A Study on Multi-ship Avoidance System for Unmanned Surface Vehicles Using the Quaternion Ship Domain and Collision Risk Index

Collision avoidance is essential for the safe navigation of unmanned surface vehicles (USVs). This paper proposes a multi-ship collision avoidance algorithm that integrates the quaternion ship domain (QSD) and collision risk index (CRI) to assess collision risk and generate optimal avoidance trajectories. The algorithm was used to evaluate target ships (TS) using the CRI values, adapt the ship domain based on QSD, and use the velocity obstacle (VO) method to determine feasible avoidance paths while adhering to the International Regulations for Preventing Collisions at Sea (COLREGs). Through 14 simulation scenarios, including single-ship, two-ship, and multi-ship encounters with unexpected TS maneuvers, the results revealed the ability of the algorithm to maintain safe separation distances in complex maritime traffic. The CRI-based approach allows real-time risk assessment and adaptive timing, while QSD adjusts dynamically based on the ship speed and encounter conditions, minimizing unnecessary maneuvers. This integration of QSD, CRI, and VO ensures COLREGs-compliant navigation, efficient path deviation management, and enhanced safety, contributing to the development of intelligent USV navigation systems.