Bi-CL: A Reinforcement Learning Framework for Robots Coordination Through Bi-level Optimization

TL;DR

This paper presents Bi-CL, a bi-level reinforcement learning framework for multi-robot coordination that improves learning efficiency and scalability through a novel optimization structure and alignment mechanism.

Contribution

It introduces a bi-level optimization approach within reinforcement learning for multi-robot systems, addressing information mismatch and enhancing training efficiency.

Findings

Bi-CL learns more efficiently than traditional methods.

Achieves comparable coordination performance to existing baselines.

Effective in route-based and graph-based multi-robot scenarios.

Abstract

In multi-robot systems, achieving coordinated missions remains a significant challenge due to the coupled nature of coordination behaviors and the lack of global information for individual robots. To mitigate these challenges, this paper introduces a novel approach, Bi-level Coordination Learning (Bi-CL), that leverages a bi-level optimization structure within a centralized training and decentralized execution paradigm. Our bi-level reformulation decomposes the original problem into a reinforcement learning level with reduced action space, and an imitation learning level that gains demonstrations from a global optimizer. Both levels contribute to improved learning efficiency and scalability. We note that robots' incomplete information leads to mismatches between the two levels of learning models. To address this, Bi-CL further integrates an alignment penalty mechanism, aiming to minimize the discrepancy between the two levels without degrading their training efficiency. We introduce a running example to conceptualize the problem formulation and apply Bi-CL to two variations of this example: route-based and graph-based scenarios. Simulation results demonstrate that Bi-CL can learn more efficiently and achieve comparable performance with traditional multi-agent reinforcement learning baselines for multi-robot coordination.