Probabilistic quantification of ship collision risk considering trajectory uncertainties

Abstract

This paper presents the probabilistic quantification of collision risk for surface ships to safely navigate in ocean environments. A semi-analytical approach is proposed which allows an accurate and fast evaluation of collision probability considering time-varying trajectory uncertainties of both the navigating vehicle and the other ships. For efficient computation, the collision probability is separated into diffusion and drift components based on the concept of probability flow, and these two probability components are combined in order to obtain the instantaneous collision probability (ICP). To demonstrate the validity of the proposed approach, numerical simulations were performed for marine traffic scenarios and the feasibility was assessed by field experiment data obtained in a real-sea environment. ？ 2016