苹果采摘机器人的视觉伺服定位研究

摘要

This paper describes the design process and feasibility of apple picking robot.The robot arm is 4 DOF,which is RRRP.The waist,major arm,minor arm are controlled by the Delta servo motion control system,and the Festo pneumatic servo motion control system is used to control the pusher.The end effector installed on the end of the pusher grabs the apple identified and located by the camera.Image processing procedures are wrote by OpenCV to get the difference value between the Apple's central point and the image central point,and the manipulator controls the robot arms to move toward the apple according to the difference value.When the camera is aimed at the apple,the Festo pusher moves toward to apple then stops,and the end effector graps the apple.The feasibility of the design is verified by the testing.The results show that the apple picking robot can realize the automatic picking of apple.%文中研究了基于视觉的苹果采摘机器人伺服控制系统的定位精度问题.机器人本体采用RRRP四自由度机械臂,利用台达伺服运动控制系统完成腰部、大臂、小臂三个自由度的运动控制,采用Festo气动伺服运动控制系统完成第四个自由度气动推杆的运动控制,并在气动推杆末端安装末端执行器用以完成苹果的抓取.实验结果表明,该苹果采摘机器人能够自动准确对准目标苹果中心点,腰部与小臂关节的定位误差小于20个像素,Festo气动推杆的定位误差小于0.5mm,这两项定位精度指标满足实际苹果采摘的要求.