摘要
为解决粒子滤波SLAM中存在的计算效率高及粒子退化造成的估计精度低等问题,结合精确稀疏滞后状态信息滤波估计精度高以及精确稀疏扩展信息滤波计算效率高的优点,将两者混合应用于粒子滤波SLAM算法中,不但在保证计算效率的条件下提高了状态估计精度,并且还克服了机器人转动状态以及环境特征疏密带来的应用缺陷。实验结果表明了该方法的有效性与可行性。
In order to solve the problems of much computation and low estimation precision caused by particle degradation.The advantage of ESDF is high estimation accuracy and ESEIF’s advantage is high efficiency.This paper used a mixture of them in the particle filter SLAM,which not only used the historical information maintained by information matrix of the relationship between the robot pose and characteristics fully and improved the accuracy of the estimate,but also overcame the defects in the application made by robot rotational state and characteristics density.Experimental results show that the proposed algorithm is valid and feasible.
出处
《计算机应用研究》
CSCD
北大核心
2013年第7期1988-1990,1994,共4页
Application Research of Computers
基金
国家自然科学基金资助项目(61101146)
陕西省教育厅科学研究计划资助项目(12JK0518)
西安工程大学博士科研启动基金资助项目(BS1207)
关键词
同时定位与地图创建
精确稀疏滞后状态滤波
精确稀疏扩展信息滤波
粒子滤波
历史信息
simultaneous localization and map building（SALM）
exactly sparse delayed-state filter（ESDF）
exactly sparse extended information filter（ESEIF）
particle filter
historical information
