【作者】 于克；

【导师】 黄真； 王洪波；

【作者基本信息】 燕山大学 ， 机械电子工程， 2009， 硕士

【摘要】 定位是自主移动机器人的基本功能,也是移动机器人实现自主导航的关键,对于提高机器人自动化水平具有重要意义。论文对移动机器人的绝对定位技术进行了深入分析和系统研究。首先,论文将无线传感器网络引入移动机器人的研究中,提出了一种基于无线传感器网络ZigBee的定位方法,该方法类似于一个小型的卫星定位系统(GPS)。在不能或不宜使用卫星系统进行定位时,可以用该方法代替卫星系统对移动机器人进行定位。其次,在回顾和总结目前存在的移动机器人环境建模方法的基础上,本文提出了一种新的基于二维激光测距传感器的几何地图构建方法:即采用线段和特征点描述环境。对环境的特征提取方法进行了详细的阐述,从而建立了比较完整的环境地图。再次,在环境模型建立的基础上,论文提出一种基于完整线段匹配的定位算法:即基于完整线段的长度关系建立匹配假设,根据完整线段与特征点的相对位置关系对匹配假设进行评价,进而建立匹配矩阵和标志矩阵,根据匹配矩阵和标志矩阵,利用最佳匹配搜索算法搜索最佳匹配,基于最佳匹配,得到机器人的位姿。最后,论文将无线传感器网络ZigBee信息和激光测距传感器信息相融合,提出了室内移动机器人快速定位算法。利用无线传感器网络ZigBee实现移动机器人的粗略定位,然后从先验全局地图中抽取该位置邻近的参考地图,再与从激光测距传感器获取的信息中快速提取的环境特征进行匹配,根据匹配的环境特征实现机器人的精确定位。更多还原

【Abstract】 Localization is the basic function and the critical step towards autonomous navigation of autonomous mobile robot. It is of great significance to improve the automatization level. In this dissertation, a deep analysis and systematic study of technologies related to the mobile robot’s absolute positioning are carried on.Firstly, a wireless sensor networks is introduced into the research of mobile robot. This paper presents a localization method based on the ZigBee wireless sensor networks. This method is similar to a small global positioning system (GPS), and it can be used in place of satellite systems for mobile robot localization while which can not or should not be used.Secondly, on the basis of review and summary of methods on mobile robot environmental model, this paper proposes a new geometry map building approach based on two-dimensional laser range sensor, namely, line segment and feature point are adopted to describe environment. The method of environmental feature extraction is described in detail in order to build a relatively complete map of the environment.Thirdly, this paper presents a localization method on the basis of complete line segments matching when the environmental model is established. The paper establishes match assumptions based on the length of complete line segments, and evaluates the match assumptions using the relative relationships between complete line segments and feature points in order to establish the match matrix and the flag matrix. The robot pose is obtained based on the best match which is searched according to the match matrix and flag matrix using the best match search algorithm.Finally, the paper presents a rapid localization method for an indoor mobile robot using the integrated information from the ZigBee wireless sensor networks and laser range sensor. The algorithm achieves a rough global location of the mobile robot by using the ZigBee wireless sensor networks, and extracts the reference map close to the mobile robot from the known global map. Then, the algorithm matches the known global map and the environmental features extracted from the laser range sensor information. According to the matched environmental characteristics, the accurate position of the robot can be obtained.更多还原