基于贝叶斯理论的船舶沿岸航行风险预估研究

Research on Risk Estimation of Ship's Coastal Navigation Based on Bayesian Theory

海上船舶交通事故预估研究是政府职能部门、航运企业和沿岸航行船舶最关心的问题,海事专家大多采取时间序列预估法、马尔科夫预估法、神经网络预估法、灰色理论预估法和贝叶斯预估法等方法进行风险预估研究.本研究结合东营港沿海水域十几年来通行船舶总艘次和海上船舶交通事故发生的实际原因,对未来此海域船舶航行风险进行预估,重点探讨贝叶斯理论在船舶沿岸航行风险预估中的应用.模型验证表明,贝叶斯方法对船舶沿岸航行风险预估效果最贴近实际情况.

The study of maritime ship traffic accident prediction is the most concerned issue for government functional departments,shipping enterprises and coastal vessels.Most maritime experts use time series prediction method,Markov prediction method,neural network prediction method,grey theory prediction method and Bayesian prediction method to estimate risk.Combined with the actual causes of the total number of ships passing through the coastal waters of Dongying Port and the actual causes of ship traffic accidents in the coastal waters of Dongying Port in the past more than ten years,this study makes a detailed analysis and estimates the risk of ship navigation in this area in the future.This paper focuses on the application of Bayesian theory in the prediction of ship coastal navigation risk.The model verification shows that the prediction effect of Bayesian method on the risk of ship navigation along the coast is the most close to the actual situation.