Hierarchical Decision-Action Strategy (HDAS) for UUV mission planning in ocean environments

Mission planning for an unmanned underwater vehicle (UUV) in time-varying ocean currents is challenging because it requires tightly coupling discrete visit sequencing with continuous, obstacle-aware trajectory generation under current-aware planning constraints. This paper proposes a Hierarchical Decision-Action Strategy (HDAS) that tightly couples discrete visit-sequence optimization with continuous, current-aware trajectory generation. The Decision Layer employs enhanced Ant Colony Optimization (EACO) with a local search operator to construct and refine mission sequences, while the Action Layer uses Particle Swarm Optimization (PSO) to optimize B-spline trajectories subject to obstacle and safety constraints. The framework represents time-varying ocean-current conditions using a vortex-based flow model, and its modular design readily accommodates alternative metaheuristics. Across two ocean-flow scenarios, namely the present-state ocean-current field and the five-hour-ahead evolved ocean-current field, HDAS achieves faster convergence, higher computational efficiency, and lower mission time than the selected representative baselines. These results demonstrate that HDAS provides an effective framework for current-aware UUV mission planning in complex ocean environments.