.png){kind=icon}
.png){kind=icon}
Return
Collaborative 3D Continuous Trajectory Optimization Method for Multi-UAV Post-Disaster Emergency Exploration
.png){kind=icon}
PRE
AI
DOI：10.1142/S0218126626440081.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
With the rapid advancement of low-altitude network technologies, UAV-based regional monitoring and post-disaster exploration have gained widespread application. However, existing methods often suffer from limited flexibility, discrepancies between theoretical assumptions and practical demands, and insufficient generalization in complex disaster scenarios. To address these challenges, this paper investigates multi-UAV collaborative search tasks in 3D continuous space. A simulation environment is constructed based on real-world scenarios, where the disaster area is gridded and pre-disaster prior importance weights are introduced. By incorporating camera Field-of-View constraints and an altitude-dependent clarity model, a dynamic reward assessment mechanism is established. The multi-UAV collaborative exploration process is then modeled as a Markov Decision Process, and the Multi-Agent Twin Delayed Deep Deterministic Policy Gradient algorithm is employed to optimize the collaborative trajectories. Simulation results demonstrate that the proposed method achieves an improvement of approximately 23.94% to 496.99% in average effective information gain compared to baseline methods, validating its effectiveness in rapid post-disaster situational awareness and exploration path planning.
Keywords:
UAV
post-disaster exploration
trajectory optimization
deep reinforcement learning
multi-agent coordination
Journal
IF:
1
Papers:
336
Citations:
2.3K