基于多传感器融合的自主导航ROS智能机器人
Autonomous Navigation ROS Intelligent Robot Based on Multi-Sensor Fusion
朱明超 1吕镇炯 2邓民翰 2苏秉华 1郑裕源 2陈骏白2
作者信息
- 1. 北京理工大学珠海学院信息学院,广东 珠海 519088;北京理工大学(珠海)大湾区创新研究院,广东 珠海 519088
- 2. 北京理工大学(珠海)大湾区创新研究院,广东 珠海 519088
- 折叠
摘要
ROS智能机器人在未知环境或狭小空间下的自主导航存在环境地图精度低、信息易丢失等问题,导致难以实现快速实时导航.设计了基于多传感器融合的自主导航ROS智能机器人,通过激光雷达与深度相机采集机器人周围的环境信息,通过激光SLAM与RGB-D视觉算法实现多传感器融合,进而完成机器人室内同步定位与地图构建,实现机器人自主导航.实验表明,多传感器融合地图构建具有高定位精度和丰富信息,所建地图与实际环境匹配度较高,提升了机器人的环境感知和导航灵活性,使系统能准确自主地完成导航任务.
Abstract
In unknown or confined spaces,autonomous navigation of ROS-based intelligent robots presents challenges such as low map accuracy and information loss,hindering real-time navigation.This paper introduces a design focused on a multi-sensor fusion-based autonomous navigation ROS intelligent robot.It collects environmental data using both LIDAR and depth cameras,employing a fusion of LIDAR-based SLAM and RGB-D visual algorithms to achieve multi-sensor fusion,consequently accomplishing simultaneous localization and mapping(SLAM)indoors for robot autonomous navigation.Experimental results demonstrate that the multi-sensor fusion mapping offers high localization accuracy and rich information.The constructed maps exhibit high correspondence with the actual environment,enhancing the robot's environmental perception and navigation flexibility,enabling the system to autonomously navigate accurately.
关键词
多传感器融合/自主导航/ROS系统/深度相机/激光雷达Key words
multi-sensor fusion/autonomous navigation/ROS system/depth camera/LIDAR引用本文复制引用
基金项目
广东省"天临空地海"复杂环境智能探测重点实验室(2022KSYS016)
广东高校省级重点平台及重大科研项目特色创新类项目(本科)(2019KTSCX216)
出版年
2024
NETL
NSTL
万方数据