Abstract
If ship navigators could access precise collision risk data and guidance for avoidance, safer navigation would be feasible. This study introduces a method for devising avoidance routes based on quantitative collision risk assessment. In particular, we propose an improved isochrone method, a cell-free approach akin to real ship navigation routes. The method uses a novel heuristic function to ensure effective collision avoidance, even in scenarios where conventional heuristic algorithms fail. Through verification cases, we demonstrate that the proposed method successfully avoided collisions, even in situations with unexpected lined-up ships, reducing maximum collision risk by up to 94.0%. Applications of the proposed method with actual AIS (Automatic Identification System) data yield simulation results with successful collision avoidance route planning. Comparing the proposed method to avoidance routes using other existing methods and to the actual ship’s route, the proposed method was able to plan a route with a 37.5% improvement in maximum collision risk and a 13.6% improvement in average collision risk.
