当前位置: X-MOL 学术arXiv.cs.MA › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Reasoning about Temporary Coalitions and LTL-definable Ordered Objectives in Infinite Concurrent Multiplayer Games
arXiv - CS - Multiagent Systems Pub Date : 2020-11-07 , DOI: arxiv-2011.03724
Dimitar P. Guelev

We propose enhancing the use of propositions for denoting decisions and strategies as established in temporal languages such as CTL*, if interpreted on concurrent game models. The enhancement enables specifying varying coalition structure. In quantified CTL* this technique also enables quantifying over coalition structure, and we use it to quantify over an extended form of strategy profiles which capture temporary coalitions. We also extend CTL* by a temporal form of a binary preference operator that can be traced back to the work of Von Wright. The resulting extension of quantified CTL* can be used to spell out conditions on the rationality of behaviour in concurrent multiplayer games such as what appear in solution concepts, with players having multiple individual objectives and preferences on them, and with the possibility to form temporary coalitions taken in account. We propose complete axiomatisations for the extension of CTL* by the temporal preference operator. The decidability of the logic is not affected by that extension.

中文翻译:

关于无限并发多人游戏中临时联盟和 LTL 可定义有序目标的推理

如果在并发博弈模型上进行解释,我们建议加强命题的使用,以表示在 CTL* 等时间语言中建立的决策和策略。增强功能可以指定不同的联盟结构。在量化的 CTL* 中,这种技术还可以量化联盟结构,我们使用它来量化扩展形式的策略配置文件,这些形式捕获临时联盟。我们还通过可以追溯到 Von Wright 的工作的二元偏好算子的时间形式扩展了 CTL*。由此产生的量化 CTL* 扩展可用于阐明并发多人游戏中行为合理性的条件,例如解决方案概念中出现的内容,玩家有多个个人目标和偏好,并考虑到形成临时联盟的可能性。我们提出了通过时间偏好算子扩展 CTL* 的完整公理化。逻辑的可判定性不受该扩展的影响。
更新日期:2020-11-10
down
wechat
bug