.png){kind=icon}
.png){kind=icon}
Return
A hydrodynamic-aware hierarchical path planning framework for multi-AUV systems in realistic ocean currents
.png){kind=icon}
OA
AI
X
D
刘
G
S
Y
S
DOI：10.1016/j.apor.2026.105086.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Energy-efficient and collision-free navigation for multi-AUV systems remains a significant challenge in ocean environments with strong and spatially heterogeneous currents. To address this issue, this paper proposes a hydrodynamic-aware hierarchical path-planning framework that is designed to exploit environmental flow features. First, a realistic 3D environment is constructed using empirical data from the Hybrid Coordinate Ocean Model (HYCOM) to represent authentic hydrodynamic resistance. Recognizing that standard algorithms often struggle in such complex vector fields, a hierarchical refinement strategy is introduced to systematically decouple the optimization process. The initialization stage employs Latin hypercube sampling (LHS) to ensure broad and stratified search coverage. Subsequently, a coarse-to-fine evolution mechanism is applied. Specifically, a flow-aware 2-opt operator is used to rectify topological entanglements caused by current drift, while a hybrid escape mechanism integrating helical perturbations with simulated annealing facilitates escape from vortex-induced local minima. Furthermore, gradient-adaptive domain scaling is employed for precise trajectory refinement in high-shear regions. Finally, kinematic feasibility is strictly enforced through cubic B-spline interpolation. Simulated results in realistic ocean-current environments show that the proposed framework achieves favorable trade-offs among route efficiency, energy-related cost, and safety compared with the selected baselines.
Keywords:
Multi-AUV systems
Hydrodynamic-aware path planning
Energy efficiency
Hierarchical optimization strategy
Realistic ocean currents
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
4.4
Papers:
4.0K
Citations:
1.3W
Organization
Cited Papers