Ship collision avoidance behaviour recognition and analysis based on AIS data

A novel approach is proposed to automatically identify the ship collision avoidance behaviour from ship trajectories based on an improved Sliding Window Algorithm. The approach has three main stages: (1) determining the ships' obligations according to the Convention on the International Regulations for Preventing Collision at Sea (COLREGs); (2) assessing the rudder angle according to the true bearing of the target ship; (3) identifying the corresponding ship handling behaviour from the ship trajectory taking into account the Rate of Turn and its derivative based on the Sliding Window Algorithm. Four typical encounter scenarios are studied to demonstrate the feasibility and effectiveness of the proposed method. The approach is then applied to near collision scenarios identified in the maritime traffic off the continental coast of Portugal, which allowed the characterization of the relative and spatial distributions of the locations at which the ships take evasive manoeuvres. The results show that the approach can be applied to accurately detect the ship collision avoidance behaviour from Automatic Identification System trajectory data, and the characterization of the collision avoidance behaviour can potentially be used by situational awareness systems and as the basis for ship collision avoidance decision-making.