Simulation of action theories and an application to general game-playing robots

被引：0

作者：

Thielscher, Michael ^{[1
]}

机构：

[1] School of Computer Science and Engineering, The University of New South Wales, Sydney, 2052, NSW

来源：

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | 2015年 / 9060卷

关键词：

33;

D O I：

10.1007/978-3-319-14726-0_3

中图分类号：

学科分类号：

摘要：

We consider the problem of verifying whether one action theory can simulate a second one. Action theories provide modular descriptions of state machines, and simulation means that all possible sequences of actions in one transition system can be matched by the other.We show how Answer Set Programming can be used to automatically prove simulation by induction from an axiomatisation of two action theories and a projection function between them. Our interest in simulation of action theories comes from general game-playing robots as systems that can understand the rules of new games and learn to play them effectively in a physical environment. A crucial property of such games is their playability, that is, each legal play sequence in the abstract game must be executable in the real environment. © Springer International Publishing Switzerland.

引用

页码：33 / 46

页数：13