This study proposes a spatial modeling methodology for optimal ship navigation in coastal areas with high risks of collision with static obstacles. To precisely represent the complex topology of coastal areas, the spatial environment is modeled using ...
This study proposes a spatial modeling methodology for optimal ship navigation in coastal areas with high risks of collision with static obstacles. To precisely represent the complex topology of coastal areas, the spatial environment is modeled using a high-resolution graph structure, with each node weight defined as the risk of collision with static obstacles. This approach guides the A* algorithm, a shortest path algorithm, to determine optimal routes that are not only shorter in travel distance but also lower in navigational risk. The proposed spatial modeling utilizes ship trajectory data to calculate the risk of collision with static obstacles, and generates different graphs based on ship characteristics, thereby expressing the risks that vary according to the ship types or tonnage. The test results of the optimal ship navigation with the proposed graph-based spatial modeling have been evaluated both quantitatively and qualitatively, demonstrating that the proposed spatial representation is proper to the fast optimal ship navigation. Furthermore, the study demonstrates that it can provide safer optimal routes in coastal areas compared to those determined by the traditional A* algorithm.