Research On Plug Seedling Recognition And Control System Of Fully Automatic Transplanter

Point tray seedling transplanting is the main planting method in the production of facility vegetable and flower in our country.With the increase of planting quantity and the increase of labor cost,Manual Transplanting can no longer meet the production requirements.Most o f the existing transplanting machines are semi-automatic transplanting machines for artificial feeding seedlings,which can not fundamentally solve the problems of high labor intensity,l ow transplanting efficiency and low transplanting precision.Therefore,fully transplanting is the trend of development in the future.The research of automatic transplanter in our country is still in the exploratory stage.There are some problems in the developed automatic transpl anter,such as complex structure,low transplanting efficiency and high seedling leakage rate. Aiming at these problems,the overall structure and key mechanism of the automatic transpl anter are analyzed and designed.The recognition method of hole seedling based on machine vision is studied,and the control system of the automatic transplanter is designed.The main research contents and achievements are as follows:(1)To develop a fully automatic transplanter for potted seedlings.The seedling feeding m echanism,seedling picking mechanism,seedling picking manipulator,seedling feeding mecha nism and planting mechanism are designed by mechanical,electronic and pneumatic technolo gy.The transplanter operates coordinately under the control of the control system of the trans planter with the counting value of the transplanting counter as the basic time sequence.It can automatically complete the work of transporting seedling tray,grasping seedling,putting in an d planting seedlings.The transplanter adopts two-line working system.It consists of 6 manipu lators.The manipulator operates at the same time by taking seedlings at intervals.It can grasp 72 holes of seedlings in trays 12 times.(2)Design of Machine Vision Cave Recognition System for Cave Tray Seedlings,the har dware part of the system is mainly composed of light source,image acquisition device,indust rial computer and RS232 communication cable.Through the RS232 communication cable sys tem,the commands of the automatic transplanter control system can be received and the result s of hole identification can be transmitted.The software algorithm is mainly composed of ima ge preprocessing and hole recognition model based on convolution neural network.In the ima ge preprocessing stage,the hole seedling image is divided into 72 hole images,and the format conversion,normalization and data set construction are carried out.Through image segmenta tion,the larger hole seedling image recognition is converted into two-class recognition of hole image,that is,the rules of recognition model are reduced.Module ensures high resolution of recognition image.(3)A convolution neural network model for identifying seedlings was constructed.The re cognition model is a 12-layer network,including one input layer,four convolution layers,four pooling layers,one pavement layer,one full connection layer and one output layer.The conv olution layer uses 3×3 convolution cores,and the four convolution layers include 32,64,128 and 128 convolution cores,respectively.The pooling layer uses a 2×2 window and adopts the maximum pooling down-sampling method.The recognition model is trained by data and opti mized by function,so that the model has high recognition accuracy and generalization ability,and the recognition accuracy can meet the needs of transplanter hole seedling recognition.(4)The motion control algorithm of seedling and feeding mechanism is designed,which i ncludes: initial positioning control of seedling and feeding mechanism,keeping the actual pos ition of the mechanism in accordance with the coordinate position of the control system,elimi nating errors in the operation process and avoiding error accumulation;double closed-loop inc remental PID stepper motor control algorithm is designed to ensure the operation positioning accuracy of seedling feeding mechanism,and the establishment of stepper motor control syste m transmission.The recursive function model is simulated by MATLAB,and the S-curve ope ration control algorithm of seedling picking mechanism is designed to improve the stability an d reliability of servo motor and prolong its service life.(5)Research and development of automatic transplanter control system and pneumatic dri ve of manipulator,equipment selection of components in the system,design of control system power supply main circuit diagram,PLC wiring schematic diagram,stepper motor wiring dia gram,servo motor wiring diagram and pneumatic drive system schematic diagram,installatio n and debugging of control box,compiling software control system,realizing automatic trans planter institutions Coordination and control.The HMI information monitoring system is deve loped to collect,display and store the operation status information and operation information of the transplanter,so as to improve the information level of the transplanter.The operation te sts of seedling mechanism,feeding mechanism and manipulator were carried out.The proble ms were found through the tests,and the system design was improved.Finally,the design req uirements were achieved.