Nash Strategies and Adaptation for Decentralized Games Involving Weakly-coupled Agents

Abstract

We consider dynamic games in large population conditions where the agents evolve according to non-uniform dynamics and are weakly coupled via their individual dynamics and costs. A state aggregation technique is employed to obtain a set of decentralized control laws for the individuals which ensures closed-loop stability and possesses an é-Nash equilibrium property. We then propose a scheme for Nash strategy adaptation when the agents have unknown parameters. The issue of transient performance improvement is addressed by introducing dither signals.