資料載入中.....
|
請使用永久網址來引用或連結此文件:
http://140.128.103.80:8080/handle/310901/21502
|
| 題名: | An evolutionary strategy for a computer team game |
| 作者: | Tsai, C.-T.a , Liaw, C.a, Huang, H.-C.a, Ko, C.-H.b |
| 貢獻者: | Department of Computer Science, Tunghai University |
| 關鍵詞: | Capture the Flag;First Person Shooter;multiplayer game;non-player character;Particle Swarm Optimization;Quake III;team game |
| 日期: | 2011 |
| 上傳時間: | 2013-05-14T09:07:39Z (UTC)
|
| 摘要: | Computer team games have attracted many players in recent years. Most of them are rule-based systems because they are simple and easy to implement. However, they usually cause a game agent to be inflexible, and it may repeat a failure. Some studies investigated the learning of a single game agent, and its learning capability has been improved. However, each agent in a team is independent and it does not cooperate with others in a multiplayer game. This article explores an evolution strategy for a computer team game based on Quake III Arena. The Particle Swarm Optimization (PSO) algorithm will be applied to evolve a non-player character (NPC) team in Quake III to be more efficient and intelligent. The evolution of a single NPC, which accommodates to its team and, moreover, the team has learning and cooperating abilities, will be discussed. An efficient team is composed of various members with their own specialties, and the leader is capable of evaluating the performance of a member and assigning it a proper job. Furthermore, the leader of an intelligent team will adapt a strategy appropriate for various circumstances and obtain the team's best performance. Instead of considering the tactic of an individual bot, this article takes the strategy of a team into account. ? 2011 Wiley Periodicals, Inc. |
| 關聯: | Computational Intelligence
Volume 27, Issue 2, May 2011, Pages 218-234 |
| 顯示於類別: | [資訊工程學系所] 期刊論文
|
文件中的檔案:
| 檔案 |
大小 | 格式 | 瀏覽次數 |
|---|
| index.html | 0Kb | HTML | 135 | 檢視/開啟 |
|
在THUIR中所有的資料項目都受到原著作權保護.
|
