Research Of Risk Zoning For Ship Oil Spill Based On Bayesian Network

The ocean is a natural resource for human survival,with the rapid development of international economy and trade,a substantial increase in the volume of trade in the 21st century,land transport alone is not sufficient to meet the transport needs of the goods.Shipping with its advantage of low freight and large volumes,take the main task of China's goods importing and exporting.However,with the trade volume increasing,the trend of large ships at sea,more and more problems emerge.A series of oil spills accidents not only caused huge economic losses,but also has brought a serious impact to the marine environment,fisheries and tourism.Accident of ship oil spill has become one of those factors which lead to most seriously marine pollution.Therefore it bears both the academic and practical significance for protecting the sea environment to study risk assessment of ship oil spill and preventive measures,especially at the stage of the installation of ship oil spill response equipment for ports and formulating an contingency planning for ship oil spill.Based on the previous research work,this paper analyzes the research methods of various types of ship oil spill risk zoning,and finally selects the multidisciplinary theory Bayesian network which based on risk engineering,probability theory and system engineering,and has strong data mining and reasoning ability.By analyzing the ship oil spill accidents in the coastal areas at home and abroad in recent years,the influencing factors of the oil spill accidents are obtained,and the causal relationship between the factors is determined,and the Bayesian network structure is constructed.Combined with historical data and prior knowledge to obtain conditional probability table,to determine the final ship oil spill Bayesian network model.Based on the reasoning results,the risk level of ship oil spill and the sensitivity of influencing factors were obtained by using GeNIe software to model the simulation.Finally,the network model is applied to the Bohai area.We divide Bohai into four blocks,the risk level and influence sensitivity of each region are calculated.The study shows that it is feasible to use the Bayesian network method to carry out the risk division of ship oil spill,which can provide theoretical support for ship safety management.