An Safety Assessment Model of Ship Collision Based on Bayesian Network

Accidents often occur in the waters of inland river bridge areas, which have a great impact on bridge safety, life safety and environmental pollution. This paper build a Bayesian network (BN) to connect the various types of accidents, from may affect the safety of navigation environment, the person’s problem, ships, four aspects of the natural environment were analyzed, and the division of ship’s security level and early warning indicators, through the analysis of the opinions of the experts in this field, optimization of Bayesian networks is conditional probability tables. Using Genie to construct the Bayesian network structure diagram, the Bayesian network can be re-inferred by changing the state of nodes. The research results show that the model has a more accurate evaluation result for the navigation risk of ships in the bridge area, and can be used to remind the ships and drivers of the navigation risk in the bridge area in different situations.