Extending SATPLAN to Multiple Agents

Multi-agent planning is a core issue in the multi-agent systems field. In this work we focus on the coordination of multiple agents in a setting where agents are able to achieve individual goals that may be either independent, or necessary for the achievement of a global common goal. The agents are able to generate individual plans in order to achieve their own goals, but, as they share the same environment, they need to find a coordinated course of action that avoids harmful (or negative) interactions, and benefits from positive interactions, whenever this is possible. Moreover, agents are interested in finding plans with optimal length where preference is given to the length of the joint plan. We formalize these problems in a more general way with respect to previous works and present a coordination algorithm which provides the optimal solution in the case of two agents. In this algorithm, agents use

• Type: Journal
• Year: 2010
• Where: SGAI
• Authors: Yannis Dimopoulos, Muhammad Adnan Hashmi, Pavlos Moraitis
• Keywords: Artificial Intelligence, Multi-agent Planning Problems, Multi-agent Systems Field, Positive Interactions, SGAI 2010