Collision Avoidance for Multiple UAVs in Unknown Scenarios with Causal Representation Disentanglement

TL;DR

This paper introduces a causal representation disentanglement method for multi-UAV collision avoidance, improving generalization and robustness in unseen scenarios by isolating causal factors in visual representations.

Contribution

The paper proposes a novel causal disentanglement approach that enhances DRL-based UAV navigation by explicitly removing non-causal factors, leading to better generalization.

Findings

Achieves robust collision avoidance in unseen scenarios

Significantly outperforms existing state-of-the-art algorithms

Improves generalization ability of DRL models

Abstract

Deep reinforcement learning (DRL) has achieved remarkable progress in online path planning tasks for multi-UAV systems. However, existing DRL-based methods often suffer from performance degradation when tackling unseen scenarios, since the non-causal factors in visual representations adversely affect policy learning. To address this issue, we propose a novel representation learning approach, \ie, causal representation disentanglement, which can identify the causal and non-causal factors in representations. After that, we only pass causal factors for subsequent policy learning and thus explicitly eliminate the influence of non-causal factors, which effectively improves the generalization ability of DRL models. Experimental results show that our proposed method can achieve robust navigation performance and effective collision avoidance especially in unseen scenarios, which significantly outperforms existing SOTA algorithms.