Low-loss Picking Mechanism And Equipment Development Of High-quality Tea Shoots

Tea is an important cash crop in China.With the development of economy people’s demand for premium tea is increasing.The aging population and the loss of young rural laborers have resulted in unharvested tea leaves.Therefore,the price of premium tea is rising year by year,which seriously affects the development of premium tea in China.Most of the varieties of premium tea are made from the tender shoot of tea leaves.Therefore,the development of equipment for picking tea shoots is the optimal solution to the current problem.When the tea shoots are picked,they are very delicate leading to damage easily caused during harvesting affecting the processing quality.And the small space where the shoots grow causes the picking machine to not have enough picking space.The second problem that makes mechanized picking operations difficult to achieve is the small size of tea shoots with similar color and shape of leaves and shoots.These issues also constrain identification,location and picking from being accurate and fast.In recent years,in order to adapt to mechanized picking,the tea plantation has gradually developed towards standardization.Therefore,this study was conducted in a standard tea plantation,with tea shoots as the research object,and focused on the key technology of mechanized low-damage harvesting of tea shoots.This thesis uses mechanical measurements,simulation experiments and theoretical analysis to construct a mechanical model for manual flexible picking of tea shoots.After that,by establishing a tea flexible analysis platform and introducing deep learning theory,build an intelligent identification system for tea tender tips under complex environment.The adaptive leveling work platform is designed according to the topography of the tea growing terrain and the tea growing state.The main research of the thesis has the following parts.(1)Study on the mechanics model of manual flexible picking.In order to solve the problem of easy damage to tea shoots during mechanical picking,this thesis combines the mechanical test bench and microscopic test bench to determine the mechanical model of manual picking of tea shoots.Firstly,the mechanical exertion process and characteristics of manual picking of tea shoots were analyzed,and the mechanical model of finger flexible gripping shoot picking was established.The relationship between mechanical properties and structural deformation of tea stalks was obtained by microstructural observation and mechanical experimental analysis.Determining the mechanics flow of manual flexible harvesting of Shucha early tea shoots by joint loading test.(2)Research on flexible harvesting technology of shoot tips.In order to verify the validity of the manual picking mechanics model,a pinch-cut combination tea shoot flexible picker was designed.Firstly,it imitates the pinching,lifting and picking action flow of manual picking,and combines flexible clamping and disc-shaped cam to realize the continuous operation of pinching and cutting,sending,throwing and throwing of tea shoots.The precise and stable picking operation of the picker is achieved through the optimization of the cam curve parameters combined with the S-curve speed control function.The optimal picking depth and height were finally determined by field response surface testing to be 10 mm,with a picking accuracy angle of 0 °.The actual test comparison found that the quality of the shoots picked by flexible clamping was consistent with that of the hand-picked shoots.(3)Research on continuous flexible harvesting technology for tea shoot.In order to further improve the picking speed and efficiency,and to enhance the reliability of the field tea garden use,this thesis designs a continuous shoot flexible picker.Firstly,the picking effect of various combinations of picking tools was analyzed by explicit kinetic comparison.Then the mechanical parameters of the clamping components were determined by analyzing the rheological characteristics of the tea shoot tip clamping process.Based on the conclusions of the mechanical analysis,the design of a continuous flexible clamping component suitable for tea shoot picking is guided.In order to keep the tea shoots in an upright position when feeding and to prevent bending during the picking identification process which leads to the reduction of picking accuracy,this thesis combines the tea stem bending moment equation to determine the forward and rotation parameters of the picker when feeding the tea shoots.After that,the tea leaves are slit into the picker through the design of the barring clamping mechanism.The picking response surface test was conducted by the shoot tip picking test bench.The continuous low loss picker can meet the continuous sequential picking of tea shoots.The optimal operating parameters of the picker were a rotation speed of 0.135m/s,a picking margin of 6mm and a forward speed of 0.04m/s.(4)Research on two-dimensional perspective recognition positioning and picking device control technology.In order to solve the problems of accurate identification of tea shoots and precise positioning of picking points in field tea gardens,this study proposes a semi-closed two-dimensional perspective recognition model combined with Yolov5 recognition algorithm,followed by the prediction of stem growth curve using two-dimensional skeleton information to determine the coordinates of picking points.Firstly,This thesis establish the tea shoot data set with multiple disturbance environments,multiple angles and distances,and then obtain the optimal Yolov5 recognition model by training the model.This study then combines the identification of the shoot skeleton information in the local area to determine the stem growth curve function by indexing the number of pixel points row by row,after which this study determines the exact picking point coordinates based on the fixed distance downward from the intersection.In order to avoid the misclamped blades and vibration during the operation of the picker leading to the reduction of recognition accuracy,this study combines path planning and S-curve velocity control function to achieve accurate control of the picker motion.The shoot picking test bench shows that the error range between the actual picking point position and the model determined position is within ± 2mm,which can meet the demand for accurate positioning of tea shoot tips.(5)Research on picking leveling technology.In order to make the picker tool plane always in the horizontal plane and ensure the accuracy of picking operation,this thesis designs an adaptive leveling platform based on the parallel mechanism.This study proposes an angle adjustment method based on PID algorithm combined with inertial guidance sensor,after which the leveling operation is realized according to the function relationship between motor rotation angle and platform rotation angle.The actual test shows that the error value of the adjustment angle of the two axes is within ±0.5°.The test showed that the picking success rate of the actual tea garden picking equipment was 83.6%.The research of rheological analysis and two-dimensional viewpoint identification and positioning algorithm realized the low loss harvesting of tea shoot tips.This study can provide a theoretical basis and reference for the design of low-damage and continuous picking equipment for tea shoots.