基于ROS的室内导航机器人设计研究  

Design and research on indoor navigation robot based on ROS

作　　者：张富春[1] 李铭田 黄豪豪 ZHANG Fuchun;LI Mingtian;HUANG Haohao(School of Physics and Electronic Information,Yan’an University,Yan’an 716000,China)

机构地区：[1]延安大学物理与电子信息学院,陕西延安716000

出　　处：《延安大学学报（自然科学版）》2023年第1期26-30,38,共6页Journal of Yan'an University：Natural Science Edition

基　　金：延安市科技创新团队项目(2017CXTD-01);延安大学研究生教育教学改革研究项目(YDYJG2019016)。

摘　　要：采用基于机器人操作系统(ROS)的设计方案开展了机器人在自主导航中的应用研究,设计了一款集定位、SLAM建图、路径规划、驱动控制于一体的机器人综合系统。系统利用激光雷达扫描获取环境实时数据和里程采集信息,通过Gmapping算法对数据进行融合,以及采用自适应蒙特卡洛定位(AMCL)方法实现了环境地图绘制与定位。通过matlab对路径规划算法进行对比研究发现,路径规划任务中Djakarta算法更适应在室内多障碍物场景。而实验结果显示在模拟环境中利用Dijkstra算法和DWA算法进行避障更加有效。仿真实验结果也表明,该系统可构建高精度环境地图,并对机器人进行定位和导航,实现了在实时路径规划中成功避开障碍物到达目标点。以上研究表明在室内定位和导航任务中该机器人系统框架具有良好的可行性和稳定性。Based on the design scheme of robot operating system(ROS),the application research of robot in autonomous navigation was carried out,and a robot integrated system integrating positioning,slam mapping,path planning and drive control was designed to complete the real-time information collection of environment by lidar,odometer and camera for data fusion,and the Adaptive Monte Carlo positioning(AMCL)method was adopted.The mapping and positioning of the environment were realized by using the data obtained by lidar scanning and odometer information.Through the comparative study of path planning algorithms with MATLAB,it is found that Djakarta algorithm is more suitable for indoor multi obstacle scenes in path planning tasks.The experimental results show that Dijkstra algorithm and DWA algorithm are more effective in obstacle avoidance in the simulation environ⁃ment.The simulation results show that the system can build a high-precision environment map and locate the robot,and successfully avoid obstacles to reach the target point in real-time path planning.The above research proves that the robot framework is feasible in indoor positioning and navigation tasks.

关 键 词：机器人操作系统(ROS) Gmapping 自适应蒙特卡洛定位 路径规划 自主导航 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]