas corresponding to the pictures taken by different cameras, as well as the degree of difference between the moisture content of the tea leaves in the areas corresponding to the pictures and the overall average moisture content of the tea leaves calculated based on the weight measured by the weight sensor. According to the formula: Threshold = |Moisture content of tea leaves in the area corresponding to the photo - Overall average moisture content| / Overall average moisture content × %, if the threshold calculated based on the moisture content of tea leaves in the areas corresponding to the photos taken by the four cameras and the overall average moisture content are both ≤ 5%, it indicates that the tea leaves on the stainless steel mesh are withering evenly, and the withering process continues; if the threshold calculated based on the moisture content of tea leaves in the areas corresponding to the photos taken by one or more cameras and the overall average moisture content is > 5%, it indicates that the tea leaves on the stainless steel mesh are not withering evenly, and the adjustable tilt conveyor is activated. The mechanism tilts the stainless steel mesh by 5°, then starts the vibrating motor to vibrate the mesh, causing the tea leaves to tumble and mix evenly. If the overall average moisture content reaches the set target moisture content, the heat pump unit, hot air unit, and yellow light tube are turned off, stopping the air supply heating and yellow light irradiation. Then, the movable cover is opened, and the adjustable tilt conveyor mechanism is started, making the right side of the withering frame lower than the left side. At the same time, the vibrating motor starts, vibrating the stainless steel mesh and causing the tea leaves to flow out from the outlet on the right side of the withering area. When all the tea leaves on the stainless steel mesh have fallen from the outlet, the withering of a new batch of tea leaves begins again.

[0025] Preferably, the photos of tea leaves on the stainless steel mesh taken by the camera are processed by the data preprocessing module to obtain tea leaf morphological feature parameters, color feature parameters, and texture feature parameters, including rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrix. These seven types of data are input into the GA-BP neural network model to output the tea leaf moisture content. Based on the output tea leaf moisture content, the degree of tea leaf withering is judged, thereby ensuring that each withering process can produce tea leaves with a similar degree of withering. This standardizes the tea leaf withering process and avoids the bias caused by subjective judgment of the degree of tea leaf withering, which leads to different withering degrees in each batch of tea leaves.

[0026] The specific process of the data preprocessing module for processing tea photos is as follows: (1) Process the tea photos into color images and grayscale images; (2) Extract features from the color images and grayscale images to obtain the morphological feature parameters, color feature parameters, and texture feature parameters of the tea. The morphological feature parameters include the major axis, minor axis, rectangularity, roundness, compactness, diagonal length, and thinness of the tea. The color feature parameters are the color feature parameters of the tea, including the first moment of red, green, blue, hue, saturation, and brightness. The texture feature parameters include the contrast, second moment of angle, correlation, entropy, and inverse difference matrix of the tea.

[0027] (3) The random forest model built into the data preprocessing module is used to sort the importance of morphological feature parameters, color feature parameters, and texture feature parameters, and the important feature parameters of the withering process are selected as rectangularity, compactness, greenness, saturation, contrast, correlation and inverse difference matrix.

[0028] Preferably, the tea moisture content monitoring model is a genetic algorithm-optimized backpropagation neural network model (GA-BP model). The specific construction process is as follows: (1) The data preprocessing module calls the historical data in the data storage module. The historical data is existing photos of tea leaves during the withering process and the corresponding tea moisture content. Then, the existing photos of tea leaves during the withering process are processed to obtain the rectangle, compactness, greenness, saturation, contrast, correlation, and inverse difference matrix. Then, the obtained rectangle, compactness, greenness, saturation, contrast, correlation, and inverse difference matrix and their corresponding tea moisture content are normalized. Finally, the normalized data is divided into training set and validation set according to a 4:1 ratio. (2) The microcontroller main control module transmits the above training set and validation set to the host computer through the WIFI communication module as input data for the GA-BP model. (3) In the host computer, the genetic algorithm is used to optimize the weights and biases of the 7-10-1 structured single hidden layer BP neural network to obtain The optimized BP neural network is constructed with the following parameters: the number of iterations of the genetic algorithm is 200, the population size is 20, the selection function is normGeomSelect, the mutation function is nonUnifMutation, the crossover probability is 0.5, and the mutation probability is 0.05. The BP neural network is constructed with the following parameters: the number of neurons in the hidden layer is 10, the activation function of the hidden layer is tansig, the activation function of the output layer is tansig, the training function is Trainlm, the learning rate is 0.01, and the number of iterations is 1000. (4) The optimized BP neural network is trained and validated using the training set and the validation set. The hyperparameters of the model are adjusted according to each index to obtain the GA-BP model. The normalized rectangleness, compactness, greenness, saturation, contrast, correlation and inverse difference matrix are used as input data, and the normalized tea moisture content is used as output data. The performance of the model is evaluated using the coefficient of determination, mean square error, mean absolute error and other indicators. The calculation formula of each indicator is:

[0029]

[0030] (5) The host computer transmits the obtained GA-BP model to the microcontroller main control module, and it is loaded into the data monitoring module as the tea moisture content monitoring model.

[0031] Preferably, the operation process of the intelligent withering machine is as follows: The movable cover is opened, and fresh tea leaves are conveyed to the withering area via the upper leaf conveyor mechanism. Then, the adjustable tilt conveyor tilts the stainless steel mesh by 5°, allowing the tea leaves to slide within the mesh. Simultaneously, the vibrating motor of the leaf-turning mechanism begins to vibrate, ensuring the tea leaves are evenly spread on the stainless steel mesh. Once the tea leaves are evenly spread on the mesh, the vibrating motor stops vibrating, and the adjustable tilt conveyor mechanism returns the stainless steel mesh to a horizontal position. Next, the movable cover is closed, a rapid moisture detector obtains the initial moisture content of the tea leaves, and a weight sensor obtains the initial weight of the tea leaves, commencing the withering process. The process begins with the activation of the yellow light tubes, heat pump unit, hot air unit, and online moisture content detection device. Hot air from the heat pump unit, circulated by a fan, withers the lower and upper parts of the tea leaves from below the stainless steel mesh, while hot air from the hot air unit withers the upper and lower parts of the tea leaves from above the stainless steel mesh, respectively. Simultaneously, a weight sensor acquires the real-time weight of the tea leaves, and a camera captures a photo of the tea leaves inside the stainless steel mesh every minute. All data is stored in real-time in the data storage module. During the withering process, the data is processed based on the real-time tea leaf photos. The tea moisture content monitoring model outputs the tea moisture content of the area corresponding to the photograph. Based on the real-time tea weight, the moisture content monitoring unit calculates the real-time overall average moisture content of the tea. By comparing the tea moisture content of the area corresponding to the photograph with the overall average moisture content, the degree of difference in tea moisture content between areas photographed by different cameras is determined. Furthermore, the difference between the tea moisture content of the area corresponding to the photograph and the overall average moisture content calculated based on the weight measured by the weight sensor is controlled. If the tea moisture content is uneven, the adjustable tilt conveyor mechanism is first activated to tilt the stainless steel mesh, and then the vibration is activated. The motor vibrates the stainless steel mesh, causing the tea leaves to tumble and mix evenly. After mixing, the motor is turned off, and the adjustable tilt conveyor mechanism is started to return the stainless steel mesh to a horizontal position. When the overall average moisture content of the tea leaves in the withering area reaches the set target moisture content, the yellow light tube, heat pump unit, hot air unit, and online moisture content detection device are turned off. Then, the adjustable tilt conveyor mechanism is started to tilt the stainless steel mesh, and the motor is turned on to vibrate the stainless steel mesh, causing the tea leaves to flow from left to right and exit from the outlet on the right side of the stainless steel mesh. When all the tea leaves on the stainless steel mesh have been unloaded from the outlet, the withering of a new batch of tea leaves begins.

[0032] The calculation method for the real-time overall average moisture content of the tea is as follows: the moisture content monitoring unit calculates the current overall average moisture content Q2 according to the following formula based on the stored initial weight W1, initial moisture content Q1 and current real-time tea weight W2, and stores it in the data storage module.

[0033]

[0034] Preferably, the target moisture content is set at 58%.

[0035] A digital withering method for the aforementioned Lichuan Red intelligent withering machine includes the following steps:

[0036] Step 1: Start the Lichuan Red Intelligent Withering Machine, open the movable cover, and the tea leaves are conveyed to the stainless steel mesh through the upper leaf conveying structure. Then, start the adjustable tilt conveying mechanism to tilt the stainless steel mesh by 5° and then stop. Next, start the vibrating motor to vibrate the stainless steel mesh, so that the tea leaves slide on the stainless steel mesh until the tea leaves are evenly spread on the stainless steel mesh. Afterward, start the adjustable tilt conveying mechanism to restore the stainless steel mesh to the horizontal position. The feeding is finished, and close the movable cover.

[0037] Step 2: The microcontroller main control module sends an inquiry message to the weight sensor. After receiving the inquiry message, the weight sensor measures the initial weight of the tea leaves and then sends the initial weight of the tea leaves to the microcontroller main control module. Similarly, the rapid moisture detector measures the initial moisture content of the tea leaves, and the temperature and humidity sensor measures the temperature and humidity in the withering area and sends the initial moisture content of the tea leaves and the temperature and humidity in the withering area to the microcontroller main control module. The withering area is the temperature and humidity about 10cm above the stainless steel mesh. Specifically, the microcontroller main control module obtains the initial moisture content Q1 of the tea leaves collected by the rapid moisture detector through the RS232-1 interface of the communication module, obtains the initial weight W1 of the tea leaves collected by the weight sensor and the temperature and humidity data collected by the temperature and humidity sensor through the RS485-1 interface. Then, the microcontroller main control module sends the acquired data to the WiFi communication module and the touch screen through the UART-4 and UART-3 interfaces. The touch screen displays the data, the WiFi communication module wirelessly transmits the data to the host computer, and the microcontroller main control module sends the acquired data to the data storage module for storage through the IIC-1 interface.

[0038] Step 3: Input the target moisture content, upper temperature limit, lower temperature limit, upper humidity limit, lower humidity limit, and process program on the touch screen, and click "Start Withering". The withering process begins, activating the yellow light tubes, heat pump unit, hot air unit, online moisture content detection device, exhaust fan, and circulating fan. The hot air output from the heat pump unit is circulated from below the stainless steel mesh by the circulating fan, and the hot air output from the hot air unit is circulated from above the stainless steel mesh to wither the lower and upper parts of the tea leaves, respectively. The yellow light tubes on the upper and lower layers of the chassis wither the upper and lower parts of the tea leaves from above and below the stainless steel mesh, respectively. Simultaneously, the weight sensor acquires the real-time weight of the tea leaves, and the camera takes a picture of the tea leaves inside the stainless steel mesh every set time. All data is stored in real time in the data storage module.

[0039] When the temperature and humidity sensor detects a current temperature higher than the input upper temperature limit, the microcontroller main control module sends signals to activate the alarm, deactivate the hot air blower, and deactivate the heat pump unit. This activates the alarm, deactivates the hot air blower and heat pump unit, while the exhaust fan and circulating fan continue operating to lower the temperature in the withering area and prevent excessive moisture loss from the tea leaves. This continues until the current temperature is no higher than the upper temperature limit, at which point the alarm is deactivated and the hot air blower and heat pump unit are activated. Similarly, when the temperature and humidity sensor detects a current temperature lower than the input lower temperature limit, the microcontroller main control module sends signals to activate the alarm, increase the power of the hot air blower, and increase the power of the heat pump unit. This activates the alarm and increases the temperature of the hot air delivered by the hot air blower and heat pump unit to raise the temperature of the tea leaves during the withering process. The system adjusts the temperature of the operating area to prevent the tea leaves from becoming bitter or brittle due to excessively low temperatures. Once the current temperature is no lower than the input lower limit, the power of the heat pump unit and the hot air blower return to their original power. Similarly, when the current humidity, as monitored by the temperature and humidity sensor, is higher than the input upper limit, the microcontroller main control module sends a signal to increase the power of the exhaust fan, improving its dehumidification efficiency. This continues until the current humidity is no higher than the upper limit, at which point the exhaust fan power returns to its original power. Likewise, when the current humidity, as monitored by the temperature and humidity sensor, is lower than the input lower limit, the microcontroller main control module sends a signal to decrease the power of the exhaust fan, reducing its dehumidification efficiency. This continues until the current humidity is no lower than the lower limit, at which point the exhaust fan power returns to its original power.

[0040] Step 4: Based on the stored initial weight W1, initial moisture content Q1, and current real-time weight W2, the moisture content monitoring unit calculates the current overall average moisture content Q2 according to the following formula and stores it in the data storage module.

[0041]

[0042] Step 5: The data preprocessing module processes the tea photos to obtain important feature parameters of the withering process. The specific process is as follows: (1) Process the tea photos into color images and grayscale images; (2) Extract features from the color images and grayscale images to obtain the morphological feature parameters, color feature parameters, and texture feature parameters of the tea. The morphological feature parameters include the major axis, minor axis, rectangularity, roundness, compactness, diagonal length, and thinness of the tea. The color feature parameters are the color feature parameters of the tea, including the first moment of red, green, blue, hue, saturation, and brightness. The texture feature parameters include the contrast, second moment of angle, correlation, entropy, and inverse difference matrix of the tea; (3) Use the random forest model built into the data preprocessing module to sort the importance of the morphological feature parameters, color feature parameters, and texture feature parameters, and select the important feature parameters of the withering process as rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrix.

[0043] Step 6: The microcontroller main control module stores the rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrices obtained in Step 5 into the data storage module; simultaneously, it inputs the rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrices as input data into the tea moisture content monitoring model in the data monitoring module. The tea moisture content monitoring model outputs the tea moisture content Q3, and compares Q3 with Q2 according to the following formula:

[0044] Threshold = |Moisture content of tea leaves in the area corresponding to the photo Q3 - Overall average moisture content Q2| / Overall average moisture content Q2 × %;

[0045] If the thresholds calculated based on the tea moisture content Q3 and the overall average moisture content Q2 of the corresponding areas captured by the four cameras are both ≤5%, it indicates that the tea on the stainless steel mesh is withered evenly, and the withering process continues; if the thresholds calculated based on the tea moisture content Q3 and the overall average moisture content Q2 of the corresponding areas captured by one or more cameras are >5%, it indicates that the tea on the stainless steel mesh is not withered evenly, then the adjustable tilt conveyor mechanism is activated to tilt the stainless steel mesh by 5°, and then the excitation motor is activated to excite the stainless steel mesh, so that the tea on it is turned over and mixed evenly.

[0046] If the overall average moisture content Q2 reaches the set target moisture content, then the heat pump unit, hot air unit, and yellow light tube are turned off, and the air supply heating and yellow light irradiation are stopped. Then, the movable cover is opened, and the adjustable tilt conveyor mechanism is started to make the right side of the withering frame lower than the left side. At the same time, the excitation motor is started to excite the stainless steel mesh, causing the tea leaves to flow out from the outlet on the right side of the withering operation area. When all the tea leaves on the stainless steel mesh have been unloaded from the outlet, the withering of a new batch of tea leaves is restarted.

[0047] Compared with the prior art, the beneficial effects and advantages of the present invention are as follows:

[0048] 1. In this invention, the up-and-down vibration of the leaf-turning mechanism causes the tea leaves inside the stainless steel mesh 6 to vibrate up and down, thereby achieving uniform distribution of the tea leaves inside the stainless steel mesh 6 and spreading them flat on the stainless steel mesh 6.

[0049] 2. In this invention, by comparing the color and texture information of tea photos taken by each camera in the withering operation area, the uniformity of the tea withering process is determined, thereby controlling the operation of the vibrating turning mechanism, thereby improving the uniformity of the tea withering process and improving the quality of the tea.

[0050] 3. In this invention, the combined use of the vibrating leaf-turning mechanism and the adjustable tilt conveying mechanism enables continuous withering operations in tea processing, saving labor. By measuring the weight of the tea leaves with a weight sensor and collecting color and texture data of the tea leaves in the withering frame with a camera, the moisture content of the tea leaves during withering is predicted, thus realizing intelligent withering operations and effectively improving work efficiency.

[0051] 4. This invention uses a rapid moisture detector in conjunction with a weight sensor to automatically acquire moisture content and weight data of fresh tea leaves. The main control system automatically calculates the overall moisture content of the tea leaves. Compared with traditional methods, the moisture content data is more accurate, the operation is simple and the time consumption is shorter. It can more accurately predict the moisture content of withered leaves. Users do not need to frequently manually check the degree of withering, and the workload is greatly reduced, which is conducive to improving the stability of the quality of finished tea.

[0052] 5. This invention allows the withering process to be programmed according to production needs, enabling timed operation and saving labor. At the same time, the equipped touch screen displays production data and control switches intuitively, providing a user-friendly human-machine interface that facilitates application and promotion in production.

[0053] 6. By equipping the data storage module, this invention can record initial production data and process data, avoiding the loss of initial data due to unexpected downtime. In addition, it can transmit data to the host computer via the UART-4 port and WiFi communication module, providing data support for improving tea quality in the future.

[0054] 7. This invention simplifies the wiring process in the production environment by equipping a WiFi communication module to achieve wireless communication, data interaction and control command transmission with a host computer.

[0055] 8. This invention provides control software for the host computer, enabling data interaction and control command transmission between the computer and the main control system, real-time calculation of the current water loss rate of the wilting leaves, and prediction of the wilting completion time.

[0056] 9. This invention discloses a method for predicting the moisture content of tea leaves during the withering process based on random forest features and a genetic algorithm-optimized backpropagation neural network. The method predicts the moisture content of withered leaves by image acquisition, which is simple to operate and takes less time; it can predict the moisture content of withered leaves relatively accurately.

[0057] In summary, this invention, through the coordinated use of various mechanisms, solves problems such as poor uniformity and difficulty in automatic leaf turning and loading / unloading during the tea withering process, thereby improving the quality of Lichuan Red Tea and giving it high efficiency, high standards, and high intelligence. This is of great significance to the development of Lichuan Red Tea. Attached Figure Description

[0058] Figure 1This is a front view of the structure of the present invention;

[0059] Figure 2 This is a top view of the structure of the present invention;

[0060] Figure 3 This is a side view of the structure of the present invention from direction A;

[0061] Figure 4 This is a side view of the structure of the present invention from direction B;

[0062] Figure 5 This is a left view of the tilt-adjustable vibratory conveyor structure of the present invention;

[0063] Figure 6 This is a right view of the tilt-adjustable vibratory conveyor structure of the present invention;

[0064] Figure 7 This is a schematic diagram of the lifting structure of the tilt-adjustable vibratory conveyor of the present invention;

[0065] Figure 8 This is a schematic diagram of the support for the tilt-adjustable vibratory conveyor of the present invention;

[0066] Figure 9 This is a structural diagram of the main control system of the present invention;

[0067] Figure 10 This is the main program flowchart of the present invention;

[0068] Figure 11 This is a flowchart illustrating the wilting moisture content prediction method based on random forest-backpropagation neural network of the present invention.

[0069] Figure 12 This is a schematic diagram illustrating the working principle of the heat pump unit of the present invention;

[0070] The components in the diagram are numbered as follows: 1. Heat pump unit; 2. Rapid moisture detector; 3. Hot air blower; 4. Camera; 5. Vibration spring; 6. Stainless steel mesh; 7. Wilting work area; 8. Excitation motor; 9. Main hot air duct; 10. Hot air branch duct; 11. Exhaust fan II; 12. Adjustable tilt conveyor mechanism; 12-1. Conveyor trough; 12-2. Bushing I; 12-3. Right angle lifter; 12-4. Drive shaft; 12-5. Gear motor; 12-6. Rotary bearing I; 12-7. Slider; 12-8. Support. 12-9, Bushing II; 12-10, Rotating Bearing II; 12-11, Connecting Shaft; 12-12, Rotating Shaft; 12-13, Chassis; 13, Yellow Lamp Tube; 14, Weight Sensor; 15, Circulating Fan II; 16, Control Box; 17, Air Duct; 18, Upper Blade Conveying Mechanism; 19, Fixed Cover; 20, Moving Cover; 21, Evaporator; 22, Gas-Liquid Separator; 23, Compressor; 24, Condenser; 25, Circulating Fan I; 26, Dehumidifying Fan I; 27, Temperature and Humidity Sensor; 28, Alarm Light; 29. Detailed Implementation

[0071] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0072] Example 1

[0073] A Lichuan Red intelligent withering machine, such as Figure 1-8 It includes a withering tank, a heat pump unit, a hot air unit, a vibrating leaf turning mechanism, an adjustable tilt conveying mechanism 12, an online moisture content detection device, and a main control system.

[0074] The wilting tank includes a housing 13, an air duct 18, a stainless steel mesh 6, a wilting work area 7, and yellow light tubes 14. The housing 13 is divided into three layers from top to bottom. The upper layer is the wilting work area 7. The upper half of the middle layer is a vibrating leaf-turning mechanism, and the lower half is an adjustable tilt conveyor mechanism 12. The top of the lower layer is equipped with an array of yellow light tubes 14. Inside the housing 13, the wilting work area 7 includes a conveyor trough 12-1. The four side walls of the wilting work area 7 are equipped with an array of yellow light tubes 14. The bottom is a stainless steel mesh 6 fixedly connected to the housing 13. Two temperature and humidity sensors 28 are fixedly installed on the stainless steel mesh 6. The space above the stainless steel mesh 6 is the conveyor trough 12-1. A hot air main duct 9 and a hot air branch duct 10 are located in the middle of the upper layer. A camera 4 is installed at each of the four corners at the top of the upper layer, i.e., at the four corners inside the housing 13. In the lower layer, the two ends of the yellow lamp tubes 14 are fixed to the casing 13 by brackets, and the yellow lamp tubes 14 are linearly and equidistantly distributed. The top surface of the casing 13 has a fixed cover 20 and a movable cover 21. The stainless steel mesh 6 is used to spread the tea leaves. The yellow lamp tubes 14 installed on the upper side wall of the stainless steel mesh 6 and the yellow lamp tubes 14 installed in the lower layer provide yellow light irradiation. Yellow light withering is beneficial to increasing the amino acid content and forming aldehydes in the tea leaves, significantly enhancing the aroma of the tea. The temperature and humidity sensor 28 on the stainless steel mesh 6 is used to measure the temperature and humidity of the upper withering area 7. Suitable temperature and humidity can be beneficial to the transformation of the substances contained in the tea leaves, thereby improving the quality of the tea.

[0075] like Figure 12The heat pump unit includes an evaporator 22, a gas-liquid separator 23, a compressor 24, a condenser 25, a circulating fan I 26, and an exhaust fan I 27. The heat pump unit 1 is located on the left side of the withering tank. A duct 18 connecting the heat pump unit 1 and the withering tank is formed in the lower part between the two. A circulating fan II 16 is installed on one side of the withering tank via the duct 18, and an exhaust fan II 11 is installed in the upper right corner directly in front of the withering tank. The heat pump unit absorbs heat from the surrounding environment and outputs hot air. The hot air enters the duct 18 and is sent to the lower layer of the casing 13 by the circulating fan II 16. Then, the hot air rises and passes through the stainless steel mesh 6 to wither the lower part of the tea leaves inside. By using the heat absorbed from the surrounding environment for the withering of the tea leaves on the stainless steel mesh 6, the heat pump unit can save energy and reduce emissions of pollutants such as CO2, achieving energy conservation and emission reduction.

[0076] The hot air unit includes a hot air blower 3, a main hot air duct 9, and branch hot air ducts 10. The hot air blower 3 is mounted on a bracket between the casing 13 and the heat pump unit 1. The outlet of the hot air blower 3 is connected to the inlet of the main hot air duct 9. Branch hot air ducts 10 are equidistantly connected to the main hot air duct 9, and small holes with a diameter of 1 cm are evenly distributed on the branch hot air ducts 10. The main hot air duct 9 is mounted above a stainless steel mesh and fixed to the casing 13 by a bracket. The hot air unit is used to deliver hot air to wither the tea leaves from above on the stainless steel mesh 6. Existing withering tanks are all bottom-ventilated and heated, which easily causes uneven withering of the upper and lower tea leaves. This invention uses a hot air unit and a heat pump unit to wither the tea leaves from the upper and lower parts respectively, which can significantly improve the withering quality of the tea leaves and promote a more uniform transformation of substances inside the tea leaves.

[0077] The vibrating leaf-turning mechanism includes an excitation motor 8 and vibration springs 5. The four vibration springs 5 ​​are located below the four corners of the stainless steel mesh 6, and the excitation motor 8 is located in the middle between every two springs. The vibrating leaf-turning mechanism makes the tea leaves in the stainless steel mesh 6 more evenly dispersed, thus making the withering more uniform.

[0078] The adjustable tilt conveyor mechanism 12 includes a bushing I 12-2, a right-angle lifter 12-3, a drive shaft 12-4, a reduction motor 12-5, a rotating bearing I 12-6, a slider 12-7, a guide rail 12-8, a support 12-9, a bushing II 12-10, and a rotating bearing II 12-11. The adjustable tilt conveyor mechanism 12 is located below the vibrating leaf-turning mechanism. On the left side of the adjustable tilt conveyor mechanism 12 are two right-angle lifters 12-3, and on the right side are two supports 12-9. The two right-angle lifters 12-3 and the two supports 12-9 are respectively located below a vibrating spring 5. The bottom end of the right-angle lifter 12-3 is installed on the corner of the bottom side of the frame of the middle layer of the housing 13, and the top end is the lifting end connected to the bushing I 12-2. One end of the rotating bearing I 12-6 is connected to the bushing I 12-2 through the rotating shaft 12-13, and the other end is connected to the slider 12-11. 2-7 connection, slider 12-7 is connected to guide rail 12-8, guide rail 12-8 is installed below vibration spring 5; the bottom end of support 12-9 is installed on the corner of the bottom end of the frame of the middle layer of the machine box 13, and the top end of support 12-9 is installed with bushing II, one end of rotating bearing II 12-11 is connected to bushing II 12-10 through connecting shaft 12-12, and the other end is installed below vibration spring 5; reduction motor 12-5 is installed in the middle between two right angle lifters 12-3, and is connected to right angle lifter 12-3 through transmission shaft 12-4. Different teas have different flow characteristics. The upper leaf conveying mechanism 19 conveys the tea to the stainless steel mesh 6, then closes the movable cover, tilts the stainless steel mesh 6 by 5° through the adjustable tilt angle conveying mechanism 12, and at the same time starts the vibrating leaf turning mechanism to excite the stainless steel mesh 6, so that the tea is evenly spread on the stainless steel mesh 6.

[0079] The online moisture content detection device includes a weight sensor 15, a rapid moisture analyzer 2, and four cameras 4. The weight sensor 15 is installed below the casing 13 of the entire withering trough. Four cameras 4 are installed at the four corners of the top of the casing 13, fixed to the frame 13 by brackets, and arranged diagonally. The rapid moisture analyzer 2 is used to obtain the initial moisture content of the tea leaves. The weight sensor 15 is used to obtain the real-time weight of the tea leaves during the withering process, and based on the real-time weight and initial moisture content, the real-time overall average moisture content of the tea leaves during withering is obtained.

[0080] The main control system includes a control box 17, which is used for digital control of the withering process to determine whether the real-time overall average moisture content of the tea leaves has reached the set target moisture content. A touchscreen is mounted on the surface of the control box 17, which houses the main control board. The target moisture content, upper temperature limit, and process program are input on the touchscreen. The main control board includes a power supply module, a communication module, a WiFi communication module, a microcontroller main control module, an I / O control module, and a data storage module. The microcontroller main control module includes a moisture content monitoring unit, which in turn includes a data preprocessing module and a data monitoring module.

[0081] The output terminal of the power supply module (manufacturer: MORNSUN, model: K7805-2000R3) is electrically connected to the input terminals of the communication module, WiFi communication module, microcontroller main control module, IO control module, and data storage module, and is used to supply power to the communication module, microcontroller main control module, IO control module, and data storage module; the power supply module also serves as the power source for external devices, including electromagnetic relays, touch screens, rapid moisture detectors 2, weight sensors 15, temperature and humidity sensors 28, and alarm lights 29.

[0082] The communication module (manufacturer: MORNSUN, model: TD331S232H) connects with the weight sensor, temperature and humidity sensor, rapid moisture analyzer, and microcontroller main control module for data communication with the weight sensor (Porston, model: PSD-F1), temperature and humidity sensor (Dekong Senshe, model: DK-3000-WS), rapid moisture analyzer (Shanghai Lichen Bangxi Instrument Technology Co., Ltd., model: LC-DHS-16), and microcontroller main control module.

[0083] The WiFi communication module (Ankexin, model: ESP-07S) is connected to the microcontroller main control module to realize wireless communication with the host computer.

[0084] The output of the IO control module is connected to the input of an electromagnetic relay (manufacturer: TONGLING, model: T90-5VDC-TL-C). Specifically, the output of electromagnetic relay 1 is connected to the alarm 29, the output of electromagnetic relay 2 is connected to the hot air blower 3, the output of electromagnetic relay 3 is connected to the heat pump unit 1, the output of electromagnetic relay 4 is connected to the vibrating leaf-turning mechanism, the output of electromagnetic relay 5 is connected to the adjustable tilt conveyor mechanism 12, the output of electromagnetic relay 6 is connected to the circulating fan 11, and the output of electromagnetic relay 7 is connected to the dehumidifying fan 11. This module is used to control the start and stop of the hot air blower 3, the circulating fan 16, the dehumidifying fan 11, the heat pump unit 1, the alarm 29, the vibrating leaf-turning mechanism, and the adjustable tilt conveyor mechanism 12.

[0085] The microcontroller main control module (manufacturer: JLCPCB, model: GD32F470ZGT6) is the core of the entire main control system, responsible for the acquisition, processing and transmission of sensor data and the control of the withering process;

[0086] The data storage module (manufacturer: MICROCHIP, model: AT24C256C-SSHL-T) is connected to the microcontroller main control module and is used to store various types of data.

[0087] The communication module has RS485-1, RS232-1, and UART interfaces. These interfaces provide data exchange channels for external devices, ensuring the normal operation of the system. Specifically, the RS485-1 interface communicates with the weighing sensor and temperature / humidity sensor, while the RS232-1 interface receives data from the rapid moisture analyzer. The UART interfaces include UART-1, UART-2, UART-3, and UART-4. UART-1 transmits the level-converted RS485-1 signal into the microcontroller main control module, and UART-2 transmits the level-converted RS232-1 signal into the microcontroller main control module. UART-4 and UART-3 are directly led out from the microcontroller main control module and are used for communication with the WiFi communication module and the touchscreen, respectively.

[0088] The data storage module is connected to the microcontroller main control module via an IIC-1 interface. The microcontroller main control module receives various data and stores it in the data storage module, including initial moisture content, real-time weight, and target moisture content data. During the withering process, the microcontroller main control module transmits the current time, current weight, current moisture content, current temperature and humidity, and pre-processed feature parameters obtained from photos taken by camera 4 to the data storage module according to a set time interval. The initial moisture content, initial weight, and target moisture content are displayed on the touchscreen, and various data from the withering process can be output to the host computer.

[0089] The working process of the moisture content monitoring unit is as follows: During the withering process, camera 4 takes pictures of the tea leaves in the withering frame every 20 minutes and transmits the pictures to the data preprocessing module. The data preprocessing module preprocesses the tea leaf pictures to obtain color and grayscale images. Then, it extracts features from the color and grayscale images to obtain the morphological, color, and texture parameters of the tea leaves. The data monitoring module is loaded into the tea moisture content monitoring model. The tea moisture content monitoring model outputs the real-time moisture content of the tea leaves based on the real-time morphological, color, and texture parameters of the tea leaves. This allows the system to determine the degree of difference in the moisture content of the tea leaves in the areas corresponding to the pictures taken by different cameras 4, as well as the degree of difference between the moisture content of the tea leaves in the areas corresponding to the pictures and the overall average moisture content of the tea leaves calculated based on the weight measured by the weight sensor. According to the formula: Threshold = |Moisture content of tea leaves in the area corresponding to the photo - Overall average moisture content| / Overall average moisture content × %, if the threshold calculated based on the moisture content of tea leaves in the areas corresponding to the photos taken by the four cameras and the overall average moisture content are both ≤ 5%, it indicates that the tea leaves on the stainless steel mesh 6 are withered evenly, and the withering process continues; if the threshold calculated based on the moisture content of tea leaves in the areas corresponding to the photos taken by one or more cameras and the overall average moisture content is > 5%, it indicates that the tea leaves on the stainless steel mesh 6 are not withered evenly, then the adjustable tilt conveyor mechanism 12 is activated to make the stainless steel mesh 6 wither evenly. The steel mesh 6 is tilted at 5°, and the vibrating motor 8 is started to vibrate the stainless steel mesh 6, causing the tea leaves on it to turn over and mix evenly. If the overall average moisture content reaches the set target moisture content, the heat pump unit 1, the hot air unit, and the yellow light tube 14 are turned off, and the air supply heating and yellow light irradiation are stopped. Then, the movable cover 21 is opened, and the adjustable tilt conveyor mechanism 12 is started, so that the right side of the withering frame is lower than the left side. At the same time, the vibrating motor 8 is started to vibrate the stainless steel mesh 6, causing the tea leaves to flow out from the outlet on the right side of the withering operation area 7. When all the tea leaves on the stainless steel mesh 6 have been unloaded from the outlet, the withering of a new batch of tea leaves is started again.

[0090] The photos of tea leaves on the stainless steel mesh 6 taken by camera 4 are processed by the data preprocessing module to obtain tea leaf morphological feature parameters, color feature parameters, and texture feature parameters, including rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrix. These seven types of data are input into the GA-BP neural network model to output the tea leaf moisture content. Based on the output tea leaf moisture content, the degree of tea leaf withering is judged, thereby ensuring that each withering process can produce tea leaves with a similar degree of withering. This standardizes the tea leaf withering process and avoids the bias caused by subjective judgment of the degree of tea leaf withering by humans, which would result in different degrees of withering for each batch of tea leaves.

[0091] The specific process of the data preprocessing module for processing tea photos is as follows: (1) Process the tea photos into color images and grayscale images; (2) Extract features from the color images and grayscale images to obtain the morphological feature parameters, color feature parameters, and texture feature parameters of the tea. Specifically, the color feature parameters are obtained by using the colorhist function to calculate the color image obtained from the processed tea photo, and the texture feature parameters are obtained by using the grayscale image obtained from the processed tea photo by using the grayscale image by using the grayscale function. The morphological feature parameters are obtained by using median filtering, followed by the Otsu method, then the erosion-dilation method, and finally the regionprops function. The morphological feature parameters include the major axis, minor axis, rectangularity, roundness, compactness, diagonal length, and thinness of the tea. The expressions for these five morphological feature parameters are:

[0092]

[0093] J = D / L

[0094]

[0095] Where R is rectangularity, E is circularity, J is compactness, Len is linearity, T is slenderness, A is the area of ​​the tea outline, C is the perimeter of the tea outline, D is the diameter of the circle with the same area as the tea outline, L is the length (major axis) of the circumscribed rectangle of the tea outline, and S is the width (minor axis) of the circumscribed rectangle of the tea outline.

[0096] The color feature parameters are the color feature parameters of tea leaves, including the first moments of red, green, blue, hue, saturation, and lightness. The first moments of red, green, blue, hue, saturation, and lightness are calculated by the following formula:

[0097]

[0098] Among them, E i It is the first moment of red, green, blue, hue, saturation, or brightness, where N represents the total number of pixels in the image, and P... ij This represents the j-th image pixel value of the i-th color channel;

[0099] The texture feature parameters include the contrast, second moment of the angular matrix, correlation, entropy, and inverse difference matrix of the tea leaves;

[0100] (3) Figure 11The random forest model built into the data preprocessing module was used to rank the importance of morphological, color, and texture features. The random forest model had 200 decision trees, a minimum of 10 leaves, and a threshold of 0.5. The important features of the withering process were selected as rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrix.

[0101] The tea moisture content monitoring model is a genetic algorithm-optimized backpropagation neural network model (GA-BP model). The specific construction process is as follows: (1) The data preprocessing module calls the historical data in the data storage module. The historical data is the existing tea photos of the withering process and the corresponding tea moisture content. Then, the existing tea photos of the withering process are processed to obtain the rectangle, compactness, greenness, saturation, contrast, correlation and inverse difference matrix. Then, the obtained rectangle, compactness, greenness, saturation, contrast, correlation and inverse difference matrix and the corresponding tea moisture content are normalized. Finally, the normalized data is divided into training set and validation set according to a ratio of 4:1. (2) The microcontroller main control module transmits the above training set and validation set to the host computer through the WIFI communication module as the input data of the GA-BP model. (3) In the host computer, the genetic algorithm is used to optimize the weights and biases of the single hidden layer BP neural network with 7-10-1 structure to obtain the optimal results. The optimized BP neural network is constructed with the following parameters: the number of iterations of the genetic algorithm is 200, the population size is 20, the selection function is normGeomSelect, the mutation function is nonUnifMutation, the crossover probability is 0.5, and the mutation probability is 0.05. The number of neurons in the hidden layer is 10, the activation function of the hidden layer is tansig, the activation function of the output layer is tansig, the training function is Trainlm, the learning rate is 0.01, and the number of iterations is 1000. (4) The optimized BP neural network is trained and validated using the training set and the validation set. The hyperparameters of the model are adjusted according to each index to obtain the GA-BP model. The normalized rectangleness, compactness, greenness, saturation, contrast, correlation and inverse difference matrix are used as input data, and the normalized tea moisture content is used as output data. The performance of the model is evaluated using the coefficient of determination, mean square error and mean absolute error. The calculation formula of each index is:

[0102]

[0103] (5) The host computer transmits the obtained GA-BP model to the microcontroller main control module, and it is loaded into the data monitoring module as the tea moisture content monitoring model.

[0104] The operation process of the intelligent withering machine is as follows: The movable cover 21 is opened, and fresh tea leaves are conveyed to the withering area 7 via the upper leaf conveying mechanism 19. Then, the adjustable tilt conveying mechanism 12 tilts the stainless steel mesh 6 by 5°, allowing the tea leaves to slide within the mesh. Simultaneously, the vibrating motor 8 of the vibrating leaf-turning mechanism begins to vibrate, ensuring the tea leaves are evenly spread on the stainless steel mesh 6. Once the tea leaves are evenly spread on the mesh 6, the vibrating motor 8 stops vibrating, and the adjustable tilt conveying mechanism 12 returns the stainless steel mesh 6 to a horizontal position. Next, the movable cover 21 is closed, the rapid moisture detector 2 obtains the initial moisture content of the tea leaves, and the weight sensor 15 obtains the initial weight of the tea leaves. The withering process begins by activating the yellow light tube 14, heat pump unit 1, hot air unit, and online moisture content detection device. Hot air from the heat pump unit is circulated by fan 16 from below the stainless steel mesh 6, and hot air from the hot air unit is circulated from above the stainless steel mesh 6, withering the lower and upper parts of the tea leaves respectively. The yellow light tubes 14 on the upper and lower layers of the chassis wither the upper and lower parts of the tea leaves from above and below the stainless steel mesh 6, respectively. Simultaneously, the weight sensor 15 acquires the real-time weight of the tea leaves, and the camera captures a photo of the tea leaves inside the stainless steel mesh 6 every 20 minutes. All data is stored in real-time in the data storage module. During the withering process, based on... The system outputs real-time tea leaf photos and the tea moisture content monitoring model, showing the tea moisture content of the corresponding area in the photos. Based on the real-time tea weight, the moisture content monitoring unit calculates the real-time overall average moisture content of the tea. By comparing the tea moisture content of the corresponding area in the photos with the overall average moisture content, the system determines the degree of difference in tea moisture content between the areas captured by different cameras 4, and controls the difference between the tea moisture content of the corresponding area in the photos and the overall average moisture content calculated based on the weight measured by the weight sensor. If the tea moisture content is uneven, the adjustable tilt conveyor mechanism 12 is first activated to tilt the stainless steel mesh 6, and then the excitation motor 8 is activated. Vibrate the stainless steel mesh 6 to agitate and mix the tea leaves. After mixing, turn off the vibrating motor 8 and start the adjustable tilt conveyor mechanism 12 to return the stainless steel mesh 6 to a horizontal position. When the overall average moisture content of the tea leaves in the withering area 7 reaches the set target moisture content, turn off the yellow light tube 14, heat pump unit 1, hot air unit, and online moisture content detection device. Then, start the adjustable tilt conveyor mechanism 12 to tilt the stainless steel mesh 6, and then turn on the vibrating motor 8 to vibrate the stainless steel mesh 6, causing the tea leaves to flow from left to right and exit from the outlet on the right side of the stainless steel mesh 6. When all the tea leaves on the stainless steel mesh 6 have been unloaded from the outlet, the withering of a new batch of tea leaves begins.

[0105] The target moisture content was set at 58%.

[0106] The calculation method for the real-time overall average moisture content of the tea is as follows: the moisture content monitoring unit calculates the current overall average moisture content Q2 according to the following formula based on the stored initial weight W1, initial moisture content Q1 and current real-time tea weight W2, and stores it in the data storage module.

[0107]

[0108] Example 2

[0109] A digital withering method for the Lichuan Red Intelligent Withering Machine according to Embodiment 1 includes the following steps:

[0110] Step 1: Start the Lichuan Red Intelligent Withering Machine, open the movable cover 21, and the tea leaves are conveyed to the stainless steel mesh 6 through the upper leaf conveying structure 19. Then, start the adjustable tilt conveying mechanism 12 to tilt the stainless steel mesh 6 by 5° and then stop. Then, start the vibrating motor 8 to vibrate the stainless steel mesh 6, so that the tea leaves slide on the stainless steel mesh 6 until the tea leaves are evenly spread on the stainless steel mesh 6. Afterward, start the adjustable tilt conveying mechanism 12 to restore the stainless steel mesh 6 to the horizontal position. The feeding is finished, and the movable cover 21 is closed.

[0111] Step 2: The microcontroller main control module sends an inquiry message to the weight sensor 15. After receiving the inquiry message, the weight sensor 15 measures the initial weight of the tea leaves and then sends the initial weight of the tea leaves to the microcontroller main control module. Similarly, the rapid moisture detector 2 measures the initial moisture content of the tea leaves, and the temperature and humidity sensor 28 measures the temperature and humidity within the withering area 7 (i.e., the temperature and humidity approximately 10cm above the stainless steel mesh 6) and then sends the initial moisture content of the tea leaves and the temperature and humidity within the withering area 7 to the microcontroller main control module. The microcontroller main control module communicates via RS232... The -1 interface acquires the initial moisture content Q1 of the tea leaves collected by the rapid moisture detector 2, and acquires the initial weight W1 of the tea leaves collected by the weight sensor and the temperature and humidity data collected by the temperature and humidity sensor through the RS485-1 interface. Then, the microcontroller main control module sends the acquired data to the WiFi communication module and the touch screen through the UART-4 and UART-3 interfaces. The touch screen displays the data, the WiFi communication module wirelessly transmits the data to the host computer, and the microcontroller main control module sends the acquired data to the data storage module for storage through the IIC-1 interface.

[0112] Step 3: Input the target moisture content (58%), upper temperature limit (25℃), lower temperature limit (33℃), upper humidity limit (80%), lower humidity limit (65%), and process program on the touch screen, and click "Start Withering". The withering process begins, activating the yellow light tube 14, heat pump unit 1, hot air unit, online moisture content detection device, exhaust fan 11, and circulating fan 16. The hot air output from the heat pump unit is circulated by the circulating fan 16 from below the stainless steel mesh 6, and the hot air output from the hot air unit is circulated from above the stainless steel mesh 6 to wither the lower and upper parts of the tea leaves respectively. The yellow light tubes 14 on the upper and lower layers of the chassis wither the upper and lower parts of the tea leaves from above and below the stainless steel mesh 6 respectively. Simultaneously, the weight sensor 15 acquires the real-time weight of the tea leaves, and the camera 4 acquires a photo of the tea leaves inside the stainless steel mesh 6 every 20 minutes. All data is stored in real-time in the data storage module.

[0113] When the temperature and humidity sensor detects a current temperature higher than the input upper temperature limit, the microcontroller main control module sends signals to activate the alarm, deactivate the hot air blower, and deactivate the heat pump unit. This activates the alarm, deactivates the hot air blower and heat pump unit, while the exhaust fan 11 and circulating fan 16 continue operating to lower the temperature in the withering area 7 and prevent excessive moisture loss in the tea leaves. The alarm is deactivated and the hot air blower and heat pump unit are activated once the current temperature is no higher than the upper temperature limit. Similarly, when the temperature and humidity sensor detects a current temperature lower than the input lower temperature limit, the microcontroller main control module sends signals to activate the alarm, increase the power of the hot air blower, and increase the power of the heat pump unit. This activates the alarm and increases the temperature of the hot air delivered by the hot air blower and heat pump unit to raise the temperature of the withering area. Domain 7 temperature prevents tea from becoming bitter or brittle due to excessively low temperatures. The power of the heat pump unit and the hot air blower returns to their original power levels once the current temperature is no lower than the input lower limit. When the current humidity, as monitored by the temperature and humidity sensor, is higher than the input upper limit, the microcontroller main control module sends a signal to increase the power of the exhaust fan 11, improving its dehumidification efficiency. The power of the exhaust fan 11 returns to its original power once the current humidity is no higher than the upper limit. Similarly, when the current humidity, as monitored by the temperature and humidity sensor, is lower than the input lower limit, the microcontroller main control module sends a signal to decrease the power of the exhaust fan 11, reducing its dehumidification efficiency. The power of the exhaust fan 11 returns to its original power once the current humidity is no lower than the lower limit.

[0114] Step 4: Based on the stored initial weight W1, initial moisture content Q1, and current real-time weight W2, the moisture content monitoring unit calculates the current overall average moisture content Q2 according to the following formula and stores it in the data storage module.

[0115]

[0116] Step 5: The data preprocessing module processes the tea photos to obtain important feature parameters of the withering process. The specific process is as follows: (1) Process the tea photos into color images and grayscale images; (2) Extract features from the color images and grayscale images to obtain the morphological feature parameters, color feature parameters, and texture feature parameters of the tea. Among them, the colorhist function is used to calculate the color image obtained after processing the tea photos to obtain color feature parameters, and the graycomatrix function is used to calculate the grayscale image obtained after processing the tea photos to obtain texture feature parameters. Median filtering, followed by the Otsu method, then the erosion-dilation method, and finally the regionprops function are used to calculate the grayscale image obtained after processing the tea photos to obtain morphological feature parameters. The morphological feature parameters include the major axis, minor axis, rectangularity, roundness, compactness, diagonal length, and thinness of the tea. The expressions for the five morphological features of rectangularity, roundness, compactness, diagonal length, and thinness are as follows:

[0117]

[0118]

[0119] J = D / L

[0120]

[0121] Where R is rectangularity, E is circularity, J is compactness, Len is linearity, T is slenderness, A is the area of ​​the tea outline, C is the perimeter of the tea outline, D is the diameter of the circle with the same area as the tea outline, L is the length (major axis) of the circumscribed rectangle of the tea outline, and S is the width (minor axis) of the circumscribed rectangle of the tea outline.

[0122] The color characteristic parameters are the color characteristic parameters of tea leaves, including the first moments of red, green, blue, hue, saturation, and lightness. The first moments of red, green, blue, hue, saturation, or lightness are obtained by the following formula:

[0123]

[0124] Among them, E i It is the first moment of red, green, blue, hue, saturation, or brightness, where N represents the total number of pixels in the image, and P... ij This represents the j-th image pixel value of the i-th color channel;

[0125] The texture feature parameters include the contrast, second moment of the angular axis, correlation, entropy, and inverse difference matrix of the tea leaves; (3) such as Figure 11The random forest model built into the data preprocessing module was used to rank the importance of morphological, color, and texture features. The random forest model had 200 decision trees, a minimum of 10 leaves, and a threshold of 0.5. The important features of the withering process were selected as rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrix.

[0126] Step 6: The microcontroller main control module stores the rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrices obtained in Step 5 into the data storage module; simultaneously, it inputs the rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrices as input data into the tea moisture content monitoring model in the data monitoring module. The tea moisture content monitoring model outputs the tea moisture content Q3, and compares Q3 with Q2 according to the following formula:

[0127] Threshold = |Moisture content of tea leaves in the area corresponding to the photo Q3 - Overall average moisture content Q2| / Overall average moisture content Q2 × %;

[0128] If the thresholds calculated based on the tea moisture content Q3 and the overall average moisture content Q2 of the corresponding areas captured by the four cameras are both ≤5%, it indicates that the tea on the stainless steel mesh 6 is withered evenly, and the withering process continues; if the thresholds calculated based on the tea moisture content Q3 and the overall average moisture content Q2 of the corresponding areas captured by one or more cameras are >5%, it indicates that the tea on the stainless steel mesh 6 is not withered evenly, then the adjustable tilt conveyor mechanism 12 is activated to tilt the stainless steel mesh 6 by 5°, and then the excitation motor 8 is activated to excite the stainless steel mesh 6, so that the tea on it is turned over and mixed evenly.

[0129] If the overall average moisture content Q2 reaches the set target moisture content of 58%, then the heat pump unit 1, the hot air unit, and the yellow light tube 14 are turned off, and the air supply heating and yellow light irradiation are stopped. Then, the movable cover 21 is opened, and the adjustable tilt conveyor mechanism 12 is started, so that the right side of the withering frame is lower than the left side. At the same time, the excitation motor 8 is started, which excites the stainless steel mesh 6, causing the tea leaves to flow out from the outlet on the right side of the withering operation area 7. When all the tea leaves on the stainless steel mesh 6 have been unloaded from the outlet, the withering of a new batch of tea leaves is restarted.

[0130] In this invention, the tea moisture content Q3 output by the data monitoring module was compared with the tea moisture content corresponding to historical tea photos (i.e., the output value of the data monitoring module was compared with the actual measured value). The results are as follows:

[0131] Table 1. Comparison of tea moisture content (Q3) of different grades of tea raw materials with the tea moisture content corresponding to historical tea photos.

[0132]

[0133]

[0134] As shown in Table 1, there are differences between the moisture content of each area within the withering area 7 and the overall moisture content. This indicates that inconsistencies exist between the fresh leaf conditions and the moisture content of the tea leaves in different areas, which can easily lead to uneven withering during the process. The areas captured by different cameras only represent the moisture content of that specific area and do not reflect the overall moisture content. However, the moisture content of the tea leaves in each area within the withering area 7 is not significantly different from the measured overall average moisture content, indicating that this invention can effectively monitor the moisture content of the tea leaves. Furthermore, this demonstrates that this invention can more quickly and accurately reflect the overall degree of withering of the tea leaves.

[0135] Table 2. Sensory evaluation results of Lichuan black tea obtained using one bud and two leaves as raw material and different withering methods.

[0136]

[0137] As shown in Table 2, when using one bud and two leaves as raw material, the overall score of the withering machine's main control system is slightly higher than that of human judgment. This is mainly reflected in the aroma and taste. Tea with proper withering, judged by the withering machine's main control system, has a stronger sweet and fruity aroma and a longer aftertaste.

[0138] Table 3. Sensory evaluation results of Lichuan black tea obtained using machine-harvested leaves as raw material and different withering methods.

[0139]

[0140]

[0141] As shown in Table 3, when using machine-picked leaves as raw materials, the overall score of the withering machine's main control system is slightly higher than that of human judgment. The dry tea has a more oily and lustrous color, a sweeter aroma, and less bitterness.

[0142] This invention, through the coordinated use of various mechanisms, solves problems such as poor uniformity and difficulty in automatic leaf turning and loading / unloading during the tea withering process, thereby improving the quality of Lichuan Red tea and giving it high efficiency, high standards, and high intelligence. This is of great significance to the development of Lichuan Red tea.

[0143] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A Lichuan Red intelligent withering machine, comprising a withering trough, a heat pump unit, a hot air unit, an online moisture content detection device, and a main control system; characterized in that, The hot air unit includes a hot air blower, a main hot air duct, and a branch hot air duct; the heat pump unit is located on the left side of the withering tank, and an air duct connecting the two is formed in the lower part between the heat pump unit and the withering tank. The withering tank includes a casing; the casing is divided into three layers from top to bottom, the upper layer is the withering operation area, the upper half of the middle layer is a vibrating leaf turning mechanism, the lower half is an adjustable tilt conveyor mechanism, and the top of the lower layer is equipped with an array of yellow lamps; Inside the chassis, the withering area includes a conveyor trough. The four side walls of the withering area are equipped with an array of yellow fluorescent tubes. The bottom is a stainless steel mesh fixed to the chassis, with two temperature and humidity sensors fixedly installed on the mesh. The space above the mesh is the conveyor trough. The main hot air duct and branch hot air ducts are located in the middle of the upper layer of the chassis. A camera is installed at each of the four corners of the top of the upper layer (the four corners inside the chassis). In the lower layer, the ends of the yellow fluorescent tubes are fixed to the chassis by brackets, and the tubes are linearly and equidistantly distributed. The top surface of the chassis has a fixed cover and a movable cover. The stainless steel mesh is used to spread the tea leaves, and the yellow fluorescent tubes installed on the side walls above the mesh and in the lower layer of the chassis provide yellow light irradiation. The hot air unit and heat pump unit wither the tea leaves from the upper and lower parts, respectively. The online moisture content detection device includes a weight sensor, a rapid moisture analyzer, and cameras. The weight sensor is installed at the bottom of the casing of the entire withering tank. Four cameras are installed at the four corners of the top of the casing of the withering tank and fixed to the casing by brackets, arranged diagonally. The rapid moisture analyzer is used to obtain the initial moisture content of the tea leaves. The weight sensor is used to obtain the real-time weight of the tea leaves during the withering process, and the real-time overall average moisture content of the tea leaves during the withering process is obtained based on the real-time weight of the tea leaves and the initial moisture content. The main control system includes a control box, which is used for digital control of the withering process and determines whether the real-time overall average moisture content of the tea leaves has reached the set target moisture content during the withering process. A touchscreen is mounted on the surface of the control box, which houses the main control board. The touchscreen displays the target moisture content, upper temperature limit, and process program. The main control board includes a power supply module, a communication module, a WiFi communication module, a microcontroller main control module, an I / O control module, and a data storage module. The microcontroller main control module includes a moisture content monitoring unit. This unit comprises a data preprocessing module and a data monitoring module. The data monitoring module loads a tea moisture content monitoring model. The moisture content monitoring unit is used to determine the degree of difference in tea moisture content in areas corresponding to photos taken by different cameras, and to determine the degree of difference between the tea moisture content in the areas corresponding to the photos and the overall average moisture content calculated based on weight measurements from a weight sensor.

2. The Lichuan Red Intelligent Withering Machine according to claim 1, characterized in that, The heat pump unit includes an evaporator, a gas-liquid separator, a compressor, a condenser, a circulating fan I, and an exhaust fan I. A circulating fan II is installed on one side of the withering tank in the air duct, and the exhaust fan II is installed in the upper right corner directly in front of the withering tank. The heat pump unit absorbs heat from the surrounding environment and outputs hot air. The hot air enters the air duct and is sent to the lower layer of the casing by the circulating fan II. Then the hot air rises and passes through the stainless steel mesh to wither the lower part of the tea leaves inside. The heat pump unit uses the heat absorbed from the surrounding environment to wither the tea leaves on the stainless steel mesh.

3. The Lichuan Red Intelligent Withering Machine according to claim 1, characterized in that, The hot air blower is mounted on a bracket between the chassis and the heat pump unit. The outlet of the hot air blower is connected to the inlet of the main hot air duct. The main hot air duct is connected with hot air branch pipes at equal intervals. Small holes with a diameter of 1 cm are evenly opened on the hot air branch pipes. The main hot air duct is mounted above the stainless steel mesh and fixed to the chassis by the bracket. The hot air unit is used to deliver hot air to wither the tea leaves from above the tea leaves on the stainless steel mesh.

4. The Lichuan Red Intelligent Withering Machine according to claim 1, characterized in that, The vibrating leaf-turning mechanism includes an excitation motor and vibration springs. The four vibration springs are located below the four corners of the stainless steel mesh, and the excitation motor is located in the middle between every two springs. The vibrating leaf-turning mechanism makes the tea leaves in the stainless steel mesh more evenly dispersed, thus making the withering more uniform.

5. The Lichuan Red Intelligent Withering Machine according to claim 1, characterized in that, The adjustable tilt conveyor mechanism includes bushing I, right-angle lifters, a drive shaft, a geared motor, rotating bearing I, a slider, a guide rail, a support, bushing II, and rotating bearing II. The adjustable tilt conveyor mechanism is located below the vibrating leaf-turning mechanism. On the left side of the adjustable tilt conveyor mechanism are two right-angle lifters, and on the right side are two supports. The two right-angle lifters and two supports are respectively located below a vibrating spring. The bottom end of the right-angle lifter is installed on the corner of the bottom side of the frame of the middle layer of the machine housing, and the top end is the lifting end connected to bushing I. One end of rotating bearing I is connected to bushing I via a rotating shaft, and the other end is connected to the slider. The slider is connected to the guide rail, which is installed below the vibrating spring. The bottom end of the support is installed on the machine housing. On the other side of the bottom corner of the middle frame, a bushing II is installed at the top of the support. One end of the rotating bearing II is connected to the bushing II via a connecting shaft, and the other end is installed below the vibration spring. The geared motor is installed in the middle between the two right-angle lifters and is connected to the right-angle lifters via a transmission shaft. Different teas have different flow characteristics. The upper leaf conveying mechanism conveys the tea to the stainless steel mesh, and then the movable cover is closed. The adjustable tilting angle conveying mechanism tilts the stainless steel mesh by 5°, and at the same time, the vibrating leaf turning mechanism is activated to vibrate the stainless steel mesh, so that the tea is evenly spread on the stainless steel mesh.

6. The Lichuan Red Intelligent Withering Machine according to claim 1, characterized in that, The working process of the moisture content monitoring unit is as follows: During the withering process, the camera takes pictures of the tea leaves in the withering frame every 20 minutes and transmits the pictures to the data preprocessing module. The data preprocessing module preprocesses the tea leaf pictures to obtain color and grayscale images. Then, it extracts features from the color and grayscale images to obtain the morphological, color, and texture parameters of the tea leaves. The data monitoring module loads the tea leaf moisture content monitoring model. The tea leaf moisture content monitoring model outputs the real-time moisture content of the tea leaves based on the real-time morphological, color, and texture parameters of the tea leaves. This allows the system to determine the degree of difference in the moisture content of the tea leaves in the areas corresponding to the pictures taken by different cameras, as well as the degree of difference between the moisture content of the tea leaves in the areas corresponding to the pictures and the overall average moisture content of the tea leaves calculated based on the weight measured by the weight sensor. According to the formula: Threshold = |Moisture content of tea leaves in the area corresponding to the photo - Overall average moisture content| / Overall average moisture content × 100%, if the threshold calculated based on the moisture content of tea leaves in the area corresponding to the photos taken by the four cameras and the overall average moisture content are both ≤ 5%, it indicates that the tea leaves on the stainless steel mesh are withered evenly, and the withering process continues. If the threshold calculated based on the tea moisture content and overall average moisture content of the corresponding area captured by one or more cameras is greater than 5%, it indicates that the tea is not withering evenly on the stainless steel mesh. In this case, the adjustable tilt conveyor mechanism is activated to tilt the stainless steel mesh by 5°, and then the vibrating motor is activated to vibrate the stainless steel mesh, causing the tea on it to turn over and mix evenly. If the overall average moisture content reaches the set target moisture content, the heat pump unit, hot air unit, and yellow light tube are turned off, and the air supply heating and yellow light irradiation are stopped. Then, the movable cover is opened, and the adjustable tilt conveyor mechanism is activated again, so that the right side of the withering frame is lower than the left side. At the same time, the vibrating motor is activated to vibrate the stainless steel mesh, causing the tea to flow out from the outlet on the right side of the withering operation area. When all the tea on the stainless steel mesh has been unloaded from the outlet, the withering of a new batch of tea begins again.

7. The Lichuan Red Intelligent Withering Machine according to claim 6, characterized in that, The photos of tea leaves on a stainless steel mesh taken by the camera are processed by the data preprocessing module to obtain the tea leaf morphological feature parameters, color feature parameters, and texture feature parameters: rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrix. The rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrix are then input into the GA-BP neural network model to output the tea leaf moisture content. Based on the output tea leaf moisture content, the degree of tea leaf withering is judged, thereby ensuring that each withering process can produce tea leaves with a similar degree of withering, thus standardizing the tea leaf withering process and avoiding the bias caused by subjective judgment of the degree of tea leaf withering by humans, which would result in different degrees of withering for each batch of tea leaves. The specific process of the data preprocessing module to obtain the rectangularity, compactness, greenness, saturation, contrast, correlation and inverse difference matrix is ​​as follows: (1) Process the tea photos into color images and grayscale images; (2) Extract features from the color images and grayscale images to obtain the morphological feature parameters, color feature parameters and texture feature parameters of the tea, wherein the morphological feature parameters include the major axis, minor axis, rectangularity, roundness, compactness, diagonal length and thinness of the tea, the color feature parameters are the color feature parameters of the tea, including the first moment of red, green, blue, hue, saturation and brightness, and the texture feature parameters include the contrast, second moment of angle, correlation, entropy and inverse difference matrix of the tea; (3) Use the random forest model built into the data preprocessing module to sort the importance of the morphological feature parameters, color feature parameters and texture feature parameters, and select the important feature parameters of the withering process as rectangularity, compactness, greenness, saturation, contrast, correlation and inverse difference matrix.

8. The Lichuan Red Intelligent Withering Machine according to claim 7, characterized in that, The tea moisture content monitoring model is a GA-BP neural network model. The specific construction process is as follows: (1) The data preprocessing module calls the historical data in the data storage module. The historical data is the existing tea photos of the withering process and the corresponding tea moisture content. Then, the existing tea photos of the withering process are processed to obtain the rectangularity, compactness, greenness, saturation, contrast, correlation and inverse difference matrix. Then, the obtained rectangularity, compactness, greenness, saturation, contrast, correlation and inverse difference matrix and the corresponding tea moisture content are normalized. Finally, the normalized data are processed according to 4:

1. The ratio is divided into training set and validation set; (2) The microcontroller main control module transmits the above training set and validation set to the host computer through the WIFI communication module as input data of GA-BP model; (3) In the host computer, the genetic algorithm is used to optimize the weights and biases of the single hidden layer BP neural network with 7-10-1 structure to obtain the optimized BP neural network. The number of iterations of the genetic algorithm is 200, the population size is 20, the selection function is normGeomSelect, the mutation function is nonUnifMutation, the crossover probability is 0.5, and the mutation probability is 0.

05. The BP neural network is constructed with the following parameters: the number of neurons in the hidden layer is 10, the activation function of the hidden layer is t The output layer activation function is tansig, the training function is Trainlm, the learning rate is 0.01, and the number of iterations is 1000; (4) The optimized BP neural network is trained and validated using the training set and validation set, and the hyperparameters of the model are adjusted according to each index to obtain the GA-BP model; among them, the normalized rectangularity, compactness, greenness, saturation, contrast, correlation and inverse difference matrix are used as input data, and the normalized tea moisture content is used as output data. The performance of the model is evaluated using the coefficient of determination, mean square error and mean absolute error index; (5) The host computer transmits the obtained GA-BP model to the microcontroller main control module, and it is loaded into the data monitoring module as the tea moisture content monitoring model.

9. A digital withering method for the Lichuan Red Intelligent Withering Machine according to any one of claims 1-8, comprising the following steps: Step 1: Start the Lichuan Red Intelligent Withering Machine, open the movable cover, and the tea leaves are conveyed to the stainless steel mesh through the upper leaf conveying structure. Then, start the adjustable tilt conveying mechanism to tilt the stainless steel mesh by 5° and then stop. Then, start the excitation motor to excite the stainless steel mesh, so that the tea leaves slide on the stainless steel mesh until the tea leaves are evenly spread on the stainless steel mesh. Then, the adjustable tilt conveyor mechanism is activated to restore the stainless steel mesh to a horizontal position, the feeding is completed, and the movable cover is closed. Step 2: The microcontroller main control module sends an inquiry message to the weight sensor. After receiving the inquiry message, the weight sensor measures the initial weight of the tea leaves and then sends the initial weight of the tea leaves to the microcontroller main control module. Similarly, the rapid moisture detector measures the initial moisture content of the tea leaves, and the temperature and humidity sensor measures the temperature and humidity in the withering area and sends the initial moisture content of the tea leaves and the temperature and humidity in the withering area to the microcontroller main control module. The withering area is the temperature and humidity about 10cm above the stainless steel mesh. Specifically, the microcontroller main control module obtains the initial moisture content Q1 of the tea leaves collected by the rapid moisture detector through the RS232-1 interface of the communication module, obtains the initial weight W1 of the tea leaves collected by the weight sensor and the temperature and humidity data collected by the temperature and humidity sensor through the RS485-1 interface. Then, the microcontroller main control module sends the acquired data to the WiFi communication module and the touch screen through the UART-4 and UART-3 interfaces. The touch screen displays the data, the WiFi communication module wirelessly transmits the data to the host computer, and the microcontroller main control module sends the acquired data to the data storage module for storage through the IIC-1 interface. Step 3: Input the target moisture content, upper temperature limit, lower temperature limit, upper humidity limit, lower humidity limit, and process program on the touch screen, and click "Start Withering". The withering process begins, activating the yellow light tubes, heat pump unit, hot air unit, online moisture content detection device, exhaust fan, and circulating fan. The hot air output from the heat pump unit is circulated from below the stainless steel mesh by the circulating fan, and the hot air output from the hot air unit is circulated from above the stainless steel mesh to wither the lower and upper parts of the tea leaves, respectively. The yellow light tubes on the upper and lower layers of the chassis wither the upper and lower parts of the tea leaves from above and below the stainless steel mesh, respectively. Simultaneously, the weight sensor acquires the real-time weight of the tea leaves, and the camera acquires a photo of the tea leaves inside the stainless steel mesh every set time. All data is stored in the data storage module in real time. When the temperature and humidity sensor detects a current temperature higher than the input upper temperature limit, the microcontroller main control module sends signals to activate the alarm, deactivate the hot air blower, and deactivate the heat pump unit. This activates the alarm, deactivates the hot air blower and heat pump unit, while the exhaust fan and circulating fan continue operating to lower the temperature in the withering area and prevent excessive moisture loss from the tea leaves. This continues until the current temperature is no higher than the upper temperature limit, at which point the alarm is deactivated and the hot air blower and heat pump unit are activated. Similarly, when the temperature and humidity sensor detects a current temperature lower than the input lower temperature limit, the microcontroller main control module sends signals to activate the alarm, increase the power of the hot air blower, and increase the power of the heat pump unit. This activates the alarm and increases the temperature of the hot air delivered by the hot air blower and heat pump unit to raise the temperature of the tea leaves during the withering process. The system adjusts the temperature of the operating area to prevent the tea leaves from becoming bitter or brittle due to excessively low temperatures. Once the current temperature is no lower than the input lower limit, the power of the heat pump unit and the hot air blower return to their original power. When the current humidity, as monitored by the temperature and humidity sensor, is higher than the input upper limit, the microcontroller main control module sends a signal to increase the power of the exhaust fan, improving its dehumidification efficiency. Once the current humidity is no higher than the upper limit, the exhaust fan power returns to its original power. Similarly, when the current humidity, as monitored by the temperature and humidity sensor, is lower than the input lower limit, the microcontroller main control module sends a signal to decrease the power of the exhaust fan, reducing its dehumidification efficiency. Once the current humidity is no lower than the lower limit, the exhaust fan power returns to its original power. Step 4: Based on the stored initial weight W1, initial moisture content Q1, and real-time weight W2, the moisture content monitoring unit calculates the overall average moisture content Q2 according to the following formula and stores it in the data storage module. Step 5: The data preprocessing module processes the tea photos to obtain important feature parameters of the withering process. The specific process is as follows: (1) Process the tea photos into color images and grayscale images; (2) Extract features from the color images and grayscale images to obtain the morphological feature parameters, color feature parameters, and texture feature parameters of the tea. The morphological feature parameters include the major axis, minor axis, rectangularity, roundness, compactness, diagonal length, and thinness of the tea. The color feature parameters are the color feature parameters of the tea, including the first moment of red, green, blue, hue, saturation, and brightness. The texture feature parameters include the contrast, second moment of angle, correlation, entropy, and inverse difference matrix of the tea; (3) Use the random forest model built into the data preprocessing module to sort the morphological feature parameters, color feature parameters, and texture feature parameters by importance, and select the important feature parameters of the withering process as rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrix. Step 6: The microcontroller main control module stores the rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrices obtained in Step 5 into the data storage module; simultaneously, it inputs the rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrices as input data into the tea moisture content monitoring model in the data monitoring module. The tea moisture content monitoring model outputs the tea moisture content Q3 based on the area corresponding to the photos taken by the four cameras, and compares Q3 with Q2 according to the following formula: Threshold = |Moisture content of tea leaves in the area corresponding to the photo Q3 - Overall average moisture content Q2| / Overall average moisture content Q2 × 100%; If the thresholds calculated based on the tea moisture content Q3 and the overall average moisture content Q2 of the corresponding areas captured by the four cameras are both ≤5%, it indicates that the tea is withered evenly on the stainless steel mesh, and the withering process continues. If the threshold calculated based on the tea moisture content Q3 and the overall average moisture content Q2 of the corresponding area captured by one or more cameras is greater than 5%, it indicates that the tea is not withering evenly on the stainless steel mesh. In this case, the adjustable tilt conveyor mechanism is activated to tilt the stainless steel mesh by 5°, and then the excitation motor is activated to vibrate the stainless steel mesh, causing the tea on it to turn over and mix evenly. If the overall average moisture content Q2 reaches the set target moisture content, the heat pump unit, hot air unit, and yellow light tube are turned off, and the air supply heating and yellow light irradiation are stopped. Then, the movable cover is opened, and the adjustable tilt conveyor mechanism is activated again, so that the right side of the withering frame is lower than the left side. At the same time, the excitation motor is activated to vibrate the stainless steel mesh, causing the tea to flow out from the outlet on the right side of the withering operation area. When all the tea on the stainless steel mesh has been unloaded from the outlet, the withering of a new batch of tea begins again.

Images