Learning Distributed Stabilizing Controllers for Multi-Agent Systems

TL;DR

This paper introduces two reinforcement learning algorithms for model-free, distributed stabilization of heterogeneous multi-agent systems, with proven convergence and demonstrated effectiveness through simulations.

Contribution

It presents novel RL algorithms for distributed stabilization of multi-agent systems without prior stabilizing gains, extending to predefined interaction graphs.

Findings

Algorithms guarantee convergence under certain conditions

Successful stabilization demonstrated in simulations

Applicable to heterogeneous multi-agent systems

Abstract

We address the problem of model-free distributed stabilization of heterogeneous multi-agent systems using reinforcement learning (RL). Two algorithms are developed. The first algorithm solves a centralized linear quadratic regulator (LQR) problem without knowing any initial stabilizing gain in advance. The second algorithm builds upon the results of the first algorithm, and extends it to distributed stabilization of multi-agent systems with predefined interaction graphs. Rigorous proofs are provided to show that the proposed algorithms achieve guaranteed convergence if specific conditions hold. A simulation example is presented to demonstrate the theoretical results.