Visualization of dynamic behaviour of multi-agent systems

Abstract

The extension of our research on analysis of a single agent or agent communities combining advanced methods of visualization with traditional AI techniques is presented in this paper. Even though this approach can be used for arbitrary Multi-Agent System (MAS), it was primarily developed to analyze systems falling into Artificial Life domain. Traditional methods are becoming insufficient as Multi-Agent Systems (MAS) are becoming more complex and therefore novel approaches are needed. In this paper we present an extension of our recent visualization tools suite. The previous approach was not suitable well to present the dynamics of the MAS, even though the development of MAS state parameters in time was presented. Our new technique, which is presented in this paper, addresses this problem by visualizing the changes of the MAS along with their quality and context. This transparent approach emphasizes MAS dynamics by providing means for discovery of changes in its tendencies or in behaviour of either single agent or agent communities. A simulated artificial life environment with intelligent agents has been used as a test bed. We have selected this particular domain because our longterm goal is to model life as it could be so as to understand life, as we know it.