融合异质传感信息机器人粒子滤波定位方法.pdfVIP

第 25 卷 第 1 期 电子测量与仪器学报 Vol. 25 No. 1 · 38 · JOURNAL OF ELECTRONIC MEASUREMENT AND INSTRUMENT 2011 年 1 月 DOI: 10.3724/SP.J.1187.2011.00038 融合异质传感信息的机器人粒子滤波定位方法* 刘洞波1,2 刘国荣1,2 喻妙华2 (1. 湖南大学电气与信息工程学院, 长沙 410082; 2. 湖南工程学院计算机与通信学院, 湘潭 411104) 摘 要: 针对机器人采用视觉传感器和激光测距传感器单独定位的缺陷, 提出一种融合视觉传感器和激光测距传感器感知 信息的移动机器人 Monte Carlo 自定位方法。视觉目标识别过程中, 只采用激光测距信息单独进行粒子集更新; 当视觉目标识别 完成, 利用码盘信息对视觉定位信息进行修正, 然后融合激光测距信息进行粒子集的同步更新。视觉信息的全局性和激光测距 的快速性得到互补。实验表明, 运用异质传感器信息融合明显地加快了粒子集的收敛, 提高了移动机器人的自定位精度。 关键词: 移动机器人；Monte Carlo 算法；信息融合；自定位 中图分类号: TP24 文献标识码: A 国家标准学科分类代码: 520.2040 Method of particle filter localization of robot based on heterogeneous sensor information fusion Liu Dongbo1, 2 Liu Guorong 1, 2 Yu Miaohua2 (1. College of Electrical and Information Engineering, Hunan University, Changsha 410082, China; 2. Dept. of Computer and Communication, Hunan Institute of Engineering, Xiangtan 411104, China) Abstract: To deal with the localization disadvantage of robot only equipped with monocular camera or LRF (laser range finder sensor), a novel Monte Carlo method based on monocular camera and LRF sensor information fusion is proposed. During the vision object recognition, particle sets is updated by LRF sensor information; and the vision object recognition completed, the vision sensor information modified by the encoders is fused with that from LRF sensor information to synchronously update particle sets. Both monocular camera and LRF advantage is fully utilized. The simulation results show that particle sets is converged faster than those from ei