Design Of Citrus Picking Robotic Arm And Its Control System

As China's agricultural production level continues to increase,agricultural picking is an important part of agricultural production.Sichuan is the largest agricultural province in China and has the largest area of citrus production.At present,citrus picking is mostly carried out by manual picking,but the efficiency of manual picking is low and high cost is required.In order to improve the efficiency of citrus picking,reduce the labor intensity and realize intelligent picking,This paper designs the structure of the citrus picking robotic arm and its corresponding control system.In this paper,we first study the motion process of picking robot arm,establish the mechanical arm link coordinate system according to D-H parameter method,and carry out positive kinematics analysis through the description of spatial pose and coordinate system translation and rotation transformation,combined with the parameters of mechanical arm.A positive kinematic equation for picking the robot arm is established.Then,the inverse kinematics analysis is performed on the picking robotic arm,and the inverse kinematics solution is obtained by algebraic method.In order to improve the stability of the mechanical arm during the picking process,This paper designs a support base for the robotic arm to prevent the picking process from failing due to excessive weight of the robotic arm during the picking process.According to the growth environment of citrus and the mechanical properties of citrus itself,the pressure deformation characteristics of citrus fruits and the cutting characteristics of citrus fruit stalks were studied.Based on this,the design requirements for the picking end effector are proposed,and an picking end effector model with high picking success rate is designed.The detection algorithm YOLO-V3 algorithm is used to detect and locate the collected citrus images.The YOLO-V3 citrus detection network is built and trained by Tensorflow framework.After networks model is trained its parameters are obtained.Finally,the model detects the forward prediction of the citrus image and visualizes the middle layer feature map in the forward prediction process.Select the driving method according to the movement requirements of the picking robot arm,the STM32 is used as the main controller.Using STM32 timer to complete the control algorithm design and programming of multi-way servo...