Agent programming with temporally extended goals

In planning as well as in other areas, temporal logic has been used to specify so-called temporally extended goals. Temporally extended goals refer to desirable sequences of states instead of a set of desirable final states as the traditional notion of achievement goal does, and provide for more variety in the types of goals allowed. In this paper, we show how temporally extended goals can be integrated into the agent programming language GOAL. The result is that GOAL agents may now have both beliefs about the future as well as have temporally extended goals. We propose a new decision making mechanism that takes temporally extended goals into account, and investigate properties of this framework. Categories and Subject Descriptors I.2.11 [Artificial Intelligence]: Distributed Artificial Intelligence— Intelligent agents, languages and structures; I.2.5 [Artificial Intelligence]: Programming Languages and Software General Terms Theory, Languages Keywords Agent Programming, Goals, ...

• Type: Conference
• Year: 2009
• Where: ATAL
• Authors: Koen V. Hindriks, Wiebe van der Hoek, M. Birna van Riemsdijk
• Keywords: Agent Programming, Artificial Intelligence, ATAL 2009, Extended Goals, Temporal Logic