移动机器人同步定位与地图构建研究进展

Simultaneous Localization and Mapping for Mobile Robots

同步定位与地图构建(S im u ltaneous loca lization and m app ing,SLAM)作为能使移动机器人实现全自主导航的工具近来倍受关注。本文对该领域的最新进展进行综述,特别侧重于一些旨在降低计算复杂度的简化算法的分析上,同时对它们进行分类,并指出其优点和不足。本文首先建立了SLAM问题的一般模型,指出了解决SLAM问题的难点;然后详细分析了基于EKF的一些简化算法和基于其他估计思想的方法;最后,对于多机器人SLAM和主动SLAM等前沿课题进行了讨论,并指出了今后的研究方向。

Simultaneous localization and mapping （SLAM） algorithm plays an important role on enabling the fully autonomous navigation. This paper surveys the latest progress of SLAM especially on simplification methods to reduce the complexity. These methods are classified and their advantages and drawbacks are pointed out. Firstly, the general model of SLAM problem is constructed and main difficulties to solve the SLAM problem are presented. Then, some simplification algorithms based on extended Kalman filter （EKF） and other estimation ideas are introduced. Finally, some front subjects in multi-robot SLAM and active SLAM problems, etc. are discussed, and some other topics for future research are indicated.