CASIA OpenIR  > 毕业生  > 博士学位论文
未知环境下多机器人协调与控制的队形问题研究
其他题名RESEARCH ON FORMATION PROBLEMS OF COORDINATION AND CONTROL OF MULTI-ROBOT SYSTEM WITH UNKNOWN ENVIRONMENT
曹志强
学位类型工学博士
导师谭民
2002-04-01
学位授予单位中国科学院研究生院
学位授予地点中国科学院自动化研究所
学位专业控制理论与控制工程
关键词多机器人 协调与控制 队形 基于行为 围捕 Multi-robot Coordination And Control Formation Behavior-based Hunting
摘要随着机器人技术的发展,机器人的应用领域向更广阔的范围扩展。人们希望 机器人不仅仅能够完成一些简单的工作,而且还能在复杂、未知的环境下完成复 杂的工作任务。由于许多复杂任务是单个机器人难以或无法完成的,于是考虑将 多个机器人构成系统,通过相互间的协调协作来完成。多机器人系统以其柔性、 并行性及鲁棒性等单个机器人所无法比拟的优越性受到普遍重视。本文针对未知 环境下多机器人协调与控制的队形问题开展研究工作。 本文首先对多机器人系统的优越性、应用背景等进行了介绍,对多机器人系 统的研究概况、主要研究内容进行了综述,并对本文的选题背景、主要内容做了 介绍。 其次,根据机器人系统任务级协作的实现、移动机器人应当具备的能力,考 虑到机器人越来越复杂的应用环境、任务要求,提出了一种混合分层的移动机器 人个体控制体系结构,分为系统监控层、协作规划层和行为控制层三个层次。 第三,本文针对多机器人系统队形完全保持任务下基于行为选择机制的策略 开展研究,给出了具体的行为描述、合理的行为选择条件以及“Leader更换”方 法,确保了任务的顺利完成。 第四,针对以完成队形控制任务为目标的多个基于Motor Schema的结构的机 器人行为控制参数难以配置的问题,本文采用遗传算法进行多个机器人行为控制 参数的联合优化。基于优化的行为控制参数,机器人实现了相互之间的协调队形 运动。 第五,在动态未知环境下多机器人队形控制的研究中,针对采用固定控制参 数进行行为合成的机器人对外界变化的适应性较低及动态障碍物的存在增加了机 器人避碰难度的问题,本文提出了一种机器人行为控制参数随着条件的变化而自 动调整的方法,同时采用抛物线模型和限定记忆最小二乘法实现了对动态障碍位 置的预测。 第六,本文研究了多移动机器人在未知环境下协作围捕具有安全运动策略的 Invader问题。在对任务建模的基础上,提出了一系列策略使得多机器人相互协作 在安全无碰的运动过程中发现并捕捉到Invader,在开发的仿真系统上实现。 最后,本文对所取得的研究成果进行了总结,并指出了需要继续开展的工作。
其他摘要With the development of robotics technology, the application of robotics has been extended more and more widely. The robot is required not only to complete some simple tasks, but also to achieve complex tasks with complex and unknown environment. For many complex tasks, it is hard or impossible for single robot to accomplish by itself. Therefore the system composed of multiple robots is considered and the tasks are performed through coordination and cooperation among multiple robots. Multi-robot system has been paid more and more attention because of its advantages such as flexibility, parallelism and robustness that single robot cannot compare. This thesis focuses on the formation problems of coordination and control of multi-robot system with unknown environment. Firstly, the advantages and application of multi-robot system are introduced. The research development and main research directions of multiple robots system are reviewed. The background and structure of this thesis are also introduced. Secondly, a layered control architecture for individual mobile robot is proposed based on the achievement of task-level cooperation of multiple robots system, the abilities the mobile robot should have and the consideration of more and more complex environment and task demand. The architecture is composed of three layers: system-monitoring layer, cooperation-planning layer and behavior-control layer. Thirdly, the strategy based on the behavior selection mechanism used to complete the formation-keeping task of multi-robot system is discussed. The behaviors' description, reasonable conditions of behavior selection and "Leader changing" approach are given in order to ensure the smooth completion of the task. Fourthly, in order to solve the problem that the behaviors' control parameters of robot adopting motor-schema based architecture with the formation control task are hard to be chosen, genetic algorithm is adopted to optimize the control parameters of robots. The robots with the optimized control parameters of behaviors move in a specified formation harmoniously. Fifthly, considering the problems that robot adopting fixed control parameters used to behaviors' combination has lower adaptability to environment and the difficulty of collision avoidance is increased because of dynamic obstacles during the research on multi-robot formation control with dynamic and unknown environment, a method that the behaviors' control parameters of robot adjust automatically with the conditions is proposed, and Parabola prediction model as well as recurrence least square algorithm with restricted scale are adopted to achieve the prediction of moving obstacles' positions. Sixthly, the problem that multiple mobile robots cooperatively hunt the invader that has safety motion strategy with unknown environment is discussed. A series of strategies are proposed on the basis of task modeling in order to ensure that multiple robots find and cap
馆藏号XWLW668
其他标识符668
语种中文
文献类型学位论文
条目标识符http://ir.ia.ac.cn/handle/173211/5729
专题毕业生_博士学位论文
推荐引用方式
GB/T 7714
曹志强. 未知环境下多机器人协调与控制的队形问题研究[D]. 中国科学院自动化研究所. 中国科学院研究生院,2002.
条目包含的文件
条目无相关文件。
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[曹志强]的文章
百度学术
百度学术中相似的文章
[曹志强]的文章
必应学术
必应学术中相似的文章
[曹志强]的文章
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。