Table 7.

Single intelligence algorithms (USVs).

Problem	Resolution	Performance and Additional Explanation	Ref.
USV
Local obstacles avoiding	LROABRA (local reactive obstacle avoidance based on region analysis)	Radar, binocular vision, stereo vision, monocular vision, infrared cameras, and laser range finders are used. Stability of LROABRA is better than that of OAABHW High-speed (≥20 knots) USVs	[132]
Fast long-distance ship route planning	Multiscale visibility graph (VG) method	The number of visibility points can be reduced by half, and the VG search time can be shortened The local planning window (LPW) plays a role in greatly reducing the complexity of the VG model. Plan routes by simplifying the map using convex points of the obstacle polygon	[133]
Obstacle avoidance	Improved VFH algorithm	Partial encounter geometry model also used. Achieving collision avoidance in compliance with the international regulation COLREGSPerforming collision avoidance measures in a water environment with sudden and dynamic obstacles. Uses the CRI values of the obstacles as key parameters in the histogram and removing the grid model to speed up calculations and improve thresholds	[134]
Obstacle avoidance	Improved ant colony optimization (IACO) algorithm	Risk avoidance from steering during high-speed navigation in real and dynamic environments Implement and simulate static unknown environments and dynamic known environments (convergence, real-time performance, and stability of the improved ACO) in the cross-platform framework.	[135]
Obstacle avoidance	Genetic collision avoidance algorithm	Search ability, convergence speed, and local optimum are improved compared to ACO. Can effectively avoid multiple obstacles coming from different directions and conditions DCPA (distance of closest point of approach), TCPA (time of closest point of approach) are used. Simulation data such as the distance between the ASV and the obstacle vessel indicate that the collision avoidance behavior is safe and verify the feasibility of the proposed genetic collision avoidance algorithm.	[136]
Obstacle avoidance	Fuzzy inference algorithm	Long-range lidar, radar, and camera-based tracking technologies are used. Effective autonomous navigation and anticollision capacity Aragon USV (8 m) Calculation of fuzzy inference algorithm using TCPA and DCPA	[137]