Multi-ship collision avoidance decision-making method under complex encounter situations

This work addresses the issue of multi-ship collision avoidance decision-making complex encounter situations, and proposes a novel velocity varying-steering collision avoidance method based on an improved particle swarm optimization (IPSO) algorithm. The proposed method establishes a limited range based on the International Regulations for Preventing Collisions at Sea (COLREGs) and creates a multi-objective model, in which the collision risk of ships, the energy loss caused by velocity varying and the voyage loss caused by steering are taken into account. To obtain an optimal solution of the multi-objective model, an IPSO is introduced to determine the feasible solution domain for ship collision avoidance decision making(CADM). The proposed CADM is validated by numerical simulations and navigation simulator. The results indicate that the recommended velocity and course can effectively remove the risk of collision between the ship and target ships.