arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2310.15686cs.MA
收藏

战略能力的假设-保证验证

Assume-Guarantee Verification of Strategic Ability

Łukasz Mikulski, Wojciech Jamroga, Damian Kurpiewski

AI总结:

针对不完美信息智能体战略能力模型检验难题,提出两种交替时序逻辑的假设-保证验证方案,证明其可靠性并讨论完备性,实验展示了该方法的实际效益。

摘要:

战略能力的模型检验是一个众所周知的难题,在不完美信息智能体的现实情况下更是如此。假设-保证推理在此能提供很大帮助,它提供了一种将复杂问题分解为一小部分指数级更简单的子问题的方法。在本文中,我们提出了两种针对不完美信息交替时序逻辑的假设-保证验证方案。我们证明了两种方案的可靠性,并讨论了它们的完备性。我们通过基于已知基准的示例说明了该方法,并展示了证明该方法实际效益的实验结果。

英文摘要:

Model checking of strategic abilities is a notoriously hard problem, even more so in the realistic case of agents with imperfect information. Assume-guarantee reasoning can be of great help here, providing a way to decompose the complex problem into a small set of exponentially easier subproblems. In this paper, we propose two schemes for assume-guarantee verification of alternating-time temporal logic with imperfect information. We prove the soundness of both schemes, and discuss their completeness. We illustrate the method by examples based on known benchmarks, and show experimental results that demonstrate the practical benefits of the approach.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑