多机器人队形控制的研究.doc

多机器人队形控制的研究
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。本文以近年来队形控制方面的研究成果为基础，做了以下几方面的讨论与研究：提出了一种新的队形表示图——吸引线段式主-从队形图；同时还定义了两种类似于势场函数但又有自己特点的人工力矩——主吸引矩和从吸引矩；提出了一种时间最优队形控制的几何算法，以解决队形控制中最优收敛问题，即队形变换过程中所需要的时间最短，有益于各种控制算法的指标优化。基于层次可扩展形描述框架 HEFF，提出了基于协商的队形动态生成方 法，可以在保持原队形形状不变的情况下动态扩展队形，以容纳新加入队形的机器人，并集成最大优先指派算法，实现了机器人到队形目标点的优解分派．文中所讨论的方法简单 、灵活，能充分利用机器人之间的进行操作，快速有效地生指队形。

关键词：指派问题  最短时限   队形控制   队形切换

More robots formation control research
Abstract
With the development of scientific research on complex system, community system’s troop controls have gradually become international frontier and hot topics of the interdisciplinary research among the community behavior imitation, system complexity science and other subjects. under the special assignation，accomplish their troop behavior for target through communication among the individuals and between system and environment. Considering the current research results and problems of formation control, we have discussed the following several areas of research: Method to Solve the Assignment Problem with the Shortest Time Limit and the Least Cost Constraint An Approach for Determination of Robots’ Vertexes and Application in Formation Control Whether may set up the unified mathematical model to solve the arithmetic diversity problem caused by multi-task. Presents a new type of formation model named attraction-segment leader-follower formation graph. Then, two types of artificial moments: leader attraction moment and follower attraction moment, are defined. This algorithm has solved the constringency-optimal problem of formation control, which is beneficial to target optimization of all control algorithms. A solution to dynamically build formation based on the negotiation using the HEFF(Hierarchical Extend Formation description Frame)was proposed．The formation can be dynamically extended while keeping the same formation shape．At the$alnc time。the MSFA(Max Sum First Assignment)method was integrated into the solution to optilIlizeas signments between robots and objective locations．The solution is simple，flexible。Which utilize robots’parallel operations and be used to buildthe formation effectively and quickly.

Key words: assignment problems ;   the shortest time limit；  formation control ;    formation rcconfigum;

目  录
摘 要 I
Abstract II
第一章  绪论 1
1.1 引言 1
1.2 多机器人系统研究的主要内容 2
1.2.1  群体体系结 …………………………………………………………………...2
1.2.2  感知…………………………………………………………………………..…….4
1.2.3  通信……………………………………………………………………………...…5
1.3 研究内容 6
第二章  基于人工力矩的多机器人队形控制 7
2.1系统模型与相关定义 7
2.2 人工力矩法 11
2.3 结论 12
第三章 基于协商的队形动态生成方法 13
3.1 协商队形的生成法与以往的对比分析………………..…………………………………13
3.2  HEFF队形描述框架 14
3.3队形动态生成方法 15
3.3.1 队形的扩展 16
3.3.2 队形协商生成的实现 18
3.3.3 队形切换 20
3.3.4 优化分派问题 20
3.4 仿真研究 21
3.5 队形协商生成 21
3.6 结 语 23
第四章 结论与展望 24
参考文献 25
致 谢 27
附录A外文参考文献 28
附录B中文翻译 29