基于激光雷达的有界区域快速全局定位算法  

A fast global localization algorithm for bounded region based on LIDAR

作　　者：吴海欣 文梓豪 熊弘俊 崔彦 林军记 张辉 WU Hai-xin;WEN Zi-hao;XIONG Hong-jun;CUI Yan;LIN Jun-ji;ZHANG Hui(School of Intelligent Systems Engineering,Sun Yat-sen University,Guangzhou 510006,China;Guangdong Provincial Key Laboratory of Intelligent Transportation System,Guangzhou 510006,China)

机构地区：[1]中山大学智能工程学院,广东广州510006 [2]广东省智能交通系统重点实验室,广东广州510006

出　　处：《激光与红外》2021年第4期428-434,共7页Laser & Infrared

摘　　要：室内全局定位技术是自动驾驶车辆和机器人的核心组成部分,也是当今的研究热点。基于该时代背景,本文提出了一种基于激光雷达的有界区域快速全局定位算法,将全局地图与局部地图的特征向量进行对比匹配,实现了雷达在有界区域内的精确定位。利用仿真实验对算法的有效性和定位精度进行了分析,发现其平均位置误差比为0.791%,平均朝向误差为0.3883°。结果表明,该算法可以较稳定且精确地获取静态雷达的位姿信息,是一种研究室内全局定位的有效方法。The technology of indoor global localization is the core component of automatic vehicle and robot,and it is also a research focus.Based on this background,this paper studied a fast algorithm of global localization algorithm based on lidar.The eigenvector of the global map and the local map are compared and matched to achieve the precise positioning of the lidar.The effectiveness and positioning accuracy of the algorithm were analyzed using simulation experiments,which shows that the average ratio of position error was 0.791%and the average orientation error was 0.3883°.The results indicate that the algorithm can stably and accurately acquire the pose information of static lidar,and it is an effective method to study the indoor global positioning.

关 键 词：激光雷达 全局地图 局部地图 特征向量 

分 类 号：TN249[电子电信—物理电子学]