Cause Analysis Of Ship Traffic Accidents In NingboZhoushan Port Based On Bayesian Network

With the development of ship transportation industry becoming more high-speed,large-scale and intelligent,China’s coastal waters have become one of the busiest waters in the world.Due to the continuous increase of port traffic and port throughput,the increase of navigation density and the complication of navigation environment,the risk of ship navigation is also increasing.At the same time,the maritime department traffic safety accident statistics show that ship traffic accidents such as collision,stranding,reef hitting and sinking have not been avoided in the port.The occurrence of ship traffic accidents not only caused serious economic losses,but also caused personal casualties and pollution of the sea environment.The research on the cause analysis of ship traffic accidents is mainly through the collection of ship traffic accident data of the relevant regional maritime departments to analyze the characteristics of ship traffic accidents of different ship types or different types of ship traffic accidents,However,there are few analyses on the characteristics of ship traffic accidents in the whole port,Therefore,this paper selects NingboZhoushan Port Ship Traffic Accident as the research object,by studying the causes of ship traffic accidents in ports,it can not only reduce the risks of ships in navigation,but also prevent the occurrence of ship traffic accidents,which is of great significance for improving the safety of ship traffic and crew safety.On the basis of collating and summarizing the previous research work,this paper uses Bayesian network in the analysis of the causes of ship traffic accidents by learning the related concepts of Bayesian network.First,the basic theorem of Bayesian network and the methods of Bayesian network structure learning,parameter learning and inference learning are introduced.Then analyze the causes of ship traffic accidents,introduce the geographical characteristics of NingboZhoushan Port,select and read 80 NingboZhoushan Port Ship Traffic Accident Investigation Reports,70 of which are used for model construction and 10 for model verification and analysis The influencing factors leading to the NingboZhoushan Port ship traffic accident are subdivided into four types:human factors,ship factors,environmental factors and management factors.Then,according to the factor relationship and specific data in the ship traffic accident,combined with the Bayesian network method to construct the ship traffic accident Bayesian network model.Then use Bayesian network analysis software GeNIe to perform application verification and reasoning analysis on the established network model,and verify that the accidents with a higher probability of accident types are the same as the actual accidents,and their probability values exceed 80%,and From the verification,we can know that the types of ship traffic accidents and their probability of occurrence can be predicted by influencing factors,and the reasoning analysis can find the cause chain that leads to the ship traffic accidents in NingboZhoushan Port.Finally,according to the cause analysis results,corresponding management countermeasures and suggestions are proposed to reduce the occurrence of accidents.The research results of this paper have certain guiding significance for the crew,shipping companies and related management departments to prevent the risk of ship traffic accidents.At the same time,the Bayesian network method has certain reference significance for other types of cause analysis.