arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.04832eess.SYcs.ROcs.SY
收藏

去中心化机器人网络协调的实验研究

Experimental Study of Decentralized Robot Network Coordination

Department of Electrical and Computer Engineering, Clemson University(克莱姆森大学电气与计算机工程系)

Martyn Lemon, Yongqiang Wang

AI总结:

本文改进并实验评估面向Roomba多机器人平台的去中心化同步与去同步算法,分析参数和三种计算方法对收敛及状态维持的影响。

摘要:

网络中的同步与去同步是许多电气系统中被广泛研究的课题,但在机器人领域却明显缺乏相关研究。为机器人网络创建一种有效的去中心化同步算法,将使多个机器人能够协同完成任务,并能够实时适应机器人的加入或丢失。本研究旨在改进作者为此目的而实现和开发的算法,并对这些方法进行实验评估。研究对先前工作中发现的用于同步和去同步的最有效算法进行了修改,以改进测试并使其计算方法多样化。实验研究使用了一个由多台Roomba机器人组成的多机器人平台。对数据的观察表明,调整算法参数如何影响达到期望同步或去同步状态所需的时间,以及网络如何维持该状态。对每种算法测试三种不同方法得到了不同结果。未来协作机器人领域的工作很可能会成功使用这些算法来完成各种任务。

英文摘要:

Synchronization and desynchronization in networks is a highly studied topic in many electrical systems, but there is a distinct lack of research on this topic with respect to robotics. Creating an effective decentralized synchronization algorithm for a robotic network would allow multiple robots to work together to achieve a task and would be able to adapt to the addition or loss of robots in real-time. The purpose of this study is to improve algorithms implemented developed by the authors for this purpose and experimentally evaluate these methods. The most effective algorithm for synchronization and desynchronization found in a former study were modified to improve testing and vary its methods of calculation. A multi-robot platform composed of multiple Roomba robots was used in the experimental study. Observation of data showed how adjusting parameters of the algorithms affected both the time to reach a desired state of synchronization or desynchronization and how the network maintained this state. Testing three different methods on each algorithm showed differing results. Future work in cooperative robotics will likely see success using these algorithms to accomplish a variety of tasks.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑