透過您的圖書館登入
IP:216.73.216.156
  • 學位論文

多重代理人聯盟合作之穩定性研究

A Study on Stability Issues in Forming Cooperative Multiagent Coalitions

指導教授 : 劉立頌
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


多重代理人彼此之間的互動行為依其所面對的議題可大致分為四類:合作、協商、任務分配與溝通。合作賽局的理論著重在代理人的集合,即聯盟要如何達到單獨的代理人選擇不合作而做不到的事。合作賽局的目標在於找到有意義的聯盟形成,而聯盟的合作穩定性為其中之一。穩定性不足的聯盟中,代理人容易跳槽,已組成的聯盟容易解散,新組成的聯盟也容易因為有新加入代理人而解散,聯盟不容易組成。聯盟的合作穩定性可以透過解決方案概念這個標準來評估,其中,滿足核心狀態的解決方案概念是最嚴格與最常用的標準。本研究使用了盈餘分配法,透過設定最高購買金額,使得各個買家代理人的購買金額屬於核心狀態,且必為非空核心狀態,也就是聯盟的穩定狀態。但使用盈餘分配法這樣的聯盟組成方式並不夠穩定,所以本研究於使用盈餘分配法後,再加入第三方平台給予之聯盟補貼金來增加聯盟組成的穩定性。最後,本研究將系統實作於購買平台中以驗證研究的成果。

並列摘要


The issues about the interaction of multi-agent can be divided into four parts: cooperation, negotiation, task allocation and communication. The cooperative game theory focuses on how a coalition reaches an agreement that single agent could not reach if it chooses not to cooperate. Moreover, the goal of cooperative game theory is to form a meaningful coalition, including the stability of coalition. Low stability of coalition causes an agent to leave the original coalition easily. In addition, reforming a coalition will become difficult. Stability of coalition can be evaluated by solution concepts. Among them, the concept, core, is the strictest and most commonly used criterion in the cooperative game theory. In this study, we use the surplus sharing rule, which by setting up the highest purchase amount, makes the purchase amount of each participating agent to have the state of core and not empty core. However, the surplus sharing rule is not enough to promote coalition stability. We propose the concept of adding subsidy after using the surplus sharing rule to increase the stability of coalition. Finally, we implement the method on the marketplace.

並列關鍵字

無資料

參考文獻


[1] K. P. Sycara, “Multiagent systems,” AI magazine. vol. 19, no. 2, pp. 79-92, 1998.
[3] S. J. Russell and P. Norvig, Artificial intelligence: A modern approach. Prentice hall, pp. 31-50, 1995.
[5] N. R. Jennings, “On agent-based software engineering,” Artificial intelligence. vol. 117, no. 2, pp. 277–296, 2000.
[7] M. Wooldridge and N. R. Jennings, “Intelligent agents: Theory and practice,” The knowledge engineering review, vol. 10, no. 02, pp. 115-152, 1995.
[9] M. Wooldridge, An introduction to multiagent systems. Wiley, pp. 151-182, 2009.

延伸閱讀