Hierarchical Reinforcement Learning for Swarm Confrontation with High Uncertainty

TL;DR

This paper introduces a hierarchical reinforcement learning framework for swarm confrontation tasks under high uncertainty, effectively decoupling decision processes and improving training stability and performance in dynamic, multi-agent environments.

Contribution

It proposes a novel hierarchical RL architecture with uncertainty quantification and an integrated training method, advancing swarm confrontation strategies under uncertain conditions.

Findings

Achieves around 90% win rate in multi-agent confrontation scenarios.

Outperforms traditional methods in effectiveness and generalization.

Demonstrates robustness and stability in real-robot experiments.

Abstract

In swarm robotics, confrontation including the pursuit-evasion game is a key scenario. High uncertainty caused by unknown opponents' strategies, dynamic obstacles, and insufficient training complicates the action space into a hybrid decision process. Although the deep reinforcement learning method is significant for swarm confrontation since it can handle various sizes, as an end-to-end implementation, it cannot deal with the hybrid process. Here, we propose a novel hierarchical reinforcement learning approach consisting of a target allocation layer, a path planning layer, and the underlying dynamic interaction mechanism between the two layers, which indicates the quantified uncertainty. It decouples the hybrid process into discrete allocation and continuous planning layers, with a probabilistic ensemble model to quantify the uncertainty and regulate the interaction frequency adaptively. Furthermore, to overcome the unstable training process introduced by the two layers, we design an integration training method including pre-training and cross-training, which enhances the training efficiency and stability. Experiment results in both comparison, ablation, and real-robot studies validate the effectiveness and generalization performance of our proposed approach. In our defined experiments with twenty to forty agents, the win rate of the proposed method reaches around ninety percent, outperforming other traditional methods.