JEMS Maritime Sci. Ahead of Print: JEMS-34603

Prediction of ship trajectory and critical collision zone in sunda strait based on automatic identification system and weather data

Iis Dewi Ratih, Mochammad Reza Habibi, Kanugrahing Christy Sekar Arum
Department of Business Statistics, Institut Teknologi Sepuluh Nopember, Surabaya, Indonesia

Indonesia as the largest maritime country, which has dense shipping activities that increase the risk of ship accidents, especially in strategic areas such as the Sunda Strait. Extreme weather, such as storms and strong winds, increase this risk and require special attention to improve shipping safety. This study aims to identify high-risk areas for ship collisions in the Sunda Strait, known as the Critical Collision Zone (CCZ). The CCZ is determined through ship trajectory prediction analysis using the Bi-GRU method and clustering with the DBSCAN algorithm. Trajectory data is obtained from AIS information and weather data. AIS data includes the position, speed, and direction of the ship in real time. Its integration with weather data allows for the formation of a more accurate trajectory. After the CCZ is identified, the probability of a collision is calculated using the Monte Carlo Simulation (MCS) method. The results show that the weather data-based prediction model performs better in identifying the CCZ, indicated by lower MAE and MSE values and higher silhouette coefficients, which improve the accuracy of identifying risky areas and estimating the probability of ship collisions in the Sunda Strait.

Keywords: Automatic Identification System (AIS), Critical Collision Zone (CCZ), Ship Collision, Ship Trajectory, Weather

---

CITE