芒果采摘机器人设计与试验

打开文本图片集

中图分类号： T P3 9 1 文献标志码：A

Abstract：MangoisoneofthharacteristicfruitsinthtropicalregionTheharvestingofmangoesisprimarilydonemanualyhich haspresentedisuessuchashighlaborintensityandelevatedcosts.Therehasbeenanurgentneedtodevelopmechanizedharvesting technologiesformangoes.Inthisstudy，amangopickingrobotisdesignedbasedonmangoagronomic growthinformation，whichcan automaticallyidentifythefruitandcompletethepickingandcollection.ByusingYOLOv5formangofruittargetrecognitionand combiningthedepthinformationprovidedbythedepthcameraD435i，thethree-dimensionalinformationofthepickingpointis obtained;Ashear-clampintegratedend-efectorisdesignedtoefectivelyreducethemechanicaldamagetothemango；Amango pickingrobotcontrolsystemisdevelopedbasingontheROSsystem；Testing experimentswereconductedinlaboratorynviroment， andtheresult of theexperiments showed that theaverage time for the mango picking robot to recognize thefruit was ，the average picking time was 19.8 s，and the picking rate was 9 0 . 8 % ，which verified the control accuracy and stability of the picking robot.Thisstudyprovidestechnicalsupportformangopickingoperationandhelpstopromotetheinteligentdevelopmentof agriculture.

Keywords：mango picking;robot;end-effector;ROS system

中国是世界上第二大芒果生产国和第一大消费国[1]，芒果生产已成为我国热带和亚热带地区农业经济的主要支柱[2]，海南省地处热带和亚热带地区，是我国芒果主产区之一[3]，截至2023年海南芒果种植面积为6.54万 ，产量达到91.04万t，在海南农业经济占据重要地位[4]，但是目前我国芒果的采摘以人工采摘为主，采摘效率低，劳动强度大[5]，严重影响芒果标准化、产业化的发展。（剩余11229字）