采摘机械臂路径规划的GA-RRT算法研究  

GA-RRT Algorithm for Path Planning of Picking Manipulator

作　　者：李炳彤 李海虹[1] Li Bingtong;Li Haihong(College of Mechanical Engineering,Taiyuan University of Science and Technology,Taiyuan 030024,China)

机构地区：[1]太原科技大学机械工程学院,太原030024

出　　处：《农机化研究》2025年第8期25-32,共8页Journal of Agricultural Mechanization Research

基　　金：山西省重点研发计划项目(202102140601015)。

摘　　要：面对采摘作业的复杂环境,提出了一种终点区域RRT(Goal Area RRT,GA-RRT)算法,以提高路径生成的效率并降低路径成本。根据环境系数确定初始步长与终点区域,当拓展节点进入终点区域后,随机点生成范围缩小至终点区域,同时调整步长;然后,在此基础上引入目标概率偏向方法,提高路径搜索效率;最后,对生成的路径进行简化节点处理以减少路径代价,并使用三次B样条方法平滑路径。仿真实验结果表明:二维环境下,GA-RRT算法相较于RRT、RRT-Connect算法,耗时缩短85.15%、29.86%,路径代价减少19.18%、18.26%;机械臂仿真环境下,与引入目标概率偏向方法的RRT算法进行比较,耗时缩短54.70%,路径代价减少51.59°。利用IRB120机械臂实验平台,验证了算法的可行性。A GA-RRT algorithm was proposed to improve the efficiency of path generation and reduce the path cost for aiming at the complex environment of picking operation.The initial step size and goal area was determined by the environment coefficient.The generation range of random point was narrowed to goal area and the step size is adjusted after the expansion point entered goal area.The target guidance probability was added to improve the efficiency of path planning.Simplifying points and cubic B-spline curve were used to reduce the path cost and smooth the path.The simulation showed that,in 2D space,the GA-RRT algorithm reduced the search time by 85.15%and 29.86%,and it reduced the path cost by 19.18%and 18.26%,comparing with the RRT and RRT-connect algorithms.The GA-RRT algorithm reduced the search time by 54.70%and reduced the path cost by 51.59°,comparing with the RRT algorithm that incorporating the target guidance probability in manipulator simulation.The feasibility of the algorithm was verified by an IRB120 manipulator.

关 键 词：采摘机械臂 路径规划 快速搜索随机树算法 终点区域 

分 类 号：S225[农业科学—农业机械化工程] TP241[农业科学—农业工程]