Game−Theoretic Agent Programming in Golog under Partial Observability

Abstract

In this paper, we present the agent programming language POGTGolog, which is a combination of explicit agent programming in Golog with game-theoretic multi-agent planning in a special kind of partially observable stochastic games (POSGs). It is a generalization of the agent programming language GTGolog by partial observability. The approach allows for partially specifying a high-level control program for a system of multiple agents, and for optimally filling in missing details by viewing it as a generalization of a special POSG and computing a Nash equilibrium. We illustrate this approach along a robotic rugby example.