基于ROS2.0四足抓取机器人的设计  

Design of Quadruped Grasping Robot Based on ROS2.0

作　　者：宣捷 王璐[1] 彭颖 XUAN Jie;WANG Lu;PENG Ying(School of Mechanical Engineering,Anhui University of Technology,Ma′anshan 243002,China)

机构地区：[1]安徽工业大学机械工程学院,安徽马鞍山243002

出　　处：《洛阳理工学院学报（自然科学版）》2023年第3期52-58,共7页Journal of Luoyang Institute of Science and Technology：Natural Science Edition

基　　金：安徽省高等学校科学研究重大项目(2022AH040361)。

摘　　要：设计了一种四足抓取机器人,使用Cyberdog四足机器人可以在复杂地形行走,选用机械臂MyCobot机械臂进行抓取作业,实现了YOLOV3结合点云数据获取特定目标的空间位置、机械臂远程控制、视觉SLAM、多点自主导航、自主避障等功能。针对DWA算法进行改进,结合YOLOV3目标检测算法以及深度相机点云数据,改善了传统导航算法存在的碰撞、阻碍动态障碍物的风险,使得机器人更好更安全地融入日常工作。A quadruped grasping robot is designed,which uses Cyberdog quadruped robots to walk in complex terrain and select a mechanical arm,MyCobot,for grasping operations.The robot has the following functions:YOLOV3 combines point cloud data to obtain the spatial position of a specific target,remote control of the robot arm,visual SLAM,multi-point autonomous navigation,autonomous obstacle avoidance,and other functions.It has improved the DWA algorithm,combined with YOLOV3 target detection algorithm and depth camera point cloud data,to improve the risk of collision and obstruction of dynamic obstacles in traditional navigation algorithms,making robots better and safer integrated into people′s daily work and production.

关 键 词：四足机器人 机械臂 目标检测算法 DWA算法 ROS2 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]