作者: Jinfen Zhang , Ângelo P Teixeira , C. Guedes Soares , Xinping Yan , Kezhong Liu
DOI: 10.1111/RISA.12519
关键词: Dummy variable 、 Probability distribution 、 Operations research 、 Risk assessment 、 Poison control 、 Engineering 、 Bayesian network 、 Occupational safety and health 、 Bayes' theorem 、 Transport engineering 、 Port (computer networking)
摘要: This article develops a Bayesian belief network model for the prediction of accident consequences in Tianjin port. The study starts with statistical analysis historical data six years from 2008 to 2013. Then is constructed express dependencies between indicator variables and consequences. statistics expert knowledge are synthesized obtain probability distribution By sensitivity analysis, several that have influence on identified, including navigational area, ship type time day. results indicate most sensitive position where accidents occurred, followed by day length. also reflect risk port at acceptable level, despite there more room improvement. These can be used Maritime Safety Administration take effective measures enhance maritime safety