基于激光雷达的室内机器人的建图方法

Drawing Method of Indoor Robot Based on Lidar

近年来,移动机器人在各个领域的应用日益广泛,与之相关的技术已在国内外机器人领域掀起一股研究热。对环境的认识和定位从而实现自主导航是移动机器人智能化的重要标志和特征。在完全未知环境下的即时定位与地图构建(SLAM)也一直是移动机器人领域的研究重点。由于激光SLAM具有不易受环境影响、测量距离远且测量精度高、成本低的优点,因此本文采用改进的Hector_SLAM算法来完成室内环境下地图的创建。本文中所用系统主要包括由硬件系统和软件系统两部分。首先通过硬件系统对环境进行感知,然后提取周围环境特征信息,最后通过软件系统完成二维环境地图的实时创建。在得到环境地图的基础上,采用改进的卡尔曼滤波法对初始数据进行优化处理,以获得更为精确的环境地图。实验结果表明,本文所用系统成本低但性能好,能较准确的构建二维环境地图,并能成功使用在小型室内移动机器人的视觉导航中。

In recent years, mobile robot has been widely used in various fields, and the related technology has set off a research heat in the field of robot at home and abroad. It is an important sign and feature of intelligent mobile robot to realize autonomous navigation by understanding and locating the envi-ronment. Real time localization and mapping (SLAM) in completely unknown environment has al-ways been the research focus in the field of mobile robot. Because laser slam is not easy to be af-fected by the environment, the measurement distance is long, the measurement accuracy is high and the cost is low, so the improved Hector is used in this paper SLAM algorithm to complete the indoor environment map creation. The system used in this paper mainly consists of hardware sys-tem and software system. Firstly, the environment is perceived by the hardware system, and then the feature information of the surrounding environment is extracted. Finally, the real-time creation of two-dimensional environment map is completed by the software system. Based on the environ-mental map, the improved Kalman filter is used to optimize the initial data to obtain more accurate environmental map. The experimental results show that the system used in this paper has low cost but good performance, and can accurately construct two-dimensional environment map, and can be successfully used in the vision navigation of small indoor mobile robot.