A novel collaborative collision avoidance decision-making methodology based on potential collision areas for intelligent navigation

Abstract

Ship navigation safety is a crucial component of intelligent navigation, and collaborative collision avoidance plays a vital role in ensuring safety and conflict resolution by enhancing decision-making efficiency in multi-ship encounters. In this paper, a collaborative collision avoidance model is proposed to connect encounter risk and decision-making in a more intuitive way. It pioneers the quantification of collision risk based on potential collision areas and differentiates encountered ships, and a collaborative mechanism is further constructed to formulate a collision avoidance decision-making model. International Regulations for Preventing Collisions at Sea (COLREGs) are comprehensively considered by differentiating risk relationships among ships and specify their roles and responsibilities. A ship collaboration mechanism under action intention games is constructed based on rational thinking to form a decision-making model with game-decision cycles. The results demonstrate that the model can meet the safety requirements in case studies, providing a rational reflection, and accurately determines encounter stages. The results also indicate that ship differentiation and role assignment can adapt to abnormal actions. This research makes significant contributions to ship decision support in term of better collaboration among ships while reducing action conflicts, promoting the development of intelligent navigation technologies.