Intelligent Lichuan red withering machine and digital withering method thereof

Abstract

The invention belongs to the technical field of tea processing, and particularly relates to an intelligent Lichuan red withering machine and a digital withering method thereof, and the intelligent Lichuan red withering machine comprises a withering groove body, a heat pump unit, a hot air unit, a vibration leaf turning mechanism, an adjustable inclination angle conveying mechanism, a water content online detection device and a main control system. Withering the upper and lower parts of the tea leaves by hot air and yellow light; the water content online detection device obtains the real-time total average water content of the tea leaves; the main control system is used for judging the difference degree of the water contents of the tea leaves in the areas corresponding to the pictures shot by the different cameras and judging the difference degree of the water contents of the tea leaves in the areas corresponding to the pictures and the total average water content calculated based on the weight measured by the weight sensor. According to the method, the uniformity of the tea withering process is judged by comparing the color, texture and shape information of the tea pictures in the withering operation area collected by each camera; the vibration leaf turning mechanism is matched with the inclination-angle-adjustable conveying mechanism, so that the continuous tea withering operation can be realized.

Description

Technical Field

[0001] The invention belongs to the technical field of tea processing, and in particular relates to a Lichuan red tea intelligent withering machine and a digital withering method thereof. Background Art

[0002] Lichuan Red is distributed in Lichuan City, Enshi Prefecture, Hubei Province. Lichuan Red is a fully fermented tea. Its unique fermentation process allows the tea leaves to undergo non-enzymatic auto-oxidation under the action of water and heat, resulting in tight and straight tea leaves, dark color, golden tips, bright red soup color, rich aroma, and sweet taste. It has anti-cancer, antibacterial, antioxidant and other functions, so its sales have been getting better and better in recent years.

[0003] Withering is the first process in the processing of Lichuan Red Tea. It is a process in which the picked fresh leaves are spread out to allow the tea to lose its green smell, reduce its water content, concentrate the cell fluid, change the physical and chemical properties of the protein, and transform the enzyme from a bound state to a free state. The activity of enzymes such as oxidase and hydrolase is enhanced, and macromolecular substances begin to hydrolyze to form simple substances, laying the foundation for the formation of tea flavor substances, aromatic substances, and coloring substances. The withering equipment on the market currently includes trough-type and multi-layer circulating conveyor-type withering machines. Due to the trough-type withering equipment, the degree of withering of fresh leaves is easily uneven along the length, width, and thickness of the trough during the withering process, and the unevenness between multiple parts of the buds and leaves, and manual page turning is required. The multi-layer circulating conveyor-type withering machine has a large equipment space, and the temperature and humidity of the tea withering process are not easy to control. The tea leaves are easily damaged during the continuous circulation and transportation process. In addition, the equipment is large in size and is not suitable for small-scale production. In addition, the judgment of the withering degree in production mainly relies on manual experience, which is subjective to a certain extent. The moisture content is an important indicator for evaluating the withering degree. Studies have shown that the quality of Lichuan red tea processed at a moisture content of about 58% at the end of withering is better. With the aging of the population and the continuous reduction of rural labor force, the problem of population resources has become prominent. In view of the problems of the traditional withering process in the current market, such as high labor intensity, low withering uniformity, and difficulty in turning leaves, it is urgent to develop an intelligent withering trough and withering method to solve the problem of Lichuan black tea processing in my country and improve the economic benefits of tea in my country. Summary of the invention

[0004] In view of the deficiencies in the prior art, the object of the present invention is to provide a Lichuan red intelligent withering machine and a digital withering method.

[0005] To achieve the above object, the present invention adopts the following technical solutions:

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

[0007] The withering trough body includes a chassis, an air duct, a stainless steel net, a withering operation area, and yellow light tubes. The frame 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 inclination conveying mechanism, and the top of the lower layer is provided with an array formed by yellow light tubes. Inside the chassis, the withering operation area includes a conveying trough, the four side walls of the withering operation area are provided with an array formed by yellow light tubes, the bottom end is a stainless steel net fixedly connected to the chassis, two temperature and humidity sensors are fixedly installed on the stainless steel net, the space above the stainless steel net is a conveying trough, a hot air main duct and a hot air branch duct are provided in the middle position of the upper layer, and a camera is provided at each of the four corners at the top of the upper layer, that is, the four corners at the top of the chassis. In the lower layer, the two ends of the yellow light tube are fixed to the chassis by brackets, and the yellow light tubes are linearly and equidistantly distributed. The top surface of the chassis is a fixed cover and a movable cover. Among them, the stainless steel net is used to spread the tea leaves, and the yellow light tubes installed on the side wall above the stainless steel net and the yellow light tubes installed in the lower layer 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 to the increase of amino acid content and the formation of aldehydes in tea, significantly improving the aroma of tea; the temperature and humidity sensors on the stainless steel net are used to measure the temperature and humidity of the upper withering operation area. Appropriate temperature and humidity can be beneficial to the transformation of substances contained in tea, thereby improving the quality of tea.

[0009] The moisture content online detection device includes a weight sensor, a rapid moisture detector, and a camera. The weight sensor is installed under the chassis of the entire withering tank; the four cameras are installed at the four corners of the top of the withering tank chassis, fixed on the frame through brackets, and distributed diagonally; the rapid moisture detector 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 according to the real-time weight of the tea leaves and the initial moisture content.

[0010] 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 during the withering process reaches the set target moisture content; 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 moisture content of tea leaves in areas corresponding to photos taken by different cameras, and to determine the degree of difference between the moisture content of tea leaves in the areas corresponding to the photos and the overall average moisture content calculated based on the weight measured by the weight sensor.

[0011] Preferably, if Fig.12The heat pump unit includes an evaporator, a gas-liquid separator, a compressor, a condenser, a circulating fan I, and a dehumidifying fan I. The heat pump unit is located on the left side of the withering tank body. An air duct connecting the heat pump unit and the withering tank body is formed at the lower part between the heat pump unit and the withering tank body. A circulating fan II is installed on the withering tank side of the air duct, and a dehumidifying fan II is installed in the upper right corner directly in front of the withering tank body. The heat pump unit absorbs heat from the surrounding environment and outputs hot air. The hot air enters the air duct, and the hot air in the air duct is sent to the lower layer of the chassis through 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, which can save energy and reduce the emission of pollutants such as CO2, achieving the effect of energy saving and emission reduction.

[0012] Preferably, the hot air unit comprises a hot air blower, a hot air main pipe, and a hot air branch pipe. The hot air blower is installed 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 hot air main pipe, the hot air main pipe is equidistantly connected with hot air branch pipes, the hot air branch pipes are evenly provided with small holes with a diameter of 1 cm, the hot air main pipe is installed above the stainless steel mesh, and fixed to the chassis through a 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.

[0013] The existing withering troughs are all ventilated and heated from the bottom, which easily causes uneven withering of the upper and lower tea leaves during the withering process. The present invention adopts 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 more uniform transformation of the substances inside the tea leaves.

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

[0015] Preferably, the adjustable inclination conveying mechanism comprises a sleeve I, a right-angle lifter, a transmission shaft, a reduction motor, a rotating bearing I, a slider, a guide rail, a support, a sleeve II, and a rotating bearing II. The adjustable inclination conveying mechanism is located below the vibrating leaf-turning mechanism, and the left side of the adjustable inclination conveying mechanism is two right-angle lifters, and the right side is two supports, and the two right-angle lifters and the two supports are respectively located below a vibration spring; the bottom end of the right-angle lifter is installed on the corner at the bottom end of one side of the frame of the middle layer of the chassis, and the top end is the lifting end connected to the sleeve I, one end of the rotating bearing I is connected to the sleeve I through the rotating shaft, and the other end is connected to the slider, and the slider is connected to the guide rail, and the guide rail is installed below the vibration spring; the bottom end of the support is installed on the corner at the bottom end of the other side of the frame of the middle layer of the chassis, and the top end of the support is installed with a sleeve II, and one end of the rotating bearing II is connected to the sleeve II through the connecting shaft, and the other end is installed below the vibration spring; the reduction motor is installed in the middle between the two right-angle lifters, and is connected to the right-angle lifter through a transmission shaft. Different teas have different flow characteristics. The upper leaf conveying mechanism conveys the tea leaves to the stainless steel net, and then the movable cover is closed. The stainless steel net is tilted 5° through the adjustable inclination conveying mechanism, and the vibrating leaf turning mechanism is started to vibrate the stainless steel net, so that the tea leaves are evenly spread on the stainless steel net.

[0016] Preferably, 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 during the withering process reaches the set target moisture content. A touch screen is installed on the surface of the control box, and a main control board is arranged inside it. The target moisture content, the upper temperature limit and the process program are input on the touch screen. The main control board includes a power supply module, a communication module, a WiFi communication module, a single-chip main control module, an IO control module, and a data storage module. The single-chip main control module includes a moisture content monitoring unit, and the moisture content monitoring unit includes a data preprocessing module and a data monitoring module.

[0017] The output end of the power supply module is electrically connected to the input end of the communication module, WiFi communication module, single-chip main control module, IO control module, and data storage module, and is used to supply power to the communication module, single-chip main control module, IO control module, and data storage module; the power supply module also serves as a 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 data-connected with the weight sensor, the temperature and humidity sensor, the rapid moisture detector, and the single-chip main control module, and is used to communicate with the weight sensor, the temperature and humidity sensor, the rapid moisture detector, and the single-chip main control module;

[0019] The WiFi communication module is connected to the single-chip microcomputer main control module and is used to realize wireless communication with the host computer.

[0020] The output end of the IO control module is connected to the input end of the electromagnetic relay, wherein the output end of the first electromagnetic relay is connected to the alarm, the output end of the second electromagnetic relay is connected to the hot air blower, the output end of the third electromagnetic relay is connected to the heat pump unit, the output end of the fourth electromagnetic relay is connected to the vibration leaf-turning mechanism, the output end of the fifth electromagnetic relay is connected to the adjustable inclination conveying mechanism, the output end of the sixth electromagnetic relay is connected to the circulation fan, and the output end of the seventh electromagnetic relay is connected to the dehumidification fan, which is used to control the start and stop of the hot air blower, the circulation fan, the dehumidification fan, the heat pump unit, the alarm, the vibration leaf-turning mechanism, and the adjustable inclination conveying mechanism.

[0021] The communication module has RS485-1, RS232-1, and UART interfaces. RS485-1, RS232-1, and UART interfaces provide data exchange channels for external devices to ensure the normal operation of the system. Among them, the RS485-1 interface is responsible for communicating with the weighing sensor and the temperature and humidity sensor, and the RS232-1 interface is responsible for receiving data transmitted by the rapid moisture detector. The UART interface includes UART-1, UART-2, UART-3, and UART-4. The UART-1 is used to transmit the RS485-1 signal after level conversion into the single-chip main control module, and the UART-2 is used to transmit the RS232-1 signal after level conversion into the single-chip main control module. The UART-4 and UART-3 are directly derived from the single-chip main control module and are respectively used to communicate with the WiFi communication module and the touch screen.

[0022] Preferably, the single-chip microcomputer main control module is the core of the entire main control system, and is responsible for the collection, processing, transmission of sensor data and the control of the withering process.

[0023] Preferably, the data storage module is connected to the single-chip main control module and is used to store various data. The data storage module is connected to the single-chip main control module through the IIC-1 interface. The single-chip main control module receives various data and stores them in the data storage module, including initial moisture content, real-time weight, and target moisture content data. The single-chip main control module transmits the current time and the current weight, current moisture content, current temperature and humidity at the current time, and characteristic parameter data obtained by preprocessing the camera photos to the data storage module according to the set time interval during the withering process. The initial moisture content, initial weight, and target moisture content are displayed on the touch screen, and various data of the withering process can be output on the host computer.

[0024] Preferably, the working process of the moisture content monitoring unit is as follows: during the withering process, a camera photographs the tea leaves in the withering frame every 20 minutes, and transmits the photographed tea leaves to a data preprocessing module, which preprocesses the tea leaves to obtain a color image and a grayscale image, and then performs feature extraction on the color image and the grayscale image to obtain the morphological characteristic parameters, color characteristic parameters, and texture characteristic parameters of the tea leaves; the data monitoring module is loaded into a tea leaves moisture content monitoring model, and the tea leaves moisture content monitoring model outputs the real-time moisture content of the tea leaves according to the real-time morphological characteristic parameters, color characteristic parameters, and texture characteristic parameters of the tea leaves, so as to determine the degree of difference in the moisture content of the tea leaves in the areas corresponding to the 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 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 at this time, the thresholds 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 ≤5%, indicating that the tea leaves on the stainless steel net are withered evenly, then the withering process continues; if the threshold calculated based on the moisture content of tea leaves in the area corresponding to the photos taken by one or more cameras and the overall average moisture content is >5%, indicating that the tea leaves on the stainless steel net are withered unevenly, then the adjustable inclination conveyor is started. The mechanism tilts the stainless steel mesh by 5°, and then starts the vibration motor to vibrate the stainless steel mesh, so that the tea leaves on it are turned over and mixed; 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 inclination conveying mechanism is started again to make the right side of the withering frame lower than the left side. At the same time, the vibration motor is started to vibrate 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 are unloaded from the outlet, the withering of a new batch of tea leaves is restarted.

[0025] Preferably, the photos of tea leaves on the stainless steel net taken by the camera are processed by the data preprocessing module to obtain the tea leaves' morphological characteristic parameters, color characteristic parameters, and texture characteristic parameters including rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrix. The seven types of data are input into the GA-BP neural network model, and the moisture content of the tea leaves is output. The degree of withering of the tea leaves is judged based on the output moisture content of the tea leaves, thereby ensuring that tea leaves with a similar degree of withering can be obtained in each withering process, standardizing the tea withering process and avoiding the deviation of subjective judgment on the degree of withering of tea leaves, which leads to different degrees of withering for each batch of tea leaves.

[0026] The specific process of processing tea photos by the data preprocessing module is as follows: (1) processing tea photos into color images and grayscale images; (2) performing feature extraction on the color images and grayscale images to obtain morphological characteristic parameters, color characteristic parameters, and texture characteristic parameters of the tea leaves, wherein the morphological characteristic parameters include the major axis, minor axis, rectangularity, circularity, compactness, diagonal length, and slenderness of the tea leaves; the color characteristic parameters are the color characteristic parameters of the tea leaves, including the first-order moments of red, green, blue, hue, saturation, and brightness; and the texture characteristic parameters include the contrast, angular second-order moment, correlation, entropy, and inverse difference matrix of the tea leaves;

[0027] (3) The built-in random forest model in the data preprocessing module was used to rank the importance of morphological feature parameters, color feature parameters, and texture feature parameters, and the important feature parameters of the withering process were screened out, namely rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrix.

[0028] Preferably, the tea moisture content monitoring model is a genetic algorithm optimized back propagation neural network model (GA-BP model), and the specific process of its construction is: (1) the data preprocessing module calls the historical data in the data storage module, and the historical data is the existing tea photos of the withering process and the tea moisture content corresponding to the tea photos. Then, the existing tea photos of the withering process are processed to obtain rectangularity, compactness, green, saturation, contrast, correlation and anti-difference matrix. Then, the obtained rectangularity, compactness, green, saturation, contrast, correlation and anti-difference matrix and the corresponding tea moisture content are normalized. Finally, the normalized data is divided into a training set and a verification set according to a ratio of 4:1; (2) the single-chip microcomputer main control module transmits the above-mentioned training set and verification 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 a 7-10-1 structure to obtain To the optimized BP neural network, where 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 constructs the model 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 verified using the training set and the validation set, and the hyperparameters of the model are adjusted according to various indicators to obtain the GA-BP model; where 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 indicators such as determination coefficient, mean square error, and mean absolute error. The calculation formulas for each indicator are as follows:

[0029]

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

[0031] Preferably, the operation process of the intelligent withering machine is as follows: open the movable cover, and the fresh tea leaves are transported to the withering operation area through the upper leaf conveying mechanism; then, the adjustable inclination conveying mechanism tilts the stainless steel mesh by 5° so that the tea leaves slide in the stainless steel mesh, and at the same time, the vibration motor of the vibrating leaf-turning mechanism starts to vibrate, so that the tea leaves are evenly spread on the stainless steel mesh; after the tea leaves are evenly spread on the stainless steel mesh, the vibration motor stops vibrating, and then the adjustable inclination conveying mechanism restores the stainless steel mesh to a horizontal position; then, the movable cover is closed, the rapid moisture detector obtains the initial moisture content of the tea leaves, the weight sensor obtains the initial weight of the tea leaves, and the withering is completed. At the beginning of the process, the yellow light tube, heat pump unit, hot air unit and moisture content online detection device are started. The hot air output by the heat pump unit is withered from the bottom of the stainless steel mesh through the circulating fan, and the hot air output by the hot air unit is withered from the top of the stainless steel mesh to the bottom and top of the tea leaves respectively. The yellow light tubes on the upper layer of the chassis and the yellow light tubes on the lower layer wither the upper and lower parts of the tea leaves from the top and bottom of the stainless steel mesh respectively. Synchronously, the weight sensor obtains the real-time weight of the tea leaves, and the camera obtains the photos of the tea leaves in the stainless steel mesh every minute. All data are stored in the data storage module in real time. During the withering process, based on the real-time photos of the tea leaves, The tea moisture content monitoring model outputs the tea moisture content of the area corresponding to the photo. 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 photo with the overall average moisture content, the difference in the tea moisture content of the areas corresponding to the photos taken by different cameras can be determined, as well as the difference between the tea moisture content of the area corresponding to the photo and the overall average moisture content of the tea calculated based on the weight sensor. If the moisture content of the tea is uneven, the adjustable inclination conveying mechanism is first started to tilt the stainless steel net, and then the vibration is started. The motor vibrates the stainless steel net, causing the tea leaves on the net to turn over and mix evenly. After mixing, the vibration motor is turned off and the adjustable inclination conveying mechanism is started again to return the stainless steel net to a horizontal position. When the overall average moisture content of the tea leaves in the withering operation area reaches the set target moisture content, the yellow light tube, heat pump unit, hot air unit, and moisture content online detection device are turned off, and then the adjustable inclination conveying mechanism is started to tilt the stainless steel net. Subsequently, the vibration motor is turned on to vibrate the stainless steel net, causing the tea leaves to flow from left to right and flow out from the outlet on the right side of the stainless steel net. When all the tea leaves on the stainless steel net have been unloaded from the outlet, the withering of a new batch of tea leaves begins again.

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

[0033]

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

[0035] A digital withering method of the above-mentioned Lichuanhong intelligent withering machine comprises the following steps:

[0036] Step 1. Start the Lichuanhong intelligent withering machine, open the movable cover, and the tea leaves are transported to the stainless steel net through the upper leaf conveying structure. Then, start the adjustable inclination conveying mechanism to tilt the stainless steel net 5° and then stop. Then start the vibration motor to vibrate the stainless steel net to make the tea leaves slide on the stainless steel net until the tea leaves are evenly spread on the stainless steel net; then, start the adjustable inclination conveying mechanism to restore the stainless steel net to a horizontal position, the loading is completed, and the movable cover is closed.

[0037] Step 2, the single-chip microcomputer main control module sends an inquiry message to the weight sensor. After receiving the inquiry message, the weight sensor determines the initial weight of the tea, and then sends the initial weight of the tea to the single-chip microcomputer main control module. Similarly, after the rapid moisture detector determines the initial moisture content of the tea and the temperature and humidity sensor determines the temperature and humidity in the withering operation area, the initial moisture content of the tea and the temperature and humidity in the withering operation area are sent to the single-chip microcomputer main control module. The withering operation area is the temperature and humidity about 10 cm above the stainless steel net. Specifically: the single-chip microcomputer main control module obtains the initial moisture content Q1 of the tea collected by the rapid moisture detector through the RS232-1 interface of the communication module, and obtains the initial weight W1 of the tea 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 single-chip microcomputer 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, and the WiFi communication module transmits the data wirelessly to the host computer. The single-chip microcomputer 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, temperature upper limit, temperature lower limit, humidity upper limit, humidity lower limit and process program on the touch screen, click "Start Withering", the withering process starts, start the yellow light lamp, heat pump unit, hot air unit, moisture content online detection device, dehumidification fan and circulating fan, the hot air output by the heat pump unit withers the lower and upper parts of the tea leaves from the bottom of the stainless steel mesh through the circulating fan, and the hot air output by the hot air unit withers the upper and lower parts of the tea leaves from the top and bottom of the stainless steel mesh respectively, the yellow light lamps on the upper layer of the chassis and the yellow light lamps on the lower layer wither the upper and lower parts of the tea leaves from the top and bottom of the stainless steel mesh respectively, synchronously, the weight sensor obtains the real-time weight of the tea leaves, the camera obtains the photo of the tea leaves in the stainless steel mesh once at a set time, and all data are stored in the data storage module in real time.

[0039] When the current temperature obtained by real-time monitoring of the temperature and humidity sensor is higher than the input temperature upper limit value, the single-chip microcomputer main control module sends a signal to turn on the alarm, turn off the hot air blower, and turn off the heat pump unit. The alarm is turned on, and the hot air blower and heat pump unit are turned off. The dehumidification fan and the circulation fan continue to work to lower the temperature in the withering operation area to prevent the tea leaves from losing water too quickly. When the current temperature is not higher than the temperature upper limit value, the alarm is turned off and the hot air blower and heat pump unit are turned on. Similarly, when the current temperature obtained by real-time monitoring of the temperature and humidity sensor is lower than the input temperature lower limit value, the single-chip microcomputer main control module sends a signal to turn on the alarm, increase the power of the hot air blower, and increase the power of the heat pump unit. The alarm is turned on to increase the temperature of the hot air delivered by the hot air blower and the heat pump unit to increase the withering temperature. The temperature of the operation area is adjusted to prevent the tea from becoming astringent and gray due to low temperature. When the current temperature is not lower than the input lower temperature limit, the power of the heat pump unit and the power of the hot air blower return to their original power. When the current humidity obtained by the temperature and humidity sensor in real-time monitoring is higher than the input upper humidity limit, the main control module of the single-chip microcomputer sends a signal to increase the power of the dehumidification fan to improve the dehumidification efficiency of the dehumidification fan. When the current humidity is not higher than the upper humidity limit, the power of the dehumidification fan returns to its original power. Similarly, when the current humidity obtained by the temperature and humidity sensor in real-time monitoring is lower than the input lower humidity limit, the main control module of the single-chip microcomputer sends a signal to reduce the power of the dehumidification fan to reduce the dehumidification efficiency of the dehumidification fan. When the current humidity is not lower than the lower humidity limit, the power of the dehumidification fan returns to its original power.

[0040] Step 4: The moisture content monitoring unit calculates the current overall average moisture content Q2 based on the stored initial weight W1, initial moisture content Q1, and current real-time weight W2 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 the important characteristic parameters of the withering process, and the specific process is: (1) processing the tea photos into color images and grayscale images; (2) extracting features from the color images and grayscale images to obtain the morphological characteristic parameters, color characteristic parameters, and texture characteristic parameters of the tea leaves, wherein the morphological characteristic parameters include the major axis, minor axis, rectangularity, circularity, compactness, diagonal length, and slenderness of the tea leaves; the color characteristic parameters are the color characteristic parameters of the tea leaves, including the first-order moments of red, green, blue, hue, saturation, and lightness; and the texture characteristic parameters include the contrast, angular second-order moment, correlation, entropy, and inverse difference matrix of the tea leaves; (3) using the built-in random forest model in the data preprocessing module to rank the morphological characteristic parameters, color characteristic parameters, and texture characteristic parameters, and screening out the important characteristic parameters of the withering process, namely rectangularity, compactness, green, saturation, contrast, correlation, and inverse difference matrix.

[0043] Step 6, the single-chip main control module stores the rectangularity, compactness, green, saturation, contrast, correlation and anti-difference matrix obtained in step 5 into the data storage module; at the same time, the rectangularity, compactness, green, saturation, contrast, correlation and anti-difference matrix are input as input data into the tea moisture content monitoring model in the data monitoring module, and the tea moisture content monitoring model outputs the tea moisture content Q3, and compares Q3 with Q2 according to the following formula:

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

[0045] If the threshold values ​​calculated based on the moisture content Q3 of the tea leaves in the area corresponding to the photos taken by the four cameras and the overall average moisture content Q2 are both ≤5%, indicating that the tea leaves on the stainless steel net are withered evenly, the withering process continues; if the threshold values ​​calculated based on the moisture content Q3 of the tea leaves in the area corresponding to the photos taken by one or more cameras and the overall average moisture content Q2 are greater than 5%, indicating that the tea leaves on the stainless steel net are withered unevenly, the adjustable inclination conveying mechanism is started to tilt the stainless steel net by 5°, and then the vibration motor is started to vibrate the stainless steel net so that the tea leaves on it are turned over and mixed evenly.

[0046] 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 inclination conveying mechanism is started to make the right side of the withering frame lower than the left side. At the same time, the vibration motor is started to vibrate the stainless steel net, 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 net are unloaded from the outlet, the withering of a new batch of tea leaves is restarted.

[0047] Compared with the prior art, the present invention has the following advantages:

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

[0049] 2. In the present invention, the color and texture information of the tea photos in the withering operation area collected by each camera is compared to judge the uniformity of the tea withering process, thereby controlling the operation of the vibration turning mechanism, thereby improving the uniformity of the tea withering process and improving the quality of the tea.

[0050] 3. In the present invention, through the coordinated use of a vibrating leaf-turning mechanism and an adjustable inclination conveying mechanism, continuous withering operations in tea processing can be realized, which can save labor; the weight of the tea is measured by a weight sensor and the color and texture data of the tea in the withering frame are collected by a camera to predict the moisture content of the tea during the withering process, thereby realizing intelligent withering operations of the tea, which can effectively improve work efficiency.

[0051] 4. The present invention uses a rapid moisture detector in conjunction with a weight sensor to automatically obtain moisture content data and weight data of fresh tea leaves, and uses a main control system to automatically calculate the overall moisture content of the tea leaves. Compared with traditional methods, the moisture content data is more accurate, the operation is simple and time-saving, and the moisture content of withered leaves can be predicted more accurately. Users do not need to frequently manually detect the degree of withering, and the workload is greatly reduced, which is beneficial to improving the stability of the quality of the finished tea.

[0052] 5. The present invention can program the withering process according to production needs to achieve timed operation, which can save labor; at the same time, relying on the equipped touch screen, production data and control switches are intuitively displayed, and it has a friendly human-computer interaction interface, which is convenient for application and promotion in production.

[0053] 6. The present invention is equipped with a data storage module to record initial production data and process data to avoid loss of initial data due to accidental shutdown. In addition, data can be transmitted to the host computer through the UART-4 port via the WiFi communication module to provide data support for improving tea quality in the future.

[0054] 7. The present invention is equipped with a WiFi communication module to achieve wireless communication with the host computer, data interaction and control command transmission, which simplifies the wiring process in the production environment.

[0055] 8. The present invention has written control software for the host computer to realize data interaction and control command transmission between the computer and the main control system, calculate the current water loss rate of withering leaves in real time and realize the withering completion time prediction function.

[0056] 9. The present invention discloses a method for predicting the moisture content of tea leaves during withering process based on random forest features and genetic algorithm optimized back propagation 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 more accurately.

[0057] To sum up, the present invention solves the problems of poor uniformity in the withering process of tea leaves and difficulty in automatic leaf turning and loading and unloading through the coordinated use of various mechanisms, improves the quality of Lichuan Red, and makes the processed Lichuan Red have the characteristics of high efficiency, high standards, and high intelligence, which is of great significance to the development of Lichuan Red. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1It is a structural front view of the present invention;

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

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

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

[0062] Figure 5 It is a left view of the structure of the inclination-adjustable vibrating conveyor of the present invention;

[0063] Figure 6 It is a right view of the structure of the inclination-adjustable vibrating conveyor of the present invention;

[0064] Figure 7 It is a schematic diagram of the lifting structure of the inclination-adjustable vibrating conveyor of the present invention;

[0065] Figure 8 A schematic diagram of a support for the inclination-adjustable vibrating conveyor of the present invention;

[0066] Fig. 9 It is a structural diagram of the main control system of the present invention;

[0067] Fig.10 It is the main program flow chart of the present invention;

[0068] Fig.11 It is a schematic flow chart of the withering moisture content prediction method based on random forest-back propagation neural network of the present invention;

[0069] Fig.12 It is a schematic diagram of the working principle of the heat pump unit of the present invention;

[0070] Serial numbers in the figure: heat pump unit 1, rapid moisture detector 2, hot air blower 3, camera 4, vibration spring 5, stainless steel mesh 6, withering operation area 7, excitation motor 8, hot air main pipeline 9, hot air branch pipeline 10, dehumidification fan II 11, adjustable inclination conveying mechanism 12, conveying trough 12-1, shaft sleeve I 12-2, right-angle lifter 12-3, transmission shaft 12-4, reduction motor 12-5, rotating bearing I 12-6, slider 12-7, guide rail 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 14, weight sensor 15, circulating fan II 16, control box 17, air duct 18, upper leaf conveying mechanism 19, fixed cover 20, movable cover 21, evaporator 22, gas-liquid separator 23, compressor 24, condenser 25, circulating fan I 26, dehumidification fan I 27, temperature and humidity sensor 28, alarm light 29. DETAILED DESCRIPTION

[0071] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0072] Example 1

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

[0074] The withering tank body includes a chassis 13, an air duct 18, a stainless steel mesh 6, a withering operation area 7, and a yellow light tube 14. The frame 13 is divided into three layers from top to bottom, the upper layer is the withering operation area 7, the upper half of the middle layer is a vibrating leaf-turning mechanism, the lower half is an adjustable inclination conveying mechanism 12, and the top of the lower layer is provided with an array formed by yellow light tubes 14. In the chassis 13, the withering operation area 7 includes a conveying trough 12-1, an array formed by yellow light tubes 14 is provided on the four side walls of the withering operation area 7, the bottom end is a stainless steel mesh 6 fixedly connected to the chassis 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 conveying trough 12-1, a hot air main duct 9 and a hot air branch duct 10 are provided in the middle position of the upper layer, and a camera 4 is provided at each of the four corners at the top of the upper layer, i.e., the four corners at the top of the chassis 13. In the lower layer, both ends of the yellow lamp tube 14 are fixed to the chassis 13 by brackets, and the yellow lamp tube 14 is linearly and equidistantly distributed. The top surface of the chassis 13 is a fixed cover 20 and a movable cover 21. Among them, the stainless steel mesh 6 is used to spread tea leaves. The yellow lamp tube 14 installed on the side wall above the stainless steel mesh 6 and the yellow lamp tube 14 installed in the lower layer provide yellow light irradiation. Yellow light withering is beneficial to the increase of amino acid content and the formation of aldehyde substances in tea leaves, and significantly improves the aroma of tea leaves; the temperature and humidity sensor 28 on the stainless steel mesh 6 is used to measure the temperature and humidity of the upper withering operation area 7. Appropriate temperature and humidity can be beneficial to the transformation of substances contained in tea leaves, thereby improving the quality of tea leaves.

[0075] like Fig.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 a dehumidifying fan I 27. The heat pump unit 1 is located on the left side of the withering tank body. An air duct 18 connecting the heat pump unit 1 and the withering tank body is formed at the lower part between the heat pump unit 1 and the withering tank body. A circulating fan II 16 is installed on the withering tank side of the air duct 18, and a dehumidifying fan II 11 is installed in the upper right corner just in front of the withering tank body. The heat pump unit absorbs heat from the surrounding environment and outputs hot air. The hot air enters the air duct 18, and the hot air in the air duct is sent to the lower layer of the chassis 13 through 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 therein. The heat pump unit uses the heat absorbed from the surrounding environment for the withering of the tea leaves on the stainless steel mesh 6, which can save energy, reduce the emission of pollutants such as CO2, and achieve the effect of energy saving and emission reduction.

[0076] The hot air unit includes a hot air blower 3, a hot air main duct 9, and a hot air branch duct 10. The hot air blower 3 is installed on a bracket between the chassis 13 and the heat pump unit 1. The outlet of the hot air blower 3 is connected to the inlet of the hot air main duct 9. The hot air branch ducts 10 are equidistantly connected to the hot air main duct 9. Small holes with a diameter of 1 cm are evenly arranged on the hot air branch ducts 10. The hot air main duct 9 is installed above the stainless steel mesh and fixed to the chassis 13 through a bracket. The hot air unit is used to transport hot air to wither the tea leaves from above the tea leaves on the stainless steel mesh 6. The existing withering tanks are all ventilated and heated from the bottom, which easily causes uneven withering of the upper and lower tea leaves during the withering process. The present invention adopts 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 more uniform transformation of substances inside the tea leaves.

[0077] The vibrating leaf-turning mechanism includes a vibration motor 8 and a vibration spring 5. Four vibration springs 5 ​​are respectively located below the four corners of the stainless steel mesh 6, and the vibration motor 8 is located in the middle position between every two springs. The vibrating leaf-turning mechanism makes the tea leaves in the stainless steel mesh 6 more evenly dispersed, thereby withering more evenly.

[0078] The adjustable inclination conveying mechanism 12 includes a sleeve Ⅰ12-2, a right angle lifter 12-3, a transmission shaft 12-4, a reduction motor 12-5, a rotating bearing Ⅰ12-6, a slider 12-7, a guide rail 12-8, a support 12-9, a sleeve Ⅱ12-10, and a rotating bearing Ⅱ12-11. The adjustable inclination conveying mechanism 12 is located below the vibrating leaf turning mechanism. The left side of the adjustable inclination conveying mechanism 12 is two right angle lifters 12-3, and the right side is two supports 12-9. The two right angle lifters 12-3 and the two supports 12-9 are respectively located below a vibration spring 5; the bottom end of the right angle lifter 12-3 is installed on the bottom corner of one side of the frame of the middle layer of the chassis 13, and the top end is the lifting end connected to the sleeve Ⅰ12-2. One end of the rotating bearing Ⅰ12-6 is connected to the sleeve Ⅰ12-2 through the rotating shaft 12-13, and the other end is connected to the slider 1 2-7 is connected, the slider 12-7 is connected to the guide rail 12-8, and the guide rail 12-8 is installed under the vibration spring 5; the bottom end of the support 12-9 is installed on the corner of the bottom end of the other side of the frame of the middle layer of the chassis 13, and the top end of the support 12-9 is installed with a sleeve II, and one end of the rotating bearing II 12-11 is connected to the sleeve II 12-10 through the connecting shaft 12-12, and the other end is installed under the vibration spring 5; the reduction motor 12-5 is installed in the middle between the two right-angle lifters 12-3, and is connected to the right-angle lifter 12-3 through the transmission shaft 12-4. Different tea leaves have different flow characteristics. The upper leaf conveying mechanism 19 conveys the tea leaves to the stainless steel mesh 6, and then closes the movable cover, tilts the stainless steel mesh 6 by 5° through the adjustable inclination conveying mechanism 12, and starts the vibration leaf turning mechanism to excite the stainless steel mesh 6, so that the tea leaves are evenly spread on the stainless steel mesh 6.

[0079] The moisture content online detection device includes a weight sensor 15, a rapid moisture detector 2, and a camera 4. The weight sensor 15 is installed below the chassis 13 of the entire withering tank body; the four cameras 4 are respectively installed at the four corners of the top of the withering tank body chassis 13, fixed on the frame 13 by brackets, and distributed diagonally; the rapid moisture detector 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 the real-time overall average moisture content of the tea leaves during the withering process is obtained according to the real-time weight of the tea leaves and the initial moisture content.

[0080] The main control system includes a control box 17, which is used for digital control of the withering process and determines whether the real-time overall average moisture content of the tea leaves during the withering process reaches the set target moisture content. A touch screen is installed on the surface of the control box 17, and a main control board is arranged inside it. The target moisture content, the upper temperature limit and the process program are input on the touch screen. The main control board includes a power supply module, a communication module, a WiFi communication module, a single-chip main control module, an IO control module, and a data storage module. The single-chip main control module includes a moisture content monitoring unit, and the moisture content monitoring unit includes a data preprocessing module and a data monitoring module.

[0081] The output end of the power supply module (manufacturer: MORNSUN, model: K7805-2000R3) is electrically connected to the input end of the communication module, WiFi communication module, single-chip main control module, IO control module, and data storage module, and is used to supply power to the communication module, single-chip main control module, IO control module, and data storage module; the power supply module also serves as a power source for external devices, including an electromagnetic relay, a touch screen, a rapid moisture detector 2, a weight sensor 15, a temperature and humidity sensor 28, and an alarm light 29.

[0082] The communication module (manufacturer: MORNSUN, model: TD331S232H) is connected to the weight sensor, the temperature and humidity sensor, the rapid moisture detector, and the single-chip main control module for data communication with the weight sensor (Puston, model: PSD-F1), the temperature and humidity sensor (Dekong Senshe, model: DK-3000-WS), the rapid moisture detector (Shanghai Lichen Bangxi Instrument Technology Co., Ltd., model: LC-DHS-16), and the single-chip main control module;

[0083] The WiFi communication module (AnKeKe, model: ESP-07S) is connected to the single-chip microcomputer main control module to achieve wireless communication with the host computer.

[0084] The output end of the IO control module is connected to the input end of an electromagnetic relay (manufacturer: TONGLING, model: T90-5VDC-TL-C), wherein the output end of the electromagnetic relay 1 is connected to the alarm 29, the output end of the electromagnetic relay 2 is connected to the hot air blower 3, the output end of the electromagnetic relay 3 is connected to the heat pump unit 1, the output end of the electromagnetic relay 4 is connected to the vibration leaf-turning mechanism, the output end of the electromagnetic relay 5 is connected to the adjustable inclination angle conveying mechanism 12, the output end of the electromagnetic relay 6 is connected to the circulation fan 11, and the output end of the electromagnetic relay 7 is connected to the dehumidification fan 11, which is used to control the start and stop of the hot air blower 3, the circulation fan 16, the dehumidification fan 11, the heat pump unit 1, the alarm 29, the vibration leaf-turning mechanism, and the adjustable inclination angle conveying mechanism 12.

[0085] The single-chip microcomputer main control module (manufacturer: Jiali Chuang, model: GD32F470ZGT6) is the core of the entire main control system and is responsible for the collection, processing, transmission of sensor data and the control of the withering process;

[0086] The data storage module (manufacturer: MICROCHIP (US Microchip), model: AT24C256C-SSHL-T) is connected to the single-chip microcomputer main control module and is used to store various data.

[0087] The communication module has RS485-1, RS232-1, and UART interfaces. RS485-1, RS232-1, and UART interfaces provide data exchange channels for external devices to ensure the normal operation of the system. Among them, the RS485-1 interface is responsible for communicating with the weighing sensor and the temperature and humidity sensor, and the RS232-1 interface is responsible for receiving data transmitted by the rapid moisture detector. The UART interface includes UART-1, UART-2, UART-3, and UART-4. The UART-1 is used to transmit the RS485-1 signal after level conversion into the single-chip main control module, and the UART-2 is used to transmit the RS232-1 signal after level conversion into the single-chip main control module. The UART-4 and UART-3 are directly derived from the single-chip main control module and are respectively used to communicate with the WiFi communication module and the touch screen.

[0088] The data storage module is connected to the single-chip main control module through the IIC-1 interface. The single-chip main control module receives various data and stores them in the data storage module, including initial moisture content, real-time weight, and target moisture content data. The single-chip main control module transmits the current time and the current weight, current moisture content, current temperature and humidity at the current time, and characteristic parameter data obtained by preprocessing the photos taken by the camera 4 to the data storage module according to the set time interval during the withering process. The initial moisture content, initial weight, and target moisture content are displayed on the touch screen, and various data of the withering process can be output on the host computer.

[0089] The working process of the moisture content monitoring unit is as follows: during the withering process, the camera 4 takes pictures of the tea leaves in the withering frame every 20 minutes, and transmits the photographed tea leaves to the data preprocessing module, which preprocesses the tea leaves to obtain color images and grayscale images, and then extracts features from the color images and grayscale images to obtain the morphological characteristic parameters, color characteristic parameters, and texture characteristic parameters of the tea leaves; the data monitoring module is loaded with a tea moisture content monitoring model, and the tea moisture content monitoring model outputs the real-time moisture content of the tea leaves according to the real-time tea morphological characteristic parameters, color characteristic parameters, and texture characteristic parameters, so as to judge the degree of difference in the moisture content of the tea leaves in the areas corresponding to the photos taken by different cameras 4, and to judge 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 of the tea leaves calculated based on the weight measured by the weight sensor. According to the formula: threshold value = |moisture content of tea leaves in the area corresponding to the photo - overall average moisture content | / overall average moisture content ×%, if at this time, the threshold values ​​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 all ≤5%, indicating that the tea leaves on the stainless steel mesh 6 are withered evenly, then the withering process continues; if the threshold value calculated based on the moisture content of tea leaves in the area corresponding to the photos taken by one or more cameras and the overall average moisture content is greater than 5%, indicating that the tea leaves on the stainless steel mesh 6 are withered unevenly, then the adjustable inclination conveying mechanism 12 is started to make the stainless steel mesh 6 wither unevenly. The steel mesh 6 is tilted 5°, and then the vibration motor 8 is started to vibrate the stainless steel mesh 6, so that the tea leaves thereon are turned over and mixed; 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 inclination conveying mechanism 12 is started again to make the right side of the withering frame lower than the left side. At the same time, the vibration motor 8 is started to vibrate the stainless steel mesh 6, so that the tea leaves are prompted 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 are unloaded from the outlet, the withering of a new batch of tea leaves is restarted.

[0090] The tea leaves photographed on the stainless steel mesh 6 taken by the camera 4 are processed by the data preprocessing module to obtain tea leaves morphological characteristic parameters, color characteristic parameters, texture characteristic parameters including rectangularity, compactness, green, saturation, contrast, correlation and inverse difference matrix. The seven kinds of data are input into the GA-BP neural network model, and the moisture content of the tea leaves is output. The withering degree of the tea leaves is judged based on the output moisture content of the tea leaves, thereby ensuring that each withering process can obtain tea leaves with a similar withering degree, standardizing the tea withering process, and avoiding the deviation of subjective judgment on the withering degree of tea leaves, which leads to different withering degrees for each batch of tea leaves.

[0091] The specific process of processing tea photos by the data preprocessing module is as follows: (1) processing tea photos into color images and grayscale images; (2) extracting features from the color images and grayscale images to obtain the morphological characteristic parameters, color characteristic parameters, and texture characteristic parameters of the tea leaves, wherein the colorhist function is used to calculate the color characteristic parameters of the color image obtained by processing the tea photos, the graycomatrix function is used to calculate the texture characteristic parameters of the grayscale image obtained by processing the tea photos, and the median filter is used, followed by the Otsu method, and then the erosion-dilation method, and finally the regionprops function is used to calculate the grayscale image obtained by processing the tea photos to obtain the morphological characteristic parameters. The morphological characteristic parameters include the major axis, minor axis, rectangularity, circularity, compactness, diagonal length, and slenderness of the tea leaves. The expressions of the five morphological characteristic parameters of rectangularity, circularity, compactness, diagonal length, and slenderness are as follows:

[0092]

[0093] J=D / L

[0094]

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

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

[0097]

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

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

[0100] (3) Fig.11The built-in random forest model in the data preprocessing module was used to rank the importance of morphological feature parameters, color feature parameters, and texture feature parameters. The number of decision trees of the random forest model was 200, the minimum number of leaves was 10, and the threshold was 0.5. The important feature parameters of the withering process were rectangularity, compactness, greenness, saturation, contrast, correlation, and anti-difference matrix.

[0101] The tea moisture content monitoring model is a genetic algorithm optimized back propagation neural network model (GA-BP model), and the specific construction process is as follows: (1) the data preprocessing module calls the historical data in the data storage module, and the historical data is the existing tea photos of the withering process and the tea moisture content corresponding to the tea photos. Then, the existing tea photos of the withering process are processed to obtain the rectangularity, compactness, greenness, saturation, contrast, correlation and anti-difference matrix. Then, the obtained rectangularity, compactness, greenness, saturation, contrast, correlation and anti-difference matrix and the corresponding tea moisture content are normalized. Finally, the normalized data is divided into a training set and a verification set according to a ratio of 4:1; (2) the single-chip microcomputer main control module transmits the above training set and verification 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 of the 7-10-1 structure to obtain the optimal The optimized 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 verified using the training set and the validation set, and the hyperparameters of the model are adjusted according to various indicators to obtain the GA-BP model; the normalized rectangularity, compactness, greenness, saturation, contrast, correlation and inverse difference matrix are used as input data, and the normalized moisture content of tea is used as output data. The performance of the model is evaluated using indicators such as determination coefficient, mean square error, and mean absolute error. The calculation formulas for each indicator are as follows:

[0102]

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

[0104] The operation process of the intelligent withering machine is as follows: open the movable cover 21, and the fresh tea leaves are transported to the withering operation area 7 through the upper leaf conveying mechanism 19. Then, the adjustable inclination conveying mechanism 12 tilts the stainless steel mesh 6 by 5° so that the tea leaves slide in the stainless steel mesh 6. At the same time, the vibration motor 8 of the vibrating leaf turning mechanism starts to vibrate, so that the tea leaves are evenly spread on the stainless steel mesh 6; after the tea leaves are evenly spread on the stainless steel mesh 6, the vibration motor 8 stops vibrating, and then the adjustable inclination conveying mechanism 12 restores the stainless steel mesh 6 to a horizontal position; then, 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, the yellow light tube 14, the heat pump unit 1, the hot air unit, and the moisture content online detection device are started, the hot air output by the heat pump unit passes through the circulating fan 16 from the bottom of the stainless steel mesh 6, and the hot air output by the hot air unit passes through the top of the stainless steel mesh 6 to wither the lower and upper parts of the tea leaves respectively, the yellow light tube 14 on the upper layer of the chassis and the yellow light tube 14 on the lower layer of the chassis wither the upper and lower parts of the tea leaves from the top and bottom of the stainless steel mesh 6 respectively, synchronously, the weight sensor 15 obtains the real-time weight of the tea leaves, the camera obtains the photos of the tea leaves in the stainless steel mesh 6 every 20 minutes, and all the data are stored in the data storage module in real time; during the withering process, based on Real-time tea photos, the tea moisture content monitoring model outputs the tea moisture content of the area corresponding to the photo, 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 photo with the overall average moisture content, so as to judge the difference degree of the tea moisture content of the area corresponding to the photos taken by different cameras 4, and judge the difference degree between the tea moisture content of the area corresponding to the photo and the overall average moisture content of the tea calculated based on the weight sensor; if the moisture content of the tea is uneven, first start the adjustable inclination conveying mechanism 12 to tilt the stainless steel net 6, and then start the exciting motor 8 to excite The stainless steel mesh 6 is vibrated to make the tea leaves on the stainless steel mesh 6 turn over and mix evenly. After mixing evenly, the vibration motor 8 is turned off and the adjustable inclination conveying mechanism 12 is started again to make the stainless steel mesh 6 return to a horizontal position. When the overall average moisture content of the tea leaves in the withering operation area 7 reaches the set target moisture content, the yellow light tube 14, the heat pump unit 1, the hot air unit and the moisture content online detection device are turned off, and then the adjustable inclination conveying mechanism 12 is started to tilt the stainless steel mesh 6. Subsequently, the vibration motor 8 is turned on to vibrate the stainless steel mesh 6 to cause the tea leaves to flow from left to right and flow out from the outlet on the right side of the stainless steel mesh 6. When all the tea leaves on the stainless steel mesh 6 are unloaded from the outlet, the withering of a new batch of tea leaves is restarted.

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

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

[0107]

[0108] Example 2

[0109] A digital withering method of the Lichuan Red intelligent withering machine of embodiment 1 comprises the following steps:

[0110] Step 1, start the Lichuanhong intelligent withering machine, open the movable cover 21, and the tea leaves are transported to the stainless steel mesh 6 through the upper leaf conveying structure 19, then start the adjustable inclination conveying mechanism 12, make the stainless steel mesh 6 tilt 5° and then stop, then start the vibration 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; then, start the adjustable inclination conveying mechanism 12 to restore the stainless steel mesh 6 to a horizontal position, the loading is completed, and the movable cover 21 is closed.

[0111] Step 2, the single-chip microcomputer 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 single-chip microcomputer 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 in the withering operation area 7 (i.e., the temperature and humidity about 10 cm above the stainless steel mesh 6). Then, the initial moisture content of the tea leaves and the temperature and humidity in the withering operation area 7 are sent to the single-chip microcomputer main control module: the single-chip microcomputer main control module receives the initial moisture content of the tea leaves and the temperature and humidity in the withering operation area 7 through the RS232 of the communication module. -1 interface obtains the initial moisture content Q1 of the tea collected by the rapid moisture detector 2, and obtains the initial weight W1 of the tea 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 MCU main control module sends the obtained data to the WiFi communication module and the touch screen through the UART-4 and UART-3 interfaces. The touch screen displays the data, and the WiFi communication module transmits the data wirelessly to the host computer. The MCU main control module sends the obtained data to the data storage module through the IIC-1 interface for storage.

[0112] Step 3, input the target moisture content of 58%, the upper temperature limit of 25°C, the lower temperature limit of 33°C, the upper humidity limit of 80%, the lower humidity limit of 65% and the process program on the touch screen, click "Start Withering", the withering process starts, start the yellow light lamp 14, the heat pump unit 1, the hot air unit, the moisture content online detection device, the dehumidification fan 11, and the circulating fan 16, the hot air output by the heat pump unit passes through the circulating fan 16 from the bottom of the stainless steel mesh 6, and the hot air output by the hot air unit passes through the top of the stainless steel mesh 6 to wither the lower and upper parts of the tea leaves respectively, the yellow light lamp 14 on the upper layer of the chassis and the yellow light lamp 14 on the lower layer wither the upper and lower parts of the tea leaves from the top and bottom of the stainless steel mesh 6 respectively, synchronously, the weight sensor 15 obtains the real-time weight of the tea leaves, the camera 4 obtains the photos of the tea leaves in the stainless steel mesh 6 every 20 minutes, and all data are stored in the data storage module in real time.

[0113] When the current temperature obtained by the real-time monitoring of the temperature and humidity sensor is higher than the input temperature upper limit value, the single-chip microcomputer main control module sends a signal to turn on the alarm, turn off the hot air blower, and turn off the heat pump unit, turn on the alarm, turn off the hot air blower and the heat pump unit, and the dehumidification fan 11 and the circulation fan 16 continue to work to lower the temperature of the withering operation area 7 to prevent the tea from losing water too quickly. When the current temperature is not higher than the temperature upper limit value, the alarm is turned off and the hot air blower and the heat pump unit are turned on. Similarly, when the current temperature obtained by the real-time monitoring of the temperature and humidity sensor is lower than the input temperature lower limit value, the single-chip microcomputer main control module sends a signal to turn on the alarm, increase the power of the hot air blower, and increase the power of the heat pump unit. The alarm is turned on to increase the temperature of the hot air delivered by the hot air blower and the heat pump unit to increase the withering operation area. Domain 7 temperature, to prevent the tea from becoming astringent and gray and withered due to low temperature, until the current temperature is not lower than the input temperature lower limit, the power of the heat pump unit and the power of the hot air blower return to the original power; when the current humidity obtained by the temperature and humidity sensor in real-time monitoring is higher than the input humidity upper limit, the single-chip microcomputer main control module sends a signal to increase the power of the dehumidification fan 11, thereby improving the dehumidification efficiency of the dehumidification fan 11, until the current humidity is not higher than the humidity upper limit, the power of the dehumidification fan 11 returns to the original power; similarly, when the current humidity obtained by the temperature and humidity sensor in real-time monitoring is lower than the input humidity lower limit, the single-chip microcomputer main control module sends a signal to reduce the power of the dehumidification fan 11, thereby reducing the dehumidification efficiency of the dehumidification fan 11, until the current humidity is not lower than the humidity lower limit, the power of the dehumidification fan 11 returns to the original power.

[0114] Step 4: The moisture content monitoring unit calculates the current overall average moisture content Q2 based on the stored initial weight W1, initial moisture content Q1, and current real-time weight W2 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 the important characteristic parameters of the withering process, and the specific process is: (1) Processing the tea photos into color images and grayscale images; (2) Extracting features from the color images and grayscale images to obtain the morphological characteristic parameters, color characteristic parameters, and texture characteristic parameters of the tea leaves, wherein the colorhist function is used to calculate the color characteristic parameters of the color image obtained by processing the tea photos, the graycomatrix function is used to calculate the texture characteristic parameters of the grayscale image obtained by processing the tea photos, and the median filter is used, followed by the Otsu method, and then the erosion-dilation method, and finally the regionprops function is used to calculate the grayscale image obtained by processing the tea photos to obtain the morphological characteristic parameters. The morphological characteristic parameters include the major axis, minor axis, rectangularity, circularity, compactness, diagonal length, and slenderness of the tea leaves. The expressions of the five morphological characteristics of rectangularity, circularity, compactness, diagonal length, and slenderness are:

[0117]

[0118]

[0119] J=D / L

[0120]

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

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

[0123]

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

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

[0126] Step 6, the single-chip main control module stores the rectangularity, compactness, green, saturation, contrast, correlation and anti-difference matrix obtained in step 5 into the data storage module; at the same time, the rectangularity, compactness, green, saturation, contrast, correlation and anti-difference matrix are input as input data into the tea moisture content monitoring model in the data monitoring module, and the tea moisture content monitoring model outputs the tea moisture content Q3, and compares Q3 with Q2 according to the following formula:

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

[0128] If the threshold values ​​calculated based on the moisture content Q3 of the tea leaves in the area corresponding to the photos taken by the four cameras and the overall average moisture content Q2 are both ≤5%, indicating that the tea leaves on the stainless steel mesh 6 are withered evenly, the withering process continues; if the threshold values ​​calculated based on the moisture content Q3 of the tea leaves in the area corresponding to the photos taken by one or more cameras and the overall average moisture content Q2 are greater than 5%, indicating that the tea leaves on the stainless steel mesh 6 are withered unevenly, the adjustable inclination conveying mechanism 12 is started to tilt the stainless steel mesh 6 by 5°, and then the vibration motor 8 is started to vibrate the stainless steel mesh 6, so that the tea leaves thereon are turned over and mixed evenly.

[0129] If the overall average moisture content Q2 reaches the set target moisture content of 58%, the heat pump unit 1, the hot air unit and the yellow light tube 14 are turned off, the air supply heating and the yellow light irradiation are stopped, then the movable cover 21 is opened, and the adjustable inclination conveying mechanism 12 is started to make the right side of the withering frame lower than the left side. At the same time, the vibration motor 8 is started to vibrate the stainless steel mesh 6 to cause 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 are unloaded from the outlet, the withering of a new batch of tea leaves is restarted.

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

[0131] Table 1 Comparison of the moisture content of tea leaves of different grades of tea raw materials Q3 and the moisture content of tea leaves corresponding to the tea photos in the historical data

[0132]

[0133]

[0134] As shown in Table 1, there is a gap between the moisture content of each area in the withering operation area 7 and the overall moisture content, indicating that there are inconsistent fresh leaf conditions and different moisture contents of tea leaves in different areas, which will lead to uneven withering in the withering process of tea leaves. The areas photographed by different cameras can only represent the moisture content of the area, and cannot reflect the overall moisture content. However, the moisture content of tea leaves in each area of ​​the withering operation area 7 is not much different from the measured overall average moisture content, indicating that the moisture content of tea leaves can be well monitored by the present invention. In addition, it also shows that the present invention can more quickly and accurately reflect the withering degree of the overall tea leaves.

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

[0136]

[0137] As shown in Table 2, when using one bud and two leaves raw material, the overall score of the withering machine main control system is slightly higher than that of manual judgment, which is mainly reflected in the aroma and taste. The tea judged by the withering machine main control system to be moderately withered has a stronger sweet and fruity aroma, and the aftertaste lasts longer.

[0138] Table 3 Sensory evaluation results of Lichuan black tea leaves obtained by machine-picked leaves and different withering methods

[0139]

[0140]

[0141] As shown in Table 3, using machine-picked leaves as raw materials, the overall score of the withering machine main control system was slightly higher than that of manual judgment, and the dry tea had a more oily color, a higher sweetness in aroma, and less bitterness.

[0142] The present invention solves the problems of poor uniformity in the withering process of tea leaves and difficulty in automatic leaf turning and loading and unloading through the coordinated use of various mechanisms, improves the quality of Lichuan Red, and makes the processed Lichuan Red have the characteristics of high efficiency, high standards, high intelligence, etc., which is of great significance to the development of Lichuan Red.

[0143] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A Lichuan red intelligent withering machine, comprising a withering tank, a heat pump unit, a hot air unit, a vibrating leaf turning mechanism, an adjustable inclination conveying mechanism, an online moisture content detection device, and a main control system; characterized in that: The withering tank body includes a chassis, an air duct, a stainless steel net, a withering operation area, and a yellow light tube; the frame 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 inclination conveying mechanism, and the top of the lower layer is provided with an array formed by yellow light tubes; Inside the chassis, the withering operation area includes a conveying trough, an array of yellow light tubes is arranged on the four side walls of the withering operation area, and the bottom end is a stainless steel net fixedly connected to the chassis, two temperature and humidity sensors are fixedly installed on the stainless steel net, and the space above the stainless steel net is a conveying trough, a hot air main duct and a hot air branch duct are arranged in the middle position of the upper layer, and a camera is arranged at each of the four corners at the top of the upper layer, that is, the four corners at the top of the chassis; in the lower layer, the two ends of the yellow light tube are fixed to the chassis by brackets, and the yellow light tubes are linearly equidistantly distributed; the top surface of the chassis is a fixed cover and a movable cover; wherein, the stainless steel net is used to spread tea leaves, and the yellow light tubes installed on the side walls above the stainless steel net and the yellow light tubes installed in the lower layer provide yellow light irradiation; the hot air unit and the heat pump unit wither from the upper and lower parts of the tea leaves respectively; The moisture content online detection device includes a weight sensor, a rapid moisture detector, and a camera; the weight sensor is installed under the chassis of the entire withering tank body; four cameras are installed at the four corners of the top of the withering tank body chassis, fixed on the frame through brackets, and distributed diagonally; the rapid moisture detector 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 according to 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 during the withering process reaches the set target moisture content; 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 moisture content of tea leaves in areas corresponding to photos taken by different cameras, and to determine the degree of difference between the moisture content of tea leaves in the areas corresponding to the photos and the overall average moisture content calculated based on the weight measured by the 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 circulation fan I, and a dehumidification fan I; the heat pump unit is located on the left side of the withering tank body, and an air duct connecting the heat pump unit and the withering tank body is formed at the lower part between the heat pump unit and the withering tank body. A circulation fan II is installed on one side of the withering tank body in the air duct, and a dehumidification fan II is installed in the upper right corner directly in front of the withering tank body; the heat pump unit absorbs heat from the surrounding environment and outputs hot air, and the hot air enters the air duct, and the hot air in the air duct is sent to the lower layer of the chassis through the circulation fan II, and then the hot air rises and passes through the stainless steel mesh to wither the lower part of the tea leaves therein; the heat pump unit uses the heat absorbed from the surrounding environment to wither the tea leaves on the stainless steel mesh, which can save energy and reduce the emission of pollutants such as CO2, thereby achieving the effect of energy saving and emission reduction.

3. The Lichuan Red intelligent withering machine according to claim 1, characterized in that: The hot air unit includes a hot air blower, a hot air main duct, and a hot air branch duct; the hot air blower is installed 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 hot air main duct, the hot air main duct is equidistantly connected with hot air branch ducts, the hot air branch ducts are evenly provided with small holes with a diameter of 1 cm, the hot air main duct is installed above the stainless steel mesh, and is fixed to the chassis by a bracket; the hot air unit is used to transport 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 a vibration motor and a vibration spring. The four vibration springs are respectively located below the four corners of the stainless steel net, and the vibration motor is located in the middle position between every two springs. The vibrating leaf-turning mechanism makes the tea leaves in the stainless steel net more evenly dispersed, thereby withering more evenly.

5. The Lichuan Red intelligent withering machine according to claim 1, characterized in that: The adjustable inclination conveying mechanism includes a sleeve I, a right-angle lifter, a transmission shaft, a reduction motor, a rotating bearing I, a slider, a guide rail, a support, a sleeve II, and a rotating bearing II; the adjustable inclination conveying mechanism is located below the vibrating leaf-turning mechanism, and the left side of the adjustable inclination conveying mechanism is two right-angle lifters, and the right side is two supports, and the two right-angle lifters and the two supports are respectively located below a vibration spring; the bottom end of the right-angle lifter is installed on the corner at the bottom end of one side of the frame of the middle layer of the chassis, and the top end is the lifting end connected to the sleeve I, one end of the rotating bearing I is connected to the sleeve I through the rotating shaft, and the other end is connected to the slider, and the slider is connected to the guide rail, The guide rail is installed below the vibration spring; the bottom end of the support is installed on the corner at the bottom end of the other side of the frame of the middle layer of the chassis, and the top end of the support is installed with a sleeve II. One end of the rotating bearing II is connected to the sleeve II through a connecting shaft, and the other end is installed below the vibration spring; the reduction motor is installed in the middle between the two right-angle lifters and is connected to the right-angle lifters through a transmission shaft; different tea leaves have different flow characteristics, and the upper leaf conveying mechanism conveys the tea leaves to the stainless steel net, and then the movable cover is closed, and the stainless steel net is tilted 5° through the adjustable inclination conveying mechanism, and the vibrating leaf turning mechanism is started to vibrate the stainless steel net at the same time, so that the tea leaves are evenly spread on the stainless steel net.

6. The Lichuan Red intelligent withering machine according to claim 1, characterized in that: The main control system includes a control box, which is used for digital control of the withering process to determine whether the real-time overall average moisture content of the tea leaves during the withering process reaches the set target moisture content; a touch screen is installed on the surface of the control box, and a main control board is arranged inside it. The target moisture content, temperature upper limit and process program are input on the touch screen. The main control board includes a power supply module, a communication module, a WiFi communication module, a single-chip main control module, an IO control module, and a data storage module. The single-chip main control module includes a moisture content monitoring unit, and the moisture content monitoring unit includes a data preprocessing module and a data monitoring module.

7. The Lichuan Red intelligent withering machine according to claim 1, characterized in that: The single-chip microcomputer main control module is the core of the entire main control system and is responsible for the collection, processing, transmission of sensor data and the control of the withering process.

8. 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 of the tea leaves to the data preprocessing module, which preprocesses the pictures of the tea leaves to obtain color images and grayscale images, and then extracts features from the color images and grayscale images to obtain the morphological characteristic parameters, color characteristic parameters, and texture characteristic parameters of the tea leaves; The data monitoring module is loaded into the tea moisture content monitoring model, and the tea moisture content monitoring model outputs the real-time moisture content of the tea according to the real-time tea morphological characteristic parameters, color characteristic parameters, and texture characteristic parameters, so as to judge the difference degree of the tea moisture content in the areas corresponding to the photos taken by different cameras, and judge the difference degree between the tea moisture content in the areas corresponding to the photos and the overall average moisture content of the tea calculated based on the weight sensor; according to the formula: threshold = |tea moisture content in the area corresponding to the photos - overall average moisture content| / overall average moisture content ×%, if at this time, the threshold values ​​calculated based on the tea moisture content in the areas corresponding to the photos taken by the four cameras and the overall average moisture content are all ≤5%, indicating that the tea on the stainless steel net is withered evenly, then the withering process continues; If the threshold value calculated based on the moisture content of the tea leaves in the area corresponding to the photos taken by one or more cameras and the overall average moisture content is greater than 5%, indicating that the tea leaves on the stainless steel net are withered unevenly, the adjustable inclination conveying mechanism is started to tilt the stainless steel net by 5°, and then the vibration motor is started to vibrate the stainless steel net to make the tea leaves on it turn over and mix evenly; if the overall average moisture content reaches the set target moisture content, the heat pump unit, the hot air unit and the 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 inclination conveying mechanism is started again to make the right side of the withering frame lower than the left side. At the same time, the vibration motor is started to vibrate the stainless steel net to prompt 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 net are unloaded from the outlet, the withering of a new batch of tea leaves is restarted. Preferably, the photos of tea leaves on the stainless steel net taken by the camera are processed by the data preprocessing module to obtain the tea leaves' morphological characteristic parameters, color characteristic parameters, and texture characteristic parameters including rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrix. The seven types of data are input into the GA-BP neural network model, and the moisture content of the tea leaves is output. The degree of withering of the tea leaves is judged based on the output moisture content of the tea leaves, thereby ensuring that tea leaves with a similar degree of withering can be obtained in each withering process, standardizing the tea withering process and avoiding the deviation of subjective judgment on the degree of withering of tea leaves, which leads to different degrees of withering for each batch of tea leaves. The specific process of the data preprocessing module processing tea photos is as follows: (1) processing tea photos into color images and grayscale images; (2) extracting features from the color image and the grayscale image to obtain morphological feature parameters, color feature parameters, and texture feature parameters of the tea leaves, wherein the morphological feature parameters include the major axis, minor axis, rectangularity, circularity, compactness, diagonal length, and slenderness of the tea leaves; the color feature parameters are the color feature parameters of the tea leaves, including the first-order moments of red, green, blue, hue, saturation, and brightness; and the texture feature parameters include the contrast, second-order moment of angle, correlation, entropy, and inverse difference matrix of the tea leaves; (3) The built-in random forest model in the data preprocessing module was used to rank the importance of morphological feature parameters, color feature parameters, and texture feature parameters, and the important feature parameters of the withering process were screened out, namely rectangularity, compactness, greenness, saturation, contrast, correlation, and inverse difference matrix. Preferably, the tea moisture content monitoring model is a genetic algorithm optimized back propagation neural network model (GA-BP model), and the specific construction process is as follows: (1) the data preprocessing module calls the historical data in the data storage module, and the historical data is the existing tea photos of the withering process and the tea moisture content corresponding to the tea photos. Then, the existing tea photos of the withering process are processed to obtain the rectangularity, compactness, green, saturation, contrast, correlation and anti-difference matrix. Then, the obtained rectangularity, compactness, green, saturation, contrast, correlation and anti-difference matrix and the corresponding tea moisture content are normalized. Finally, the normalized data is divided into a training set and a verification set according to a ratio of 4:1; (2) the single-chip microcomputer main control module transmits the above-mentioned training set and verification 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 of the 7-10-1 structure to obtain the optimized BP neural network, wherein the iteration of the genetic algorithm The number of iterations 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 constructs the model 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 verified using the training set and the validation set, and the hyperparameters of the model are adjusted according to various indicators 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 indicators such as determination coefficient, mean square error, and mean absolute error; (5) The upper computer transmits the obtained GA-BP model to the microcontroller main control module, and it is loaded into the data monitoring module as a tea moisture content monitoring model. Preferably, the operation process of the intelligent withering machine is as follows: open the movable cover, and the fresh tea leaves are transported to the withering operation area through the upper leaf conveying mechanism; then, the adjustable inclination conveying mechanism tilts the stainless steel mesh by 5° so that the tea leaves slide in the stainless steel mesh, and at the same time, the vibration motor of the vibrating leaf-turning mechanism starts to vibrate, so that the tea leaves are evenly spread on the stainless steel mesh; after the tea leaves are evenly spread on the stainless steel mesh, the vibration motor stops vibrating, and then the adjustable inclination conveying mechanism restores the stainless steel mesh to a horizontal position; then, the movable cover is closed, the rapid moisture detector obtains the initial moisture content of the tea leaves, the weight sensor obtains the initial weight of the tea leaves, and the withering is completed. At the beginning of the process, the yellow light tube, heat pump unit, hot air unit and moisture content online detection device are started. The hot air output by the heat pump unit is withered from the bottom of the stainless steel mesh through the circulating fan, and the hot air output by the hot air unit is withered from the top of the stainless steel mesh to the bottom and top of the tea leaves respectively. The yellow light tubes on the upper layer of the chassis and the yellow light tubes on the lower layer wither the upper and lower parts of the tea leaves from the top and bottom of the stainless steel mesh respectively. Synchronously, the weight sensor obtains the real-time weight of the tea leaves, and the camera obtains the photos of the tea leaves in the stainless steel mesh every minute. All data are stored in the data storage module in real time. During the withering process, based on the real-time photos of the tea leaves, The tea moisture content monitoring model outputs the tea moisture content of the area corresponding to the photo. 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 photo with the overall average moisture content, the difference in the tea moisture content of the areas corresponding to the photos taken by different cameras can be determined, as well as the difference between the tea moisture content of the area corresponding to the photo and the overall average moisture content of the tea calculated based on the weight sensor. If the moisture content of the tea is uneven, the adjustable inclination conveying mechanism is first started to tilt the stainless steel net, and then the vibration is started. The motor excites the stainless steel net to make the tea leaves on the stainless steel net turn over and mix evenly. After mixing evenly, the excitation motor is turned off and the adjustable inclination conveying mechanism is started again to make the stainless steel net return to a horizontal position. When the overall average moisture content of the tea leaves in the withering operation area reaches the set target moisture content, the yellow light tube, the heat pump unit, the hot air unit, and the moisture content online detection device are turned off, and then the adjustable inclination conveying mechanism is started to tilt the stainless steel net, and then the excitation motor is turned on to vibrate the stainless steel net, so that the tea leaves flow from left to right and flow out from the outlet on the right side of the stainless steel net. When all the tea leaves on the stainless steel net are unloaded from the outlet, the withering of a new batch of tea leaves is restarted. The method for calculating the real-time overall average moisture content of the tea leaves is as follows: the moisture content monitoring unit calculates the current overall average moisture content Q2 according to the stored initial weight W1, initial moisture content Q1 and current real-time tea weight W2 according to the following formula and stores it in the data storage module; 9. A digital withering method for the Lichuanhong intelligent withering machine according to any one of claims 1 to 8, comprising the following steps: Step 1, start the Lichuanhong intelligent withering machine, open the movable cover, and transport the tea leaves to the stainless steel net through the upper leaf conveying structure, then start the adjustable inclination conveying mechanism, make the stainless steel net tilt 5° and then stop, then start the vibration motor to vibrate the stainless steel net, so that the tea leaves slide on the stainless steel net until the tea leaves are evenly spread on the stainless steel net; Then, the adjustable inclination conveying mechanism is started to restore the stainless steel mesh to a horizontal position, the loading is completed, and the movable cover is closed; Step 2, the single-chip microcomputer main control module sends an inquiry message to the weight sensor. After receiving the inquiry message, the weight sensor determines the initial weight of the tea, and then sends the initial weight of the tea to the single-chip microcomputer main control module. Similarly, after the rapid moisture detector determines the initial moisture content of the tea, and the temperature and humidity sensor determines the temperature and humidity in the withering operation area, the initial moisture content of the tea and the temperature and humidity in the withering operation area are sent to the single-chip microcomputer main control module. The withering operation area is the temperature and humidity about 10 cm above the stainless steel net. Specifically: the single-chip microcomputer main control module obtains the initial moisture content Q1 of the tea collected by the rapid moisture detector through the RS232-1 interface of the communication module, and obtains the initial weight W1 of the tea 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 single-chip microcomputer 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, and the WiFi communication module wirelessly transmits the data to the host computer. The single-chip microcomputer main control module sends the acquired data to the data storage module through the IIC-1 interface for storage; Step 3, input the target moisture content, temperature upper limit, temperature lower limit, humidity upper limit, humidity lower limit and process program on the touch screen, click "Start Withering", the withering process starts, start the yellow light tube, heat pump unit, hot air unit, moisture content online detection device, dehumidification fan, and circulation fan, the hot air output by the heat pump unit withers the lower and upper parts of the tea leaves from the bottom of the stainless steel mesh through the circulation fan, and the hot air output by the hot air unit withers the upper and lower parts of the tea leaves from the top and bottom of the stainless steel mesh respectively, the yellow light tube on the upper layer of the chassis and the yellow light tube on the lower layer wither the upper and lower parts of the tea leaves from the top and bottom of the stainless steel mesh respectively, synchronously, the weight sensor obtains the real-time weight of the tea leaves, the camera obtains the photos of the tea leaves in the stainless steel mesh every set time, and all data are stored in real time in the data storage module; When the current temperature obtained by real-time monitoring of the temperature and humidity sensor is higher than the input temperature upper limit value, the single-chip microcomputer main control module sends a signal to turn on the alarm, turn off the hot air blower, and turn off the heat pump unit. The alarm is turned on, and the hot air blower and heat pump unit are turned off. The dehumidification fan and the circulation fan continue to work to lower the temperature in the withering operation area to prevent the tea leaves from losing water too quickly. When the current temperature is not higher than the temperature upper limit value, the alarm is turned off and the hot air blower and heat pump unit are turned on. Similarly, when the current temperature obtained by real-time monitoring of the temperature and humidity sensor is lower than the input temperature lower limit value, the single-chip microcomputer main control module sends a signal to turn on the alarm, increase the power of the hot air blower, and increase the power of the heat pump unit. The alarm is turned on to increase the temperature of the hot air delivered by the hot air blower and the heat pump unit to increase the withering temperature. The temperature of the operation area is adjusted to prevent the tea from becoming astringent and withered due to low temperature. When the current temperature is not lower than the input lower temperature limit, the power of the heat pump unit and the power of the hot air blower return to their original power. When the current humidity obtained by the temperature and humidity sensor in real-time monitoring is higher than the input upper humidity limit, the main control module of the single-chip microcomputer sends a signal to increase the power of the dehumidification fan to improve the dehumidification efficiency of the dehumidification fan. When the current humidity is not higher than the upper humidity limit, the power of the dehumidification fan returns to its original power. Similarly, when the current humidity obtained by the temperature and humidity sensor in real-time monitoring is lower than the input lower humidity limit, the main control module of the single-chip microcomputer sends a signal to reduce the power of the dehumidification fan to reduce the dehumidification efficiency of the dehumidification fan. When the current humidity is not lower than the lower humidity limit, the power of the dehumidification fan returns to its original power. Step 4: The moisture content monitoring unit calculates the current overall average moisture content Q2 according to the stored initial weight W1, initial moisture content Q1, and current real-time weight W2 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 characteristic parameters of the withering process, and the specific process is: (1) processing the tea photos into color images and grayscale images; (2) extracting features from the color images and grayscale images to obtain the morphological characteristic parameters, color characteristic parameters, and texture characteristic parameters of the tea leaves, wherein the morphological characteristic parameters include the major axis, minor axis, rectangularity, circularity, compactness, diagonal length, and slenderness of the tea leaves; the color characteristic parameters are the color characteristic parameters of the tea leaves, including the first-order moments of red, green, blue, hue, saturation, and lightness; and the texture characteristic parameters include the contrast, angular second-order moment, correlation, entropy, and inverse difference matrix of the tea leaves; (3) using the built-in random forest model in the data preprocessing module to rank the morphological characteristic parameters, color characteristic parameters, and texture characteristic parameters, and screening out the important characteristic parameters of the withering process, which are rectangularity, compactness, green, saturation, contrast, correlation, and inverse difference matrix; Step 6, the single-chip main control module stores the rectangularity, compactness, green, saturation, contrast, correlation and anti-difference matrix obtained in step 5 into the data storage module; at the same time, the rectangularity, compactness, green, saturation, contrast, correlation and anti-difference matrix are input as input data into the tea moisture content monitoring model in the data monitoring module, and the tea moisture content monitoring model outputs the tea moisture content Q3, and compares Q3 with Q2 according to the following formula: Threshold value = |moisture content of tea leaves in the area corresponding to the photo Q3 - overall average moisture content Q2| / overall average moisture content Q2×%.

10. The digital withering method according to claim 9, characterized in that: If the threshold values ​​calculated based on the tea moisture content Q3 and the overall average moisture content Q2 of the area corresponding to the photos taken by the four cameras are both ≤5%, indicating that the tea leaves on the stainless steel net are withered evenly, the withering process continues; If the threshold values ​​calculated based on the moisture content Q3 of the tea leaves in the area corresponding to the photos taken by one or more cameras and the overall average moisture content Q2 are greater than 5%, indicating that the tea leaves on the stainless steel net are withered unevenly, the adjustable inclination conveying mechanism is started to tilt the stainless steel net by 5°, and then the vibration motor is started to vibrate the stainless steel net to make the tea leaves on it turn over and mix evenly; if the overall average moisture content Q2 reaches the set target moisture content, the heat pump unit, the hot air unit and the 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 inclination conveying mechanism is started again to make the right side of the withering frame lower than the left side. At the same time, the vibration motor is started to vibrate the stainless steel net to prompt 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 net are unloaded from the outlet, the withering of a new batch of tea leaves is restarted.

Images