Improving Generalization in Collision Avoidance for Multiple Unmanned Aerial Vehicles via Causal Representation Learning

Lin, Che; Han, Gaofei; Wu, Qingling; Wang, Boxi; Zhuang, Jiafan; Li, Wenji; Hao, Zhifeng; Fan, Zhun

doi:10.3390/s25113303

This is an early access version, the complete PDF, HTML, and XML versions will be available soon.

Open AccessArticle

Improving Generalization in Collision Avoidance for Multiple Unmanned Aerial Vehicles via Causal Representation Learning^†

by

Che Lin

^1,‡

,

Gaofei Han

^1,‡,

Qingling Wu

¹,

Boxi Wang

¹,

Jiafan Zhuang

^1,*,

Wenji Li

¹,

Zhifeng Hao

¹ and

Zhun Fan

^2,*

¹

Department of Electronic Engineering, Shantou University, Shantou 515063, China

²

Shenzhen Institute for Advanced Study, University of Electronic Science and Technology of China, Shenzhen 518000, China

^*

Authors to whom correspondence should be addressed.

^†

This paper is an extension version of the conference paper: Han, G.; Wu, Q.; Wang, B.; Lin, C.; Zhuang, J.; Li, W. Deep Reinforcement Learning Based Multi-UAV Collision Avoidance with Causal Representation Learning. In Proceedings of the 2024 10th International Conference on Big Data and Information Analytics (BigDIA), Chiang Mai, Thailand, 25–28 October 2024.

^‡

These authors contributed equally to this work.

Sensors 2025, 25(11), 3303; https://doi.org/10.3390/s25113303 (registering DOI)

Submission received: 7 April 2025 / Revised: 20 May 2025 / Accepted: 22 May 2025 / Published: 24 May 2025

(This article belongs to the Special Issue Autonomous Robotics and Unmanned Aerial Vehicles: Addressing Technical Challenges and Advancing Innovation)

Download Versions Notes

Abstract

Deep-reinforcement-learning-based multi-UAV collision avoidance and navigation methods have made significant progress. However, the fundamental challenge of those methods is their restricted capability to generalize beyond the specific scenario in which they are trained on. We find that the cause of the generalization failures is attributed to spurious correlation. To solve this generalization problem, we propose a causal representation learning method to identify the causal representations from images. Specifically, our method can neglect factors of variation that are irrelevant to the deep reinforcement learning task through causal intervention. Subsequently, the causal representations are fed into the policy network for action prediction. Extensive testing reveals that our proposed method exhibits better generalization results compared to state-of-the-art methods in different testing scenes.

Keywords: deep reinforcement learning; multi-UAV collision avoidance; generalization failure; causal representation learning; causal intervention

Share and Cite

MDPI and ACS Style

Lin, C.; Han, G.; Wu, Q.; Wang, B.; Zhuang, J.; Li, W.; Hao, Z.; Fan, Z. Improving Generalization in Collision Avoidance for Multiple Unmanned Aerial Vehicles via Causal Representation Learning. Sensors 2025, 25, 3303. https://doi.org/10.3390/s25113303

AMA Style

Lin C, Han G, Wu Q, Wang B, Zhuang J, Li W, Hao Z, Fan Z. Improving Generalization in Collision Avoidance for Multiple Unmanned Aerial Vehicles via Causal Representation Learning. Sensors. 2025; 25(11):3303. https://doi.org/10.3390/s25113303

Chicago/Turabian Style

Lin, Che, Gaofei Han, Qingling Wu, Boxi Wang, Jiafan Zhuang, Wenji Li, Zhifeng Hao, and Zhun Fan. 2025. "Improving Generalization in Collision Avoidance for Multiple Unmanned Aerial Vehicles via Causal Representation Learning" Sensors 25, no. 11: 3303. https://doi.org/10.3390/s25113303

APA Style

Lin, C., Han, G., Wu, Q., Wang, B., Zhuang, J., Li, W., Hao, Z., & Fan, Z. (2025). Improving Generalization in Collision Avoidance for Multiple Unmanned Aerial Vehicles via Causal Representation Learning. Sensors, 25(11), 3303. https://doi.org/10.3390/s25113303

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Menu

Improving Generalization in Collision Avoidance for Multiple Unmanned Aerial Vehicles via Causal Representation Learning^†

Abstract

Share and Cite

Article Metrics

Article Access Statistics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI

Article Menu

Improving Generalization in Collision Avoidance for Multiple Unmanned Aerial Vehicles via Causal Representation Learning †

Abstract

Share and Cite

Article Metrics

Article Access Statistics

Further Information

Guidelines

MDPI Initiatives

Follow MDPI

Improving Generalization in Collision Avoidance for Multiple Unmanned Aerial Vehicles via Causal Representation Learning^†