Sensorimotor Coordinated Multi-UAV Coverage Path Planning

Siyuan Liu,Hongyu Fang,Ziyang Deng,Dingsheng Luo
DOI: https://doi.org/10.1109/ro-man60168.2024.10731246
2024-01-01
Abstract:We propose a novel on-line multi-UAV coverage path planning method for 3D reconstruction. UAVs play homogeneous roles in most existing multi-UAV path planning methods, where their flexibility is not fully exploited. The fact that UAVs can explore with different missions and strategies at different heights simultaneously is not well considered. Therefore, we give UAVs different identities by presenting a global-local pattern, where global UAV locates the interesting regions while local UAV explores these regions in succession. Leveraging the mechanism of sensorimotor coordination, our strategy adjusts the paths of local UAV on-line according to the input from global UAV in order to gather more information and save time. In addition, our method is compatible with real-time SLAM systems. Simulation shows that our strategy takes less time than other strategies based on area-division, and indicates that it gathers more information for better reconstruction.
What problem does this paper attempt to address?