复杂环境下多机器人协作构建地图的方法
An Approach to Cooperative Multi-robot Map Building in Complex Environments
作者：潘薇(1.中国科学技术信息研究所，北京 100038；2.中南大学 信息科学与工程学院, 湖南 长沙 410083)；蔡自兴(中南大学 信息科学与工程学院, 湖南 长沙 410083)；陈白帆(中南大学 信息科学与工程学院, 湖南 长沙 410083)
Author：Pan Wei(1.Inst. of Scientific and Technical Info. of China,Beijing 100038,China;2.School of Info. Sci. and Eng., Central South Univ., Changsha 410083,China)；Cai Zixing(School of Info. Sci. and Eng., Central South Univ., Changsha 410083,China)；Chen Baifan(School of Info. Sci. and Eng., Central South Univ., Changsha 410083,China)
收稿日期：2009-01-08 年卷（期）页码：2010,42(1):143-148
期刊名称：工程科学与技术
Journal Name：Advanced Engineering Sciences
关键字：多机器人；复杂环境；地图构建；距离变换；遗传算法
Key words：multi-robot；complex environments；map building；distance transform；genetic algorithm
基金项目：国防基础科研资助项目（A1420060159）
中文摘要
针对复杂环境以及无法获知机器人位置信息的情况，提出一种多移动机器人协作构建地图的新方法。该方法采用独立探索、集中建图的探索策略，对环境建立局部栅格地图并予以融合。在地图融合过程中，无需已知机器人的相对位置信息，而是以栅格地图相似度为度量标准，利用距离变换和改进的遗传算法高效、快速地搜索各局部地图之间的最大重叠部分，进而予以融合。此外，考虑到局部地图无重叠部分的情况，引入判断函数对融合结果予以度量。实验结果验证了该方法在复杂环境下的可行性和有效性。
英文摘要
Aiming at complex environments and the situation without any pose information of robots, a novel approach to multi-robot cooperative map building was presented. The approach let all robots operate individually and then tried to merge the different local grid maps into a single global one. Without using any relative pose information of robots, the process of map merging was performed by measuring the similarity between grid maps. Distance transforms and an improved genetic algorithm were used to effectively search the maximum overlap at which the local maps can be joined together. Furthermore, considering the situation that there may be no overlapped area among local maps, a value function was introduced to judge the merging results. Experimental results showed the feasibility and effectiveness of the present approach in complex environments.
【关闭】
