RESEARCH ON SHIP COLLISION RISK ASSESSMENT USING REAL DATA AND HIERARCHICAL CLUSTERING ALGORITHM

Abstract

Ship collisions in maritime accidents is one of the most serious incidents, causing significant consequences for both people and property. Collisions often occur due to various reasons, primarily due to the complacency and lack of attention of the navigators, poor weather conditions, or technical failures. When two or more vessels collide, it not only results in damage to the ships but can also lead to accompanying incidents such as oil spills, environment pollution, and impacts on maritime safety. To minimize risks, compliance with maritime regulations, improving monitoring technology, and training crew members are essential. In this paper, we identify and analyze the factors that may lead to collision incidents between nearby groups of vessels, and subsequently propose effective preventive measures. By collecting real-time vessel position data from maritime traffic monitoring systems in the surveyed area and applying hierarchical clustering algorithms to classify specific high-risk situations and conditions, that can assess and provide necessary forecasts and recommendations for vessel groups that are likely to collide.