Table 9.

Multi robots and swarm intelligence algorithms (USVs).

Problem	Resolution	Performance and Additional Explanation	Ref.
USV
Underwater cooperative navigation techniques	SFE algorithms	Navigate with frame providing spatial density of plastics over sea. Differential evolution algorithm for control SFE algorithm is better suited for plastic collection than is ACO Development of SFE algorithm based on stigmergy and flocking for marine plastic collection	[166]
Obstacle avoidance	Combining restricted A* algorithms	Path planning by a constrained A* algorithm leader–follower formation control Maneuverability that allows for improved path-following performance for navigation and reduction of cross-track errors All followers are affected by the leader and all other USVs in the group, which is also applicable to UAVs Combining a limited A* algorithm using an artificial potential field based on USV various maneuvering response time capabilities	[167]
Obstacle avoidance	APF-DQN (artificial potential field-deep Q-learning network)	N: local dynamic path planning G: APF-DQN C: Markov decision process Performance of DRL-based method works better on the global trajectory A deep reinforcement learning and artificial potential field (APF)-based path planning method that complies with the International Regulations for Preventing Collisions at Sea (COLREGS) rules. Improvement of action space and reward function of a deep Q-learning network (DQN) by utilizing the APF method Eliminate USV with known local dynamic environment information Solve collision path planning challenge	[168]