首页 | 本学科首页   官方微博 | 高级检索  
     

分布式视觉机器人导航中的定位算法研究
引用本文:高业坤,朱劲,蒋平.分布式视觉机器人导航中的定位算法研究[J].数字社区&智能家居,2008(10):230-233.
作者姓名:高业坤  朱劲  蒋平
作者单位:同济大学控制理论与控制工程,上海201804
基金项目:基金项目:国家高技术研究发展计划资助项目(863-2006AA042222)
摘    要:生物启发的无线复眼导航技术是新型的机器人导航方案,将分布在环境中的分布式智能代替了传统的集中式智能。蒙特卡洛定位是近来流行的机器人自主定位算法,将这种算法应用在分布式视觉传感器机器人的定位中,并针对多视觉传感器观测值的最优选择,提出了一种分布式的基于熵的观测量选择方法,目的是选择那些对提高定位精度更有效的观测信息,在保证定位精度的前提下,提高了定位的卖时性和可靠性。仿真实验结果证明了这种算法的可行性。

关 键 词:机器人定位  贝叶斯滤波  蒙特卡洛定位  分布式视觉  

Research on Distributed Vision based Robot Localization
GAO Ye-kun,ZHU JinJIANG Ping.Research on Distributed Vision based Robot Localization[J].Digital Community & Smart Home,2008(10):230-233.
Authors:GAO Ye-kun  ZHU JinJIANG Ping
Affiliation:(Control Theory and Control Engineering,Tongji University, Shanghai 201804,China)
Abstract:Wireless Mosaic Eye system (WIME) is a new solution for robot navigation,robots are navigated by distributed intelligence in an enviroment instead of the conventional centraIized intelligence. Monte Carlo Localization (MCL) for mobile robots has become popular these days. In this paper, the MCL algorithm is applied in a scenario of distributed vision guided robot navigation. Taking into account observations from multiple cameras, this paper presents an observation selection method for enhancement of localization precision. It evaluates entropy of each observation and selects one with the richest information for robot localization. This method can increase localization precision and improve computational efficiency. Simulations are carried out to verify its feasibility.
Keywords:robot localization  bayes filter  monte carlo location  distributed visions  entropy
本文献已被 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号