Mathematical Analysis of Multi-Agent Systems

Abstract

We review existing approaches to mathematical modeling and analysis of multi-agent systems in which complex collective behavior arises out of local interactions between many simple agents. Though the behavior of an individual agent can be considered to be stochastic and unpredictable, the collective behavior of such systems can have a simple probabilistic description. We show that a class of mathematical models that describe the dynamics of collective behavior of multi-agent systems can be written down from the details of the individual agent controller. The models are valid for Markov or memoryless agents, in which each agents future state depends only on its present state and not any of the past states. We illustrate the approach by analyzing in detail applications from the robotics domain: collaboration and foraging in groups of robots.