This Paper:Browse 88 Download 0 |
码上扫一扫! |
A hierarchical multi-UAV cooperative framework for infrastructure inspection and reconstruction |
ChuanxiangGao1,XinyiWang1,XiChen1,BenM.Chen1 |
|
(1 Department of Mechanical and Automation Engineering, The Chinese University of Hong Kong, Hong Kong, Shatin, NT, China) |
|
摘要: |
Unmanned aerial vehicles (UAVs) are emerging as a powerful tool for inspections and repair works in large-scale and
unstructured 3D infrastructures, but current approaches take a long time to cover the entire area. Planning using UAVs for
inspections and repair works puts forward a requirement of improving time efficiency in large-scale and cluster environments.
This paper presents a hierarchical multi-UAV cooperative framework for infrastructure inspection and reconstruction to
balance the workload and reduce the overall task completion time. The proposed framework consists of two stages, the
exploration stage and the exploitation stage, resolving the task in a sequential manner. At the exploration stage, the density
map is developed to update global and local information for dynamic load-balanced area partition based on reconstructability
and relative positions of UAVs, and the Voronoi-based planner is used to enable the UAVs to reach their best region. After
obtaining the global map, viewpoints are generated and divided while taking into account the battery capacity of each UAV.
Finally, a shortest path planning method is used to minimize the total traveling cost of these viewpoints for obtaining a
high-quality reconstruction. Several experiments are conducted in both a simulated and real environment to show the time
efficiency, robustness, and effectiveness of the proposed method. Furthermore, the whole system is implemented in real
applications. |
关键词: Multi-UAV · Coverage path planning · Infrastructure inspection and reconstruction |
DOI:https://doi.org/10.1007/s11768-024-00202-0 |
|
基金项目:This work was supported in part by the Research Grants Council of Hong Kong SAR (Nos. 14209020, 14206821) and in part by the Hong Kong Centre for Logistics Robotics (HKCLR). |
|
A hierarchical multi-UAV cooperative framework for infrastructure inspection and reconstruction |
Chuanxiang Gao1,Xinyi Wang1,Xi Chen1,Ben M. Chen1 |
(1 Department of Mechanical and Automation Engineering, The Chinese University of Hong Kong, Hong Kong, Shatin, NT, China) |
Abstract: |
Unmanned aerial vehicles (UAVs) are emerging as a powerful tool for inspections and repair works in large-scale and
unstructured 3D infrastructures, but current approaches take a long time to cover the entire area. Planning using UAVs for
inspections and repair works puts forward a requirement of improving time efficiency in large-scale and cluster environments.
This paper presents a hierarchical multi-UAV cooperative framework for infrastructure inspection and reconstruction to
balance the workload and reduce the overall task completion time. The proposed framework consists of two stages, the
exploration stage and the exploitation stage, resolving the task in a sequential manner. At the exploration stage, the density
map is developed to update global and local information for dynamic load-balanced area partition based on reconstructability
and relative positions of UAVs, and the Voronoi-based planner is used to enable the UAVs to reach their best region. After
obtaining the global map, viewpoints are generated and divided while taking into account the battery capacity of each UAV.
Finally, a shortest path planning method is used to minimize the total traveling cost of these viewpoints for obtaining a
high-quality reconstruction. Several experiments are conducted in both a simulated and real environment to show the time
efficiency, robustness, and effectiveness of the proposed method. Furthermore, the whole system is implemented in real
applications. |
Key words: Multi-UAV · Coverage path planning · Infrastructure inspection and reconstruction |