Next Article in Journal
UTUAV: A Drone Dataset for Urban Traffic Analysis
Previous Article in Journal
MCB-RT-DETR: A Real-Time Vessel Detection Method for UAV Maritime Operations
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Drones
Volume 10
Issue 1
10.3390/drones10010014
drones-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
This is an early access version, the complete PDF, HTML, and XML versions will be available soon.
Article

Generalization and Exploitation: Meta-GSAC for Multi-Task UAV Path Planning and Obstacle Avoidance

by
Jingyi Huang
1,
Shuangxia Bai
2,
Liangliang Huai
3,
Yujie Cui
1,
Bo Li
1,* and
Kaifang Wan
1
1
School of Electronics and Information, Northwestern Polytechnical University, 127 Youyi West Road, Xi’an 710072, China
2
College of Computing, City University of Hong Kong, 83 Tat Chee Avenue, Kowloon Tong, Hong Kong Special Administrative Region, China
3
Leihua Electronic Technology Research Institute, Aviation Industry Corporation of China, Wuxi 641100, China
*
Author to whom correspondence should be addressed.
Drones 2026, 10(1), 14; https://doi.org/10.3390/drones10010014 (registering DOI)
Submission received: 4 November 2025 / Revised: 24 December 2025 / Accepted: 25 December 2025 / Published: 27 December 2025
(This article belongs to the Section Artificial Intelligence in Drones (AID))
Versions Notes

Abstract

Deep reinforcement learning (DRL) is extensively applied in autonomous unmanned aerial vehicle (UAV) control yet faces critical challenges regarding adaptability and generalization in dynamic environments. To address these limitations, this paper proposes the Meta Gated Transformer-XL Soft Actor-Critic (Meta-GSAC) algorithm. This framework integrates a Gated Transformer-XL module to capture long-term temporal dependencies from multimodal inputs and incorporates the Reptile algorithm to facilitate multi-task meta-learning. Experimental results demonstrate that Meta-GSAC significantly outperforms standard baselines. Notably, it achieves optimal policy convergence with approximately 50% fewer training epochs while effectively eliminating the high-frequency control oscillations observed in the GSAC baseline. Moreover, the proposed method exhibits superior few-shot adaptation capabilities, enabling the UAV to rapidly adapt to novel task scenarios with minimal gradient updates.
Keywords: meta-learning; attention mechanism; few-shot; UAV control decision; reinforcement learning generalization; multi-task training meta-learning; attention mechanism; few-shot; UAV control decision; reinforcement learning generalization; multi-task training

Share and Cite

MDPI and ACS Style

Huang, J.; Bai, S.; Huai, L.; Cui, Y.; Li, B.; Wan, K. Generalization and Exploitation: Meta-GSAC for Multi-Task UAV Path Planning and Obstacle Avoidance. Drones 2026, 10, 14. https://doi.org/10.3390/drones10010014

AMA Style

Huang J, Bai S, Huai L, Cui Y, Li B, Wan K. Generalization and Exploitation: Meta-GSAC for Multi-Task UAV Path Planning and Obstacle Avoidance. Drones. 2026; 10(1):14. https://doi.org/10.3390/drones10010014

Chicago/Turabian Style

Huang, Jingyi, Shuangxia Bai, Liangliang Huai, Yujie Cui, Bo Li, and Kaifang Wan. 2026. "Generalization and Exploitation: Meta-GSAC for Multi-Task UAV Path Planning and Obstacle Avoidance" Drones 10, no. 1: 14. https://doi.org/10.3390/drones10010014

APA Style

Huang, J., Bai, S., Huai, L., Cui, Y., Li, B., & Wan, K. (2026). Generalization and Exploitation: Meta-GSAC for Multi-Task UAV Path Planning and Obstacle Avoidance. Drones, 10(1), 14. https://doi.org/10.3390/drones10010014

Article Metrics

Article metric data becomes available approximately 24 hours after publication online.
Back to TopTop