CASIA OpenIR  > 毕业生  > 硕士学位论文
Alternative TitleResearch on Multi-robot Hunting Task Coordination and Target Recognition
Thesis Advisor曹志强
Degree Grantor中国科学院研究生院
Place of Conferral中国科学院自动化研究所
Degree Discipline控制理论与控制工程
Keyword多机器人系统 协调 任务分配 动态联盟 目标识别 Multi-robot Systems Coordination Task Allocation Dynamic Alliance Target Recognition
Abstract随着控制技术、计算机和人工智能的快速发展,以及人们需求的日益复杂,多机器人系统以其分布式、柔性、鲁棒性等优势发挥着越来越重要的作用。要使多个机器人有效地运行,就必须对系统加以组织。目前,多机器人协调与控制的研究已经成为机器人学研究的重要方向之一。本文针对多机器人系统围捕任务协调与目标识别开展研究工作,主要内容如下: 首先,本文对多机器人系统进行了综述,简要介绍了有代表性的多机器人系统以及典型的多机器人仿真平台,综述了多机器人任务分配、围捕任务和基于视觉的目标识别研究现状,并对论文的选题背景和结构安排做了介绍。 其次,提出了一种基于任务案例匹配的分层式任务分配方法。该方法基于分层式多机器人组织形式,将含有历史任务信息的任务案例与合同网、熟人网的任务分配方式结合起来,应用于多机器人围捕仿真中。 第三,针对多机器人围捕任务中入侵者多于一个的情况,在基于上述任务分配方法得到围捕特定入侵者的机器人小队后,为了更好的执行任务,提出了一种基于圆形-椭圆双包围圈的动态联盟围捕策略,给出了机器人小队间的联盟条件,通过仿真进行了验证。 第四,给出一种基于参照光线阈值动态调整的机器人目标识别算法。在多组确定的参照光线下进行相应颜色的阈值标定后,根据当前图像的亮度信息,动态进行阈值调整,在颜色识别的基础上结合形状信息完成目标的识别。 最后,论文对所取得的研究成果进行了总结,并阐述了下一步的工作。
Other AbstractAs the rapid development of control, computer and artificial intelligence as well as the complexity of demand, multi-robot systems play a more and more important role because of the advantages of distribution, flexibility, robustness. Multi-robot coordination and control have become an important research direction in robotics. This paper considers multi-robot hunting task coordination and target recognition, which are as follows: Firstly, multi-robot systems are introduced, and several representative multiple robots systems and typical simulation platforms are described. The current research work on task allocation, multi-robot hunting, and target recognition are summarized. We also give the background and content arrangement of the paper. Secondly, a hierarchical task allocation method based on task case is proposed. Based on a layered organization, the method combines the task cases with history information, contract net, and acquaintance net together. The proposed method is tested in multi-robot hunting simulation. Thirdly, for the case of more than one evader in multi-robot hunting, after the robotic teams to pursue the specific evaders are determined by the allocation method above mentioned, we propose a hunting strategy with dynamic alliance based on circular-elliptical besieging circle for a better performance, and give the alliance conditions. The simulations are conducted to verify the proposed strategy. Fourthly, a target recognition method with threshold dynamic adjustment based on reference lights is given. After calibrating all thresholds of colors in different reference lights, we adjust the thresholds of current image to get the belongings of all pixels. On the basis of colors, the targets are recognized with shape information. Finally, the research work is concluded and the future research is described.
Other Identifier200628014628011
Document Type学位论文
Recommended Citation
GB/T 7714
马莹. 多机器人系统围捕任务协调与目标识别的研究[D]. 中国科学院自动化研究所. 中国科学院研究生院,2009.
Files in This Item:
File Name/Size DocType Version Access License
CASIA_20062801462801(13264KB) 暂不开放CC BY-NC-SAApplication Full Text
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[马莹]'s Articles
Baidu academic
Similar articles in Baidu academic
[马莹]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[马莹]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.