.png){kind=icon}
.png){kind=icon}
Return
Adaptive Coverage Path Planning for Multiple Fixed-wing UAVs With Learning-based Initialization
.png){kind=icon}
PRE
AI
DOI：10.1007/s12555-025-0263-7.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This study presents a real-time feasible and dynamically adaptive coverage path planning (CPP) method for multiple fixed-wing unmanned aerial vehicles (UAVs), specifically designed to address the challenges of unstructured and mission conditions requiring rapid re-planning. To support efficient re-planning in such conditions, the proposed algorithm follows a two-stage framework consisting of region partitioning and path planning. A learning-based network is employed to generate high-quality initial solutions for region partitioning, improving computational efficiency within the baseline algorithm. In addition, a novel method is introduced to generate adaptive paths that respect the dynamic constraints of fixed-wing UAVs, thereby enhancing the practical applicability of CPP in real-world scenarios. The proposed method is validated through simulations under time-constrained conditions, demonstrating both computational efficiency and reduced risk. Experimental results further confirm its feasibility for field deployment.
Keywords:
Fixed-wing unmanned aerial vehicle
multi-agent coverage path planning
robot learning
Journal
IF:
2.9
Papers:
216
Citations:
6.5K