“…Liu and Shi (2005) used neural networks and fuzzy set theory to assess the most threatening target ship and generate a corresponding collision-avoidance maneuver action in accordance with COLREGS, encounter forms, and the speed ratio and bearing of the ship. Tam and Bucknall (2013) effectively solved for an optimal collision avoidance path under in a multi-ship encounter situation, in accordance with the COLREGS knowledge base, the maneuvering characteristics of the ships, and the ship motion equations, but this study assumed that data on the mechanical properties and hydrodynamic characteristics (propulsion, resistance, radius of gyration, and stopping distance) of all the ships are known.…”