基于轨迹预测与改进人工势场法的机械臂避障规划方法  

Dynamic Obstacle Avoidance for Manipulator Based on Trajectory Prediction and Improved Artificial Potential Field Method

作　　者：田应仲[1] 韩通 杨晓东[2] 倪雨嘉 李龙[1] TIAN Yingzhong;HAN Tong;YANG Xiaodong;NI Yujia;LI Long

机构地区：[1]上海大学机电工程与自动化学院 [2]上海市儿童医院

出　　处：《计量与测试技术》2025年第1期74-77,共4页Metrology & Measurement Technique

摘　　要：在人机共存的环境中,人可能会成为机器人执行任务过程中的动态障碍物,因此在机器人在运动过程中需动态的避障规划,从而避免机器人危害到用户安全。本文通过预测手臂的运动构建虚拟障碍物,用于改进传统人工势场法。实验表明:该方法可有效地进行动态避障,且规划的路径更短、更平滑。In the environment of human-machine coexistence,humans may become dynamic obstacles during the execution of tasks by robots.Therefore,dynamic obstacle avoidance planning is needed during the movement of robots to avoid endangering user safety.This article constructs virtual obstacles by predicting arm movements to improve traditional artificial potential field methods.The experiment shows that this method can effectively perform dynamic obstacle avoidance,and the planned path is shorter and smoother.

关 键 词：轨迹预测 改进人工势场法 动态避障 路径规划 

分 类 号：TP241[自动化与计算机技术—检测技术与自动化装置]