基于改进的RRT*-Connect机械臂路径规划  

Dynamic Obstacle Avoidance Path Planning for Robotic Arm Based on Improved RRT-Connect

作　　者：张晨 袁宇浩 ZHANG Chen;YUAN Yu-hao(College of Electrical Engineering and Control Science,Nanjing Tech University,Nanjing Jiangsu 211816,China)

机构地区：[1]南京工业大学电气工程与控制科学学院,江苏南京211816

出　　处：《计算机仿真》2024年第11期458-464,554,共8页Computer Simulation

基　　金：国家自然科学基金项目(61773200)。

摘　　要：面对复杂地形和具有动态障碍物环境中机械臂避障路径规划的场景,RRT*-Connect算法存在效率低和路径差等问题,提出一种基于改进的RRT*-Connect机械臂动态避障路径规划算法。首先在RRT*-Connect算法中加入中间采样点,引导两棵随机树向中间采样点生长,另外对采样点加入可调约束参数,通过调整约束参数实现对目标偏向采样和随机采样空间的控制。在动态避障阶段,加入机械臂与动态障碍物距离监测算法,仅删除被障碍物遮挡和危险距离以内的路径,再次使用改进后的算法进行局部路径规划并与原始路径进行连接。最后,通过Matlab仿真验证了上述算法在复杂静态场景和动态环境中的路径规划效果,并在ROS平台中验证了所提算法在机械臂动态避障路径规划中的实用性。In the face of complex terrain and dynamic obstacle environments,the RRT*-Connect algorithm has problems such as low efficiency and path difference in obstacle avoidance path planning for robotic arms.A dynamic obstacle avoidance path planning algorithm based on an improved RRT*-Connect robotic arm is proposed.Firstly,the middle sampling point is added to the RRT*-Connect algorithm to guide the two random trees to grow toward the middle sampling point,and in addition,adjustable constraint parameters are added to the sampling point to achieve the control of target bias sampling and random sampling space by adjusting the constraint parameters.In the dynamic ob-stacle avoidance stage,the distance monitoring algorithm between the robotic arm and dynamic obstacles is added,and only the paths that are obscured by obstacles and within the danger distance are deleted,and the improved algorithm is used again for local path planning and connected with the original path.Finally,the path planning effect of this paper's algorithm in complex static scenes and dynamic environment is verified by Matlab simulation,and the practi-cality of this paper's algorithm in dynamic obstacle avoidance path planning of robotic arm is verified in ROS platform.

关 键 词：机械臂 动态避障 中间采样点 路径规划 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]