Quantifying ship-bridge collision risk through extreme value theory: a new methodological framework

Ship-bridge collisions pose significant safety risks, yet traditional risk assessment methods often struggle due to the scarcity of accident data and reliance on restrictive distributional assumptions. To address these limitations, this study proposes a novel risk assessment framework that integrate...

Descripción completa

Detalles Bibliográficos
Autores: Grifoll Colls, Manel|||0000-0003-4260-6732, Chen, Xin, Zhou, Hongzhu, Zhou, Yusheng|||0000-0002-8795-3423, Liu, Jiao, Zheng, Pengjun, Li, Meng
Tipo de recurso: artículo
Fecha de publicación:2026
País:España
Institución:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglés
OAI Identifier:oai:upcommons.upc.edu:2117/452355
Acceso en línea:https://hdl.handle.net/2117/452355
https://dx.doi.org/10.1016/j.oceaneng.2025.124023
Access Level:acceso abierto
Palabra clave:Merchant marine--Safety measures
Collisions at sea
Ship-bridge collision
Surrogate safety measure
Extreme value theory
Risk assessment
Abordatges
Marina mercant--Mesures de seguretat
Àrees temàtiques de la UPC::Nàutica::Seguretat marítima::Accidents marítims
Descripción
Sumario:Ship-bridge collisions pose significant safety risks, yet traditional risk assessment methods often struggle due to the scarcity of accident data and reliance on restrictive distributional assumptions. To address these limitations, this study proposes a novel risk assessment framework that integrates Extreme Value Theory (EVT) with Surrogate Safety Measures (SSM). Three indicators are constructed to quantify potential collision risk: Spatial Clearance (SC), Trajectory Deviation (TD), and Time of Risk Exposure (TRT). Collectively, these indicators capture both static and dynamic dimensions of collision risk, ranging from spatial proximity to navigational behavior beyond the safe channel. To identify high-risk trajectories, the framework employs both the Block Maxima (BM) and Peak-Over-Threshold (POT) models from EVT. Comparative analysis reveals that the POT model offers more efficient data utilization and higher precision in identifying high-risk cases. Among the indicators, SC demonstrates the highest identification precision (98.53 %), while TRT, although yielding slightly lower precision (95.32 %), provides valuable complementary insights into dynamic risk conditions. Overall, this integrated approach significantly improves the accuracy and practicality of ship-bridge collision risk identification. It also offers important technical support for intelligent maritime traffic management systems and the real-time monitoring of bridge-area safety.