運用基因型案例推理產生足球機器代理人團隊策略

Use hybrid approach CBR-GA to RoboCup’s Strategy

沈月笙(Yue-Sheng Shen)

指導教授 ： 郭忠義

國立臺北科技大學/電資學院/資訊工程系研究所/碩士(2009年)

https://doi.org/10.6841/NTUT.2009.00157

摘要

機器人足球競賽(RoboCup Soccer)是近十年國際上發展的一種機器人競賽，規則與人類足球比賽相似。目前足球機器人的發展仍有許多困難必須克服，例如足球機器人高階的行為設計、控制機器人低階基本動作時的不確定性、足球機器人彼此合作與協調等。本研究將針對足球機器代理人團隊的策略產生的設計提出解決方法。我們以基因型案例式推論(Genetic Case-Based Reasoning, CBR-GA)來發展足球機器人的策略產生，使機器人團隊能根據目前的情境產生適合的策略。本論文將足球代理人分成兩種型態：教練代理人(Coach Agent)為決策者，執行代理人是執行者。多代理人系統如同人類社會體系，有決策者和執行者，根據環境的不同，決策者可訂出有效的策略，而執行者遵循策略彼此合作以達成目標。教練代理人根據當下競賽情境，負責分配執行代理人扮演的角色與工作任務，使用基因型案例式推論方法產生團隊策略。執行代理人根據策略計畫執行，將所得到的結果寫入案例庫，以保存這次策略執行的經驗。利用經驗的累積，達到策略學習的效果，成功的經驗能幫助未來在相似的情境之成功率，而失敗的經驗能使代理人避開相同的錯誤。

關鍵字

RoboCup ； CBR-GA ； 策略

並列摘要

RoboCup (Robot world cup tournament) soccer game is a competition game. It has already become a popular research domain in recent years. Developing RoboCup agents involves solving the cooperation of multiple agents, the learning of adaptive behavior, and the resolution to noisy data handling. In this paper, we use hybrid approach CBR-GA(Case Based Reasoning-Genetic Algorithm) to provide better strategy for robots planning base on all kinds of conditions and save the related experience for reusing to improve the drawback of CBR-GA. We show our robots can be self-learning and self-organization and cooperate with other robots in a RoboCup game.The robots not only learn how to score but also know how to avoid the same mistakes. And we show the effectiveness of the proposed method by implementing a prototype and by comparing with another learning approach.

並列關鍵字

RoboCup ； CBR-GA ； strategy

參考文獻

[3] T.Srinivasan, K.Aarthi, S.Aishwarya Meenakshi, M.Kausalya, "CBRRoboSoc : An Efficient Planning Strategy for Robotic Soccer using Case Based Reasoning", International Conference on Computational Intelligence for Modeling Control and Automation, and International Conference on Intelligent Agents, Web Technologies and Internet Commerce, pp.113-113, 2006.

[5] Nguyen Hoang Phuong, Nadipuriun R. Prasad, Dang Huu Hung, Jeffery T. Drake, “Approach to Combining Case Based Reasoning with Rule Based Reasoning for Lung Disease Diagnosis”, International Conference on IFSA World Congress, Vol. 2, pp.883-888, 2001.

[6] Lai Chaoan, “The expert system of product design based on CBR and GA”, International Conference on Computational Intelligence and Security Workshops, pp.144-147, 2007.

[7] A. Rao, and M. Georgeff, “BDI Agents: From Theory to Practice”, InternationalConference on Multi-Agent Systems, 1995.

[11] C. Vasudevan, “An Experience-Based Approach to Software Project Management”, International Conference on Tools with Artificial Intelligence,. Proceedings of 6th, pp.624-630, 1994.