Developing a Ship Collision Risk Assessment Model with Internal and External Factors: Focused on South Korea Maritime Environment

Abstract

Maritime collisions pose significant risks, prompting the need for robust risk assessment models to enhance safety measures. This study endeavors to develop a comprehensive ship collision risk model reflecting the intricate marine traffic environment in South Korea. Through a survey of experienced maritime personnel and a random forest model analysis, an evaluation model integrating internal and external factors was devised. Internal factors were determined through conjoint analysis, emphasizing encounter relationships, separation distance, and vessel speed. External risk factors were established using a random forest model based on historical collision data. The model’s efficacy was then applied to and validated in the vicinity of Busan Port, a region with complex marine traffic. The resulting risk map highlighted high-risk areas, offering valuable insights for risk management and policy formulation. This model provides a foundational framework for maritime safety policy decisions, representing a significant contribution to collision risk assessment methodologies.