一种改进的AMCL机器人定位方法  

An improved adaptive Monte Carlo localization method for robot

作　　者：王宁 王坚 李丽华[2] WANG Ning;WANG Jian;LI Lihua(School of Geomatics and Urban Spatial Informatics,Beijing University of Civil Engineering and Architecture,Beijing 102616,China;School of Land Science and Technology,China University of Geosciences,Beijing 100083,China)

机构地区：[1]北京建筑大学测绘与城市空间信息学院,北京102616 [2]中国地质大学土地科学技术学院,北京100083

出　　处：《导航定位学报》2019年第3期31-37,共7页Journal of Navigation and Positioning

基　　金：国家重点研发计划项目(2016YFC0803103)

摘　　要：针对移动机器人技术中的蒙特卡罗定位算法存在计算量大、实时处理能力差、粒子退化等问题,提出一种基于AMCL算法的自适应定位模型:改进ROS中的move_base节点的路径规划功能,并加载学院楼CAD地图实现路径规划;基于ROS中slam_gmapping节点在未知环境下创建OGM地图。仿真实验结果表明:AMCL定位模型能够减小定位误差,提高路径规划的准确性和实时性;采样部分利用重采样与KLD采样交替进行的方法,根据粒子在状态空间的分布情况,能够实时在线调整粒子数,有效减少算法计算量。Aiming at the problems of large computation,poor real-time processing ability and particle degradation in Monte Carlo location algorithm,the paper proposed an adaptive positioning model based on AMCL algorithm:the path planning function of move_base node in ROS was improved,and the CAD map of the college building was loaded to realize the path planning;an OGM map was created in an unknown environment based on slam_gmapping node in ROS.Simulational result showed that AMCL positioning model could reduce the positioning errors and improve the accuracy and real time of path planning;and the alternate sampling between resampling and KLD sampling could adjust the number of particles online in real time so that the calculation of the algorithm could be reduced effectively according to the distribution of the particles in the state space.

关 键 词：自适应蒙特卡罗定位模型 机器人操作系统 静态地图 移动机器人 路径规划 

分 类 号：P228[天文地球—大地测量学与测量工程]