天津职业技术师范大学学报

针对狭窄、紧凑的环境下使用传统转向方式的智能小车通过性较差的问题，以搭载ROS操作系统、IMU、里程计、单线激光雷达等传感器的麦克纳姆轮智能小车为研究对象，基于单线激光雷达SLAM，在陌生环境中构建地图，并在狭窄环境下进行导航实验。实验发现，由于膨胀半径参数过大，智能小车在自主导航时，经常发生停滞、通过性较差等情况，影响导航效果。经过参数调整优化，结果表明：在狭窄环境下，智能小车的通过性与敏捷性有显著提升，导航路线选取更加简洁，对空间的利用效率也更好。

To solve the problem of poor trafficability of smart cars with traditional steering modes in narrow and compact environments,the Mecanum wheel smart car equipped with ROS operating system,IMU,odometer,single line laser radar and other sensors is taken as the research object. Based on the Gmapping algorithm in single line laser radar SLAM(Simultaneous Localization and Mapping),the map is built in unfamiliar environments,and navigation experiments are conducted in narrow environments. The experimental result reveals that the intelligent car often stops and has poor trafficability during autonomous navigation due to the large expansion radius parameter,affecting the navigation effect. After adjusting and optimizing the parameters,the experimental results show that the trafficability and agility of the smart car have been significantly improved in a narrow environment,the navigation route selected is more concise,and the space utilization more efficient.