移动机器人平台低成本自主导航系统设计与实现  

Design and Implementation of Low-Cost Autonomous Navigation System for Mobile Robot Platform

作　　者：赵世鑫 王春彦 王佳楠 李春雨 ZHAO Shixin;WANG Chunyan;WANG Jianan;LI Chunyu(School of Aerospace Engineering,Beijing Institute of Technology,Beijing 100081,China)

机构地区：[1]北京理工大学宇航学院,北京100081

出　　处：《无人系统技术》2022年第4期50-62,共13页Unmanned Systems Technology

基　　金：国家自然科学基金(61803032,61873031)。

摘　　要：定位、建图与轨迹规划是实现移动机器人在未知环境自主运动的核心技术,也是目前的研究热点。首先,针对未知环境建图和定位问题,设计一种基于视觉的低成本导航方案。该方案利用深度相机获取未知环境信息,构建环境地图,并基于ORB-SLAM2算法完成自身定位。其次,为保障移动机器人运行平稳性,设计了一种A*优化方案。该方案结合轮式机器人运动特点,对传统A*算法的节点扩展过程进行优化,并通过改进搜索算法的碰撞检测部分实现地面移动平台的三维避障功能,同时基于B样条曲线完成轨迹平滑。最后,在移动机器人实验平台上完成自主导航系统部署并进行实验验证。实验结果表明,系统单次规划所需时间在10 ms以内,建图平均误差为0.14 m。此外,所提自主导航系统所需硬件设备简单,便于快速部署,具有广阔的应用前景。Localization,mapping and trajectory planning are the core technologies to realize autonomous movement of mobile robots in unknown environments,and are also the research hotspots at present,First,a low-cost navigation system based on single vision sensor is designed.This system uses unknown environmental data obtained from depth camera to build the unknown environment map and find its own location based on ORB-SLAM2.Second,in order to ensure the smoothness of mobile robot operation,an A*optimization scheme is designed.The node expansion process of the traditional A*algorithm is optimized by combining the motion characteristics of the mobile robot.Three-dimensional obstacle avoidance of the mobile robot platform is achieved by improving the searching algorithm in the collision detection part.The constraint conditions are constructed to further smooth the trajectory based on the B-spline curve.Finally,the autonomous navigation system is tested on a mobile robot platform and experiments.The results show that the time required for single planning is less than 10 ms,and the average error of mapping is 0.14 m.In addition,the hardware required by the proposed autonomous navigation system is simple,convenient for rapid deployment,and has a wider application prospect.

关 键 词：移动机器人 路径规划 轨迹优化 改进A*算法 SLAM ROS B样条曲线 

分 类 号：V279[航空宇航科学与技术—飞行器设计]