CASIA OpenIR  > 毕业生  > 博士学位论文
Alternative TitleResearch on Cooperative Hunting and Control of Multiple Mobile Robots
Thesis Advisor侯增广 ; 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 围捕 动态规划场 局部感知 通道 Multi-robot System Coordination Hunting Dynamic Programming Field Local Sensing Passageway
Abstract随着机器人应用领域的不断拓展,多机器人系统以其柔性、并行性及鲁棒性等特性受到普遍重视,围捕作为代表性的多机器人任务以其天然的动态性和对抗性成为研究热点。本文主要针对多移动机器人协调围捕与控制开展研究工作,主要内容如下: 首先,介绍了多机器人系统研究现状,同时综述了多机器人围捕方面的发展现状,并对研究背景和论文结构做了介绍。 其次,研究了基于位置分布的多机器人围捕问题,引入禁区和自由区概念,提出了无障碍物环境追捕机器人和入侵者等速条件下的通用围捕策略,以及成功围捕的充分必要条件,并利用李雅普诺夫稳定性理论进行了证明,同时分析了速度方向和障碍物对围捕的影响。 第三,提出用动态规划场描述环境信息的方法。动态规划场通过GAP-RBF网络或支持向量机学习得到,能直接计算已知环境中任意两点之间考虑障碍物情况下的近似最短距离。在此基础上,设计了基于动态规划场的多机器人围捕策略,基于感知范围内同伴信息的变化动态改变网络结构以提高协调性,最终实现对采用基于动态规划场逃跑策略的入侵者的围捕。 第四,提出一种基于局部感知的移动机器人目标跟踪方法。视觉、码盘、超声和红外传感器信息融合处理后,产生决策空间,得到障碍物分布点集,在此基础上,引入通道概念,提出可行通道决策算法选取合适的通道方向,实现未知环境下无碰的目标跟踪,同时应对可能出现的局部极小。 第五,提出一种基于模糊控制协调策略的多机器人围捕方法。追捕机器人根据局部感知获取的目标和左右最邻角同伴信息,选取不同的模糊控制器用以控制与同伴之间的夹角,进而结合基于局部感知的目标跟踪方法,实现对智能入侵者的协调围捕。 最后,论文对所取得的研究成果进行了总结,并阐述下一步的工作。
Other AbstractWith the expansion of the robot application, multiple robots systems have been paid much attention with the characteristics of flexibility, parallelism and robustness. As a representative task of multi-robot system, the hunting by multiple robots has become a hotspot because of its natural dynamism and antagonism. This thesis focuses on cooperative hunting and control of multiple mobile robots. The contents are as follows: Firstly, the research development of multi-robot system are described. The survey of multi-robot hunting is presented. The background and structure of this thesis are also introduced. Secondly, the hunting with position distribution of multiple robots is discussed. The concepts of forbidden zone and freedom zone are introduced. Under the condition that the predator robots and the invader have the same velocity, a general hunting strategy with non-obstacle environments is proposed, and the necessary and sufficient condition of successful hunting is proved based on the idea of lyapunov stability. Also, the impacts of speed direction and obstacles are analyzed. Thirdly, a dynamic programming field based environment information description method is proposed. Dynamic programming field, which may be used to compute the approximate shortest distance between any two points with the consideration of obstacles in known environments, can be trained by GAP-RBF network or support vector machine. On this basis, a multi-robot hunting strategy based on dynamic programming field is designed and the structure of network may be varied dynamically according to the information of teammates within sensing range to improve the coordination of system, which is used to pursue the invader with the escape strategy based on dynamic programming field. Fourthly, a local sensing based target tracking approach for mobile robot is proposed. The information provided by CCD camera, encoder, sonar and infrared sensors are combined to produce decision-making space, and the points set of the distribution of obstacles are given. On this basis, the concept of passageway is introduced and a passageway-based decision-making algorithm is proposed to obtain the appropriate passageway direction, which is used to achieve the collision-free target tracking in unknown environments and cope with possible local minimum. Fifthly, a hunting approach based on fuzzy coordination is proposed. According to the local sensing information of target and the angle-minimal neighbor teamma...
Other Identifier200718014628024
Document Type学位论文
Recommended Citation
GB/T 7714
袁瑗. 多移动机器人协调围捕与控制研究[D]. 中国科学院自动化研究所. 中国科学院研究生院,2011.
Files in This Item:
File Name/Size DocType Version Access License
CASIA_20071801462802(10174KB) 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
[袁瑗]'s Articles
Baidu academic
Similar articles in Baidu academic
[袁瑗]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[袁瑗]'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.