中文摘要：

通过对港口船舶靠港装卸作业溢油相关统计资料和文献的分析,识别船舶靠港装卸作业溢油风险的影响因素,并根据各因素的因果关系构建船舶靠港装卸作业溢油风险贝叶斯网络拓扑结构,确定各节点的取值范围;综合运用贝叶斯条件概率模型及三角模糊数处理方法,获得各节点变量的条件概率;在此基础上利用GeNIe软件进行概率推理和风险因素灵敏度分析;最后,以我国沿海四个港口为例进行了实际评价.研究结果表明运用贝叶斯网络模型评价船舶靠港装卸作业溢油风险问题是合理和有效的,不仅可为港口和船舶安全管理等制度的制定提供理论依据,也验证了应用贝叶斯网络评价船舶靠港装卸作业溢油风险的方法是一种科学有效的方法.

英文摘要：

This article firstly identifies risk factors based on analysis on relative data and literatures,and a Bayesian network structure diagram and the nodes＇ value ranges are established according to the causality of nodes variables;Secondly,conditional probability of each variable in the Bayesian network is acquired by combining with Bayesian conditional probability model and the triangular fuzzy number approach;Based on this,the software GeNIe is used to compute accident probability of ship oil spill and to analyze sensitivity of risk factors;Finally,through the case study of our country＇ s coastal ports,the results show that the model is reasonable and effective,not only does it provides a theoretical basis to develop policies to strengthen the management of the port and ship management,but also verified the scientific and effective way to evaluate the application of Bayesian network loading and unloading of ships berthing method is an oil spill risk..