Maritime transportation plays a vital role in global trade, yet increasing congestion and complexity in waterways elevate the risk of accidents, underscoring the need for advanced ship trajectory prediction techniques.

Effective trajectory prediction is essential for preventing collisions, optimizing routes, and enhancing overall navigation safety and efficiency.

Path planning is crucial for the autonomous navigation of Unmanned Surface Vehicles (USVs), impacting their efficiency, safety, and mission success.

As USVs become increasingly significant in maritime studies, advancements in communication, navigation, and artificial intelligence are driving their development.

Various algorithms have been developed for route planning, including bio-inspired algorithms, graph-based A* algorithms, and artificial potential field methods, each with its strengths and weaknesses.

The artificial potential field method (APF) is favored for its real-time performance, although it faces limitations such as local minimum traps and complexities in path execution.

Recent research proposes an innovative algorithm that integrates the APF with Deep Q-Networks (DQNs), enhancing path planning by considering vessel dynamics and environmental variability.

Studies incorporating DQNs have shown promise in various maritime tasks, although the design of the reward function remains a significant challenge.

The performance of the newly proposed Mamba model is being compared against benchmark models like LSTM and Transformer to ensure a fair assessment under uniform parameter settings.

Testing of the Mamba model using AIS data from the Beijing–Hangzhou Canal demonstrates its capability to predict vessel behavior in complex waterways.

Experimental results indicate that the MAPF-DQN algorithm outperforms traditional methods in terms of safety, efficiency, and path feasibility.

With over 80% of naval accidents attributed to human factors, intelligent automated navigation systems can significantly reduce maritime accidents by minimizing human error and enhancing safety.