Effective Multi-Agent Deep Reinforcement Learning Control with Relative Entropy Regularization

TL;DR

This paper introduces MACDPP, a novel multi-agent reinforcement learning method that improves learning efficiency and performance in complex control tasks by incorporating relative entropy regularization into the CTDE framework.

Contribution

The paper proposes MACDPP, a new MARL algorithm that enhances policy update consistency and sample efficiency using relative entropy regularization within an Actor-Critic framework.

Findings

MACDPP outperforms existing MARL algorithms in cooperation and competition tasks.

MACDPP demonstrates superior sample efficiency in benchmark control tasks.

The approach effectively handles complex multi-agent control scenarios.

Abstract

In this paper, a novel Multi-agent Reinforcement Learning (MARL) approach, Multi-Agent Continuous Dynamic Policy Gradient (MACDPP) was proposed to tackle the issues of limited capability and sample efficiency in various scenarios controlled by multiple agents. It alleviates the inconsistency of multiple agents' policy updates by introducing the relative entropy regularization to the Centralized Training with Decentralized Execution (CTDE) framework with the Actor-Critic (AC) structure. Evaluated by multi-agent cooperation and competition tasks and traditional control tasks including OpenAI benchmarks and robot arm manipulation, MACDPP demonstrates significant superiority in learning capability and sample efficiency compared with both related multi-agent and widely implemented signal-agent baselines and therefore expands the potential of MARL in effectively learning challenging control scenarios.