摘要
解决同时定位与地图构建(SLAM)问题是实现机器人自主导航的核心.目前,Rao-Blackwellized粒子滤波器(RBPF)是解决机器人同时定位与地图构建的有效方法.该方法在计算提议分布时,通常只考虑移动机器人的里程计信息,因此存在需要大量的采样粒子造成的计算量和复杂度增大的问题.本文提出一种改进算法,在计算提议分布时将机器人里程计信息和激光传感器采集的距离信息进行融合,有效地减少了所需粒子的数量并降低了滤波器预测阶段机器人位姿的不确定性.本文在机器人操作系统(robot operating system,ROS)平台上,使用配有URG激光器的Pioneer3-DX机器人进行了实验.结果表明,采用本文方法能够实时在线地创建高精度的栅格地图,为机器人在未知环境中的SLAM和导航提供了新途径.
Solving simultaneous localization and mapping （SLAM） problem is the core to realize the robot autonomous navigation. At present, Rao-Blackwellized particle filters （RBPF） is considered an effective mean to solve SLAM problem. By using RBPF, we only use the odometer information of the robot to calculate the proposal distribution which requires large-scale particle sampling, so the RBPF is still restricted by the calculation complexity. This paper provides a novel algorithm which fuses the robot＇s odometer information and laser sensor information in the proposal distribution computation for effectively reducing the particle sample scale and the uncertainty of robot pose estimation in the forecast period of particle filter. This approach, based on the robot operating system （ROS）, runs on platform Pioneer 3-DX robot equipped with a URG laser range finder. Experimental results show that this method successfully realizes the online real-time highprecision grid-map mapping, providing the robot for a new mean for simultaneous localization and mapping and navigation in unknown environments.
出处
《控制理论与应用》
EI
CAS
CSCD
北大核心
2015年第2期267-272,共6页
Control Theory & Applications
基金
国家自然科学基金项目(51075420)
国家科技部国际合作项目(2010DFA12160)资助~~
