A Big Data Method For The Evaluation Of Ship Domain In Regional Historical Data

The shipping industry has long been recognized as a high-risk and high-uncertainty sector, where ship collisions represent particularly grave accidents that can lead to catastrophic consequences. Ship domain, a widely employed navigational approach, establishes a safety zone around the ship to ensure safe passage. Extensive research has been conducted on ship domain, with previous studies predominantly investigating the impact of static and dynamic ship parameters, or the effects of the ship's surrounding environment on the model. This study proposes a novel ship domain method based on regional historical data, with a primary focus on the influence of risk data on the size of the ship domain in the vessel's navigation area. In this paper, the AIS data of ship in the Zhoushan waters, China was collected, and the method of detecting possible near miss ship-ship collisions from AIS data is used to find the historical risk point data in the ship's driving area firstly. Through the historical risk point data, the density of historical ship collision accidents in the region was classified into first, secondary, and tertiary levels. The first level indicated frequent occurrences of ship accidents, thereby highlighting areas of high collision risk. Conversely, the third level represented low-frequency historical ship accidents, signifying a low risk of collision. Finally, the shape and size of the ship domain model were determined based on the hierarchical index of historical ship collision accidents. Simulation experiments confirmed the effectiveness of the proposed method, indicating that it can provide accurate estimations of collision risks in ship encounter scenarios.