Risk-aware adaptive APF-RRT* path planning method for complex seabed terrain

The terrain slope of deep-sea mining areas varies drastically, and the operation path planning of mining vehicles is crucial for reducing energy consumption and ensuring safety. When applied to deep-sea environments, existing RRT-based algorithms have two limitations: Firstly, they fail to fully account for the influence of ocean current disturbances on path feasibility; secondly, the random sampling mechanism tends to fall into local optima. To address these issues, this paper proposes a risk-aware adaptive APF-RRT* algorithm that enhances search performance through four core mechanisms: (1) Hybrid sampling mechanism; (2) Adaptive potential field parameter updating mechanism; (3) APF expansion and node generation mechanism; (4) Risk consistency cost optimization mechanism. Simulation results demonstrate that the proposed algorithm outperforms A*, RRT, RRT*, and Informed-RRT* algorithms in both path quality and stability across three terrains of varying complexity. Furthermore, this paper additionally fuses ocean current factors with terrain risk to establish a deep-sea mining vehicle path planning model that accounts for static flow field disturbances, thereby verifying the robustness of the proposed algorithm under marine environmental perturbations.