Modeling self satisfaction and altruism to handle action selection and reactive cooperation

Abstract

This paper proposes a method for integrat-ing cooperative behaviors among distributed au-tonomous agents based on local communications. We dene agent satisfaction as a signal to handle action selection and cooperative interaction. The resolution of spatial problems is handled with an altruistic behavior dened with the vector eld model. This solution ensure the spatial coordina-tion of multiple agents in a dynamic system and the selection of an altruistic behavior when neces-sary. After presenting the agent architecture we describe experimental results. They illustrate the ability of the model to achieve eÆcient naviga-tion, adaptive and reactive cooperation in collec-tive robotics application. 1.