摘要： 在移动机器人同时定位与地图创建(SLAM)过程中，机器人本身位置不确定，其所处环境也不可预知，针对这些不确定性因素，应用贝叶斯规则作为理论基础，建立移动机器人SLAM算法的概率表示模型，通过扩展卡尔曼滤波器实现SLAM算法，并介绍一种激光雷达数据与特征地图的数据关联方法。实验结果表明，该方法为实现SLAM算法提供了一种有效可靠的途径。

关键词: 机器人, 概率论, 同时定位与地图创建, 扩展卡尔曼滤波器

Abstract: During the mobile robot Simultaneous Localization And Mapping(SLAM), the location is unknown and the environment round is also unpredictable. Aiming at these uncertain factors, the Bayes rule is as a theory foundation, the probability model of the mobile robot SLAM is founded, the realization process of the SLAM by Extended Kalman Filter(EKF) is discussed. A data association method between the laser radar and the feature map is introduced. Experimental results show this method is effective and reliable to realize SLAM.

Key words: robot, probability theory, Simultaneous Localization And Mapping(SLAM), Extended Kalman Filter(EKF)