신경망을 이용한 유조선 기름 유출사고에 따른 환경비용 추정에 관한 연구

The accident risks in the marine environment are increasing because of the tendency to build faster and larger ships. To secure ship safety, risk-based ship design (RBSD) was recently suggested based on a formal safety assessment (FSA). In the process of RBSD, a ship designer decides which risk reduction option is most cost-effective in the design stage using a cost-benefit analysis (CBA). There are three dimensions of risk in this CBA: fatality, environment, and asset. In this paper, we present an approach to estimate the environmental costs based on the size of an oil tanker involved in an accident using a neural network. An appropriate neural network model is suggested for the estimation,and the neural network is trained using IOPCF data. Finally,the learned neural network is compared with the cost regression equation by IMO MEPC 62/WP.13 (2011).