多機器足球員協作策略開發

Development of Multi-Soccer Robot Collaboration Strategy

康育銘(Yu-Ming Kang)

指導教授 ： 李世安

淡江大學/工學院/電機工程學系機器人工程碩士班/碩士(2020年)

https://doi.org/10.6846/TKU.2020.00138

摘要

本論文提出一多機器人合作策略設計方法。本論文使用機器人作業系統2(Robot Operating System 2, ROS2)建立多機器人通訊協定，並分享各自的機器人狀態和觀察到的球場資訊。之後透過資訊融合方法建立起雲端虛擬球場地圖。本論文根據球場的資訊去設計多機器人的合作策略。此多機器人合作策略根據場上機器人的位置，進行(1)群體機器人進攻、(2)聯合防守及(3)互相合作等三種策略。多機器人合作策略與單機器人策略不同的地方，在於多機器人會互相通訊後一起執行策略，以藉此增加進攻或防守的成功率。
在實驗結果部分，本文所提出的多機器人合作策略，可以得到更多的進攻機會與增加防守的成功率。

關鍵字

多機器人協作 ； 移動型機器人 ； 機器人作業系統2

並列摘要

This thesis proposes a design method to multi-robot collaboration strategy. This thesis uses Robot Operating System 2 (Robot Operating System 2, ROS2) to establish a multi-robot communication protocol and share their robot status and observed information. After that, a field was established through information fusion. This thesis designs multi-robot cooperation strategies based on the information of the virtual field in the cloud. This multi-robot cooperation strategy is according the position of the robots on the field to carry out (1) attack with group robot, (2) joint defense and (3) mutual cooperation. The difference between the multi-robot cooperation strategy and the single-robot strategy is that the multi-robots will communicate with each other to execute the strategy together to increase the success rate of offense or defense.
In the experimental results, the multi-robot cooperation strategy proposed in this thesis can get more offensive opportunities and increase the success rate of defense.

並列關鍵字

Behavior Decision ； Mobile Robot ； Robot Operating System 2(ROS2)

參考文獻

[1] FIRA, URL: http://www.fira.net

[2] RoboCup, URL: http://www.robocup.org

[3] FIFA, URL: http://www.fifa.com/

[4] ROS, URL: http://www.ros.org

[5] A. Y. Ng, S. Gould, M. Quigley, A. Saxena and E. Berger, “STAIR: Hardware and Software Architecture.” AAAI 2007 Robotics Workshop, 2007.