聯(lián)盟生成是多Agent 系統(tǒng)的一個(gè)關(guān)鍵問題，主要研究如何在多Agent 系統(tǒng)中動(dòng)態(tài)生
成面向任務(wù)的最優(yōu)Agent 聯(lián)盟。本文討論了Agent 聯(lián)盟的生成，并將遺傳算法應(yīng)用于Agent聯(lián)盟生成中，對(duì)比實(shí)驗(yàn)結(jié)果表明，遺傳算法可以快速、高效地找出合適的Agent 聯(lián)盟，并在解的質(zhì)量和收斂速度上均優(yōu)于相關(guān)算法。
關(guān)鍵詞： Agent 聯(lián)盟；遺傳算法；聯(lián)盟生成
Abstract: Coalition formation of the Multi-Agent System is a key issue, mainly on how
Multi-Agent System dynamically generate optimal task-oriented Agent Coalition. This article will discuss the formation of Agent coalition, and then applied genetic algorithm to the formation of Agent coalition, Experimental results show that the genetic algorithm can quickly and efficiently identify Agent coalition, and the performance and convergence rate are better than correlation algorithm.
Key words: Agent coalition；Genetic algorithm；Coalition formation