Regional Ship Collision Risk Assessment: An Integrated Approach Using Velocity Obstacle and Complex Network

Regional ship collision risk identification and assessment is significant for reducing collision accidents and improving maritime transport management. To improve maritime safety management in busy waters, we propose a novel regional ship collision risk assessment approach with the integration of Velocity Obstacle (VO) and complex network in this paper. Firstly, a density-based spatial clustering method is utilized to classify the ships into different clusters, and then a comprehensive collision risk index is proposed. This index combines VO-based time-varying collision risk and network-based complexity measurement of ship encounter situations to estimate the collision risk for single ships. Finally, aiming at assessing the collision risk from multiple perspectives, the Shapley value method is introduced to obtain the macro-regional collision risk by measuring the collision risk contribution of each ship and cluster. To test the performance and feasibility of the proposed approach, a case study was conducted in Zhoushan, China using historical AIS data. The results show that the proposed approach can accurately identify ships with high collision risk and evaluate the spatiotemporal distribution characteristics of collision risk in selected regions, thereby improving the safety management capabilities of maritime surveillance operators.