| Nowadays,China’s economy has been deeply integrated into the progress of economic globalization,and its dependence on foreign trade is increasing.As the main mode of transportation in foreign trade,the safety and stability of transportation channel plays a decisive role in China’s current and future economic development.At the present stage,pirate attacks are constantly reported in the global waters,which not only brings great threat to the safety of ship operation,but also brings great instability to the sea transportation channel,and affects the development of China’s economy.In this case,predicting the probability of ship piracy attack risk and studying the main causes and related factors of ship piracy attack risk are conducive to ensuring the safety of ship operation,maintaining the safety and stability of maritime transportation channels,and further promoting the development of China’s economy.This paper mainly studies the risk of ship piracy attack through Bayesian network model.Firstly,the mechanism of the risk of ship piracy attack is clarified,and the risk of ship piracy attack is mainly composed of five risk factors: ship,environment,protection,piracy action and naval management.Secondly,the Bayesian network nodes are determined by risk factor analysis;the Bayesian network structure is preliminarily determined by using the causal relationship between nodes,and the preliminary Bayesian network structure is modified by correlation analysis based on the data of ship piracy attacks in different regions,and finally the Bayesian network structure is determined;the conditional probability table of each node is calculated by EM algorithm to complete the Bayesian network model The construction of the system.Finally,the connection tree algorithm is used to complete the reasoning of the model.This paper selects the West African waters for empirical analysis of the risk of ship piracy attack Bayesian network.The rationality of the Bayesian network model for the risk of ship piracy attack is determined;the model can be used to predict the probability of the risk of ship piracy attack is verified by an example;then the key causes and related influencing factors of the risk of ship piracy attack in West Africa are determined by using the diagnostic reasoning ability of Bayesian network,and the corresponding countermeasures and suggestions are put forward Suggestions.The research results of this paper clarify the mechanism of ship piracy attack risk,enrich the method of Bayesian network model construction,prove that using Bayesian network model can carry out the research of ship piracy attack risk,realize the prediction of ship piracy attack risk probability based on limited information,as well as the research of the main causes and related effects of ship piracy attack risk. |
