Multi-agent assignment via state augmented reinforcement learning

TL;DR

This paper introduces a distributed multi-agent assignment method using state-augmented reinforcement learning, leveraging dual variable oscillations and communication networks to coordinate agents without sharing local states, with theoretical guarantees and numerical validation.

Contribution

It proposes a novel distributed multi-agent assignment protocol based on state augmentation and dual variable oscillations, overcoming limitations of standard regularization techniques.

Findings

The protocol achieves effective multi-agent assignment without sharing local states.

Theoretical feasibility guarantees are established for the proposed method.

Numerical experiments validate the approach's effectiveness.

Abstract

We address the conflicting requirements of a multi-agent assignment problem through constrained reinforcement learning, emphasizing the inadequacy of standard regularization techniques for this purpose. Instead, we recur to a state augmentation approach in which the oscillation of dual variables is exploited by agents to alternate between tasks. In addition, we coordinate the actions of the multiple agents acting on their local states through these multipliers, which are gossiped through a communication network, eliminating the need to access other agent states. By these means, we propose a distributed multi-agent assignment protocol with theoretical feasibility guarantees that we corroborate in a monitoring numerical experiment.