【作者】 李鹏；

【导师】 杨名利；

【作者基本信息】 哈尔滨工业大学 ， 控制科学与工程， 2011， 硕士

【摘要】 为了实现移动机器人对家庭环境地图构建和自身的精确定位,利用智能空间的概念,可以将移动机器人同环境中的各种设备进行无线互连,扩展了移动机器人的感知能力,通过智能空间中的设备计算移动机器人携带传感器采集到的数据,减轻了移动机器人的负担。本文旨在以家庭环境为背景,构建智能空间,使得机器人能够快速准确地完成环境地图的构建和自身的精确定位。并在此前提下,进行路径规划,使移动机器人完成各项服务。主要研究内容如下:首先,给出了研究服务机器人的意义,介绍了智能空间这一新兴概念,并给出了智能空间构建的基本技术要求。在构建的智能空间下,利用图像处理的方法,得到二维环境地图。其次,对移动机器人的同时定位与建图技术进行了介绍,论述了该技术应用过程中所遇到的一些难点,给出了扩展卡尔曼滤波方法和Rao-Blackwellised粒子滤波方法解决移动机器人同时定位与建图的一般过程,通过将机器人同时定位与建图过程与智能空间技术相结合,采用了一种混合式的同时定位与建图方法,结合了Rao-Blackwellised粒子滤波和扩展卡尔曼滤波方法的优点,避免了两种方法的缺点,利用Rao-Blackwellised粒子滤波进行局部地图的创建,然后融合到全局地图中,通过与扩展卡尔曼滤波方法的仿真结果进行比较。最后,在智能空间构建地图的前提下,采用了人工势场法进行路径规划。该方法具有反应速度快,安全可靠的优点,可以保证移动机器人在环境中每个坐标点都有对应的运动方向,在家庭环境中,具有一定应用价值。更多还原

【Abstract】 In order to make the mobile robot mapping and locate itself more accuracy, intelligent space should have flexibility and scalability so that the mobile robot can navigate easier.This paper is mainly aiming at home environment that is a dynamic and non-structured. We can treat the lab as a home, based on the lab, for the mobile robot, there are some technologies researched in the intelligent space as follows:First of all, this paper introduced the background of resarch and the concept of intelligent space, gave the reserch status at home and abroad in recent years. Besides, using the intelligent space technologies, we present that how to structure the intelligent sapce and give the basic elements. Based on it, we can get the 2d environment map.Second, in the intelligent space, mobile robot first needs to know the map, then based on the map, it can locate itself. This paper introduces Simultaneous localization and mapping technology, and gives some difficulties. Combining Simultaneous localization and mapping and intelligent sapce, the mobile robot can locate itself better, and estimated the map is more accuracy. We use a new SLAM algorithm, combining Rao-Blackwellised particle filter and Extended Kalman filter, we can use the advantages of two algorithms. Rao-Blackwellised particle filter produces the local map, which are saved into the intelligent space, and fused into EKF. Compared with EKF, we can avoid the weakness of two popular mapping strategies.Finally, based on the global map that is produced by intelligent space, the mobile robot can use artificial potential field achiveing motion palnning. The algorithm is safe and reliable. The mobile robot can move in every direction.更多还原