搜救环境中的仿生蛇形机器人SLAM方法  

SLAM of bionic snake-like robot in rescue environment

作　　者：王超杰[1] 苏中[1] 连晓峰[2] 赵旭[3] 

机构地区：[1]北京信息科技大学高动态导航技术北京市重点实验室,北京100101 [2]北京工商大学计算机与信息工程学院,北京100048 [3]北京理工大学自动化学院,北京100084

出　　处：《计算机工程与设计》2015年第8期2098-2102,共5页Computer Engineering and Design

基　　金：国家自然科学基金项目(61261160497);北京市科技计划课题基金项目(Z121100001612007);北京市教委科技成果转化和产业化基金项目(PXM2013_014224_000074)

摘　　要：针对搜救环境中蛇形机器人实时地图构建中计算量大、实时性差等问题,提出一种MiniSLAM算法。采用基于Monte-Carlo定位的粒子滤波算法对机器人位姿进行估计和更新,采用改进的Bresenham算法对环境地图进行更新。实验平台采用基于激光测距仪的自主研制的变结构仿生蛇形机器人,实验结果表明,与传统SLAM相比,该算法计算量较少,实时性高,在复杂环境中精度高,适用于搜救环境下蛇形机器人的应用。Aiming at the problems of large amount of calculation and poor real-time in real-time map building of snakelike robot in search and rescue environment, a MiniSLAM algorithm was presented. Based on Monte-Carlo particle filter positioning algo- rithmm, the pose of the robot was estimated and the modified Bresenham algorithm was used to update the environment map. Experimental platform used biomimetic snake-like robot with independently developed variable structure based on a laser range finder. Experimental results show that, compared with the traditional SLAM, the algorithm is of small amount of calculation, high real-time and high accuracy in complex environments. It is suitable for snake-like robot applications in search and rescue en- vironment.

关 键 词：蛇形机器人 MiniSLAM算法 激光测距仪 Monte-Carlo定位 BRESENHAM算法 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]