Collision avoidance path planning in multi-ship encounter situations

Collision avoidance path planning is still one of the essential problems in the design and application of an intelligent maritime navigation system. Its main obstacle is how to determine effective and cooperative collision avoidance maneuvers within a multi-ship encounter situation. By deconstructing a multi-ship encounter, this study adopted ship domain around target ships to assess the collision danger that own ship should avoid. Subsequently, the fitness function that has multiple dynamic obstacle constraints was designed in a two-dimensional map. Based on DE algorithm, a path-planning method was developed to compute collision-free and optimal navigation paths for ships. Simulation results show that the algorithm can generate a safe and suitable path from each perspective in a multi-ship encounter. The results also validate the practicality of the generated paths, consistency of the algorithm outputs and performance of the algorithm. It would be expected to provide a reference for collision avoidance decision making as well as contribute to the development of autonomous navigation systems.