Work

Multi-AUV Path Planning Based on Hierarchical Opponent Multiactor-Attention-Critic Reinforcement Learning Framework