山东科学 ›› 2022, Vol. 35 ›› Issue (5): 130-139.doi: 10.3976/j.issn.1002-4026.2022.05.016

• 其他研究论文 • 上一篇    下一篇

基于机器人操作系统的液压机器人自主导航系统设计与实现

刘广亮(), 马争光, 朱琳, 肖永飞, 张艳芳*()   

  1. 齐鲁工业大学(山东省科学院) 自动化研究所 山东省机器人与制造自动化技术重点实验室,山东 济南 250014
  • 收稿日期:2021-11-02 出版日期:2022-10-20 发布日期:2022-10-10
  • 通信作者: 张艳芳 E-mail:liugl@sdas.org;365733524@qq.com
  • 作者简介:刘广亮(1971—),男,硕士,副研究员,研究方向为机器人系统与机器视觉。E-mail: liugl@sdas.org
  • 基金资助:
    济南市“高校20条”资助项目(2020GXRC035);山东省重点研发计划(2020CXGC010205)

Design and implementation of autonomous navigation system comprising a hydraulic driven robot based on robot operating system

LIU Guang-liang(), MA Zheng-guang, ZHU Lin, XIAO Yong-fei, ZHANG Yan-fang*()   

  1. Provincial Key Laboratory of Robot and Manufacturing Automation Technology, Institute of Automation, Qilu University of Technology (Shandong Academy of Sciences), Jinan 250014, China
  • Received:2021-11-02 Online:2022-10-20 Published:2022-10-10
  • Contact: ZHANG Yan-fang E-mail:liugl@sdas.org;365733524@qq.com

摘要:

在机器人两轮差速运动模型分析的基础上,根据运动模型的航迹演算公式可实时计算机器人轨迹姿态。基于机器人操作系统架构平台设计了机器人本体、机器人底座、驱动液压马达、导向轮、激光雷达等7个关节和连接的结构模型。各机器人关节坐标系变换关系经过Tf实时发布,用于计算机器人位置坐标信息。针对液压机器人没有配备编码器和视觉系统的不足,系统采用2D平面激光里程计模型(RF2O),通过建立连续激光扫描点对距离的流约束方程获得机器人平面运动估计,进而得出激光雷达的速度和机器人实际运行轨迹。设计了自主导航软件系统,实现了导航地图构建、定点任务导航和多机器人管理等功能,并对机器人定位导航精度进行了测试,分析对比多次指定位置和导航位置的数据差异,结果表明机器人导航定位精度达到设计要求。

关键词: 液压驱动机器人, 2D平面里程计, 自主导航, 机器人操作系统, 两轮差速驱动

Abstract:

In this study, a robot track equation is formulated to estimate the state and position of the robot based on a two-wheel differential drive model. Using the robot operating system(ROS) stack package, we design a robot structural model with seven joints and links, including a hydraulic robot body, a robot base, hydraulic motors, guide wheels, and lidar. The transformation of coordinate frames is shown over the Tf of ROS to obtain the coordinates of the robot's position. Because our robot does not possess an encoder and vision system, we use a two-dimensional (2D) planar lidar odometric model (RF2O). The planar motion estimation of the robot is performed by imposing the range flow constraint equation on consecutive lidar scan pairs. Then, the velocity of the lidar sensor and the real-time trajectory of the robot can be estimated. We also implement an autonomous navigation software system with functions such as navigation map building, navigation targeting, and multirobot management. We evaluate the accuracy of the system several times. Based on a comparative analysis of the difference in the data of assigned and navigated locations, the accuracy of robot navigation and position is found to satisfy the design requirements.

Key words: hydraulic driven robot, 2D planar odometry, autonomous navigation, robot operating system, two-wheel differential drive

中图分类号: 

  • TP242.3