融合惯性导航与激光的服务机器人自定位研究

Autonomous Localization of Service Robots Based on IMU and Laser Data

针对家庭服务机器人在导航中的自定位算法进行分析,并通过引入惯性导航环节采集数据对激光和里程计数据进行补偿和融合,得到了更为精确的地图,从而实现更精确的定位。首先通过分析SLAM问题,得到在只有激光和里程计的信息情况,机器人位姿数据的误差较大。通过融合惯导数据信息,基于Rao-Blackwellized粒子滤波方法得到的更高概率的粒子点,进行更准确的位姿估计。同时也会对得到更准确的地图,有效提高鲁棒性。并在家具场景中进行了实验,通过实验可看出,经过数据融合改进的机器人导航系统中自定位的位姿数据,可以较好的完成导航的任务。

The analyzing of this paper is the self- localization algorithm of home service robot in navigation and compensates and integrates the laser and odometer data by introducing the inertial navigation link data to obtain a more accurate map to achieve more accurate positioning. This paper first analyzes the SLAM problem and obtains the error of the robot pose data in the case of only laser and odometer. By combining the IMU data,a higher probability particle point based on the Rao- Blackwellized particle filter method is used to obtain more accurate pose estimation. But also to get more accurate map,effectively improve the robustness. In this paper,experiments were carried out in the furniture scene,we can see that the self- location pose data of the robot navigation system improved by data fusion can accomplish the navigation task well.