Container Ship Fire Risk Assessment Based on Dynamic Bayesian Network Model

Fire risk is one of the most crucial safety issues considered in transportation. Historical data indicates that accidents involving fires on ships usually cause serious consequences. Container ship fire accidents are dynamic and influenced by multiple factors, but the question of which factors are of greater concern has not been clarified. This paper analyzes the 38 fundamental factors involved in container ship fire accidents and establishes a fault tree. Considering the effect of time on the model, the fault tree is transformed into a dynamic Bayesian network model. The result demonstrates the critical risk factors such as short circuits, insufficient execution, the concealment of dangerous goods, and cargo declaration errors that require more attention. The model and results have practical applications that can improve the safety management level of container ship fire risks and help relevant personnel identify potential risk factors, ensuring the safety of personnel and property.