基于改进Informed RRT^(*)算法的大棚采摘机械臂路径规划  

Path planning of a robotic arm for greenhouse picking based on improved Informed RRT^(*)algorithm

作　　者：郑泽斌 郑书河[1,2] 翁武雄 林添良 郭雷 ZHENG Zebin;ZHENG Shuhe;WENG Wuxiong;LIN Tianliang;GUO Lei(College of Mechanical and Electrical Engineering,Fujian Agriculture and Forestry University,Fuzhou,Fujian 350002,China;Fujian University Engineering Research Center for Modern Agricultural Equipment,Fuzhou,Fujian 350002,China;Fujian Key Laboratory of Green Intelligent Drive and Transmission for Mobile Machinery,Xiamen,Fujian 361000,China)

机构地区：[1]福建农林大学机电工程学院,福建福州350002 [2]福建省现代农业装备高校工程研究中心,福建福州350002 [3]福建省移动机械绿色智能驱动与传动重点实验室,福建厦门361000

出　　处：《福建农林大学学报(自然科学版)》2025年第2期279-288,共10页Journal of Fujian Agriculture and Forestry University:Natural Science Edition

基　　金：福建省科技厅引导性项目(2022N0009);福建省移动机械绿色智能驱动与传动重点实验室开放基金课题项目(GIDT-2023XX);福建农林大学茶全产业链创新与服务体系建设项目(K1520005A05)。

摘　　要：【目的】提出一种机械臂路径规划算法,以解决多自由度机械臂在大棚采摘作业中路径规划速度慢、路径成本高等问题,为采摘机械臂高效作业提供依据。【方法】基于Informed RRT^(*)机械臂路径规划算法,引入自适应目标偏置策略,结合贪婪思想的双向非优化路径搜索方式,增加动态概率节点拒绝策略。提出一种动态拒绝的知情RRT^(*)贪婪连接算法(dynamic rejection informed RRT^(*) greedy connect algorithm, DR-IRRT^(*)-GC),并通过Matlab多障碍物场景、ROS(机器人操作系统,robot operating system)机械臂场景的仿真对比试验进行验证。【结果】DR-IRRT^(*)-GC算法在路径规划的过程中具有运行速度快、路径成本低、路径规划成功率高等优势。与Informed RRT^(*)算法相比,DR-IRRT^(*)-GC算法在首次路径搜索用时上缩短了97.36%,路径规划成功率提高了283.33%。【结论】DR-IRRT^(*)-GC算法具有较强的实用性。【Objective】A robotic arm path planning algorithm was proposed to address issues such as slow path planning speed and high path cost for multi⁃degree⁃of⁃freedom robotic arms in greenhouse picking tasks,to provide basis for the efficient operation of picking robots.【Method】Based on the Informed RRT^(*)robotic arm path planning algorithm,an adaptive target bias strategy was in⁃troduced,combined with a bidirectional unoptimizable path search method in combination with greedy principle,along with a dy⁃namic probability node rejection strategy.A dynamic rejection informed RRT^(*)greedy connect algorithm(DR⁃IRRT^(*)⁃GC)was pro⁃posed.The approach was validated through comparative simulations in Matlab multi⁃obstacle scenarios and robotic arm scenarios in robot operating system(ROS).【Result】DR⁃IRRT^(*)⁃GC algorithm demonstrated advantages in path planning process with faster op⁃erating speed,lower path cost,and higher success rate in path planning in simulation test.Compared with the Informed RRT^(*)algo⁃rithm,the DR⁃IRRT^(*)⁃GC algorithm cut down the time for the initial path search by 97.36%,and achieved higher success rate in path planning by 283.33%.【Conclusion】The DR⁃IRRT^(*)⁃GC algorithm demonstrates strong practicality.

关 键 词：采摘机械臂 路径规划 改进Informed RRT^(*)算法 贪婪思想 动态概率 

分 类 号：S23[农业科学—农业机械化工程] TP241[农业科学—农业工程]