大部分现有的移动机器人地图构建方法都是基于静态环境的假设,而实际应用中移动机器人的工作环境是随时间变化的。综述了动态环境中移动机器人地图构建的最新研究进展,介绍了基于地图、基于运动和基于跟踪的检测动态障碍物的各种方法,分析比较了动态环境中移动机器人过滤运动障碍物传感器观测信息和结合运动障碍物传感器观测信息构建环境地图的主要方法,并总结了各种方法的优缺点。探讨了动态环境中移动机器人地图构建存在的难点问题,并展望了该领域的研究方向。
Most existing map building methods for mobile robots are based on the assumption of static environments, while the working environments of mobile robots in real applications change over time. Advances on map building with a mobile robot in dynamic environments are overviewed. Various methods of detecting moving obstacles for a mobile robot such as map-based, motion-based and track-based methods are introduced. Approaches to map building in dynamic environments with a mobile robot which filter out and integrate sensor observations of moving obstaclos are analyzed and their advantages and disadvantages are further summarized. The difficult issues of map building with a mobile robot in dynamic environments are discussed and future research of this field is also proposed.