Symmetries-enhanced Multi-Agent Reinforcement Learning
This work addresses generalization and scalability issues in multi-agent reinforcement learning for distributed swarming tasks, representing an incremental advancement by extending symmetry-enhanced methods to systems with limited intrinsic symmetries.
The paper tackles the challenge of insufficient intrinsic symmetries in multi-agent reinforcement learning by embedding extrinsic symmetries into system dynamics, enabling improved generalization and scalability. The method achieves significant reductions in collision rates and enhances task success rates across diverse scenarios and varying swarm sizes.
Multi-agent reinforcement learning has emerged as a powerful framework for enabling agents to learn complex, coordinated behaviors but faces persistent challenges regarding its generalization, scalability and sample efficiency. Recent advancements have sought to alleviate those issues by embedding intrinsic symmetries of the systems in the policy. Yet, most dynamical systems exhibit little to no symmetries to exploit. This paper presents a novel framework for embedding extrinsic symmetries in multi-agent system dynamics that enables the use of symmetry-enhanced methods to address systems with insufficient intrinsic symmetries, expanding the scope of equivariant learning to a wide variety of MARL problems. Central to our framework is the Group Equivariant Graphormer, a group-modular architecture specifically designed for distributed swarming tasks. Extensive experiments on a swarm of symmetry-breaking quadrotors validate the effectiveness of our approach, showcasing its potential for improved generalization and zero-shot scalability. Our method achieves significant reductions in collision rates and enhances task success rates across a diverse range of scenarios and varying swarm sizes.