A ship collision is one of the most serious maritime accidents. It usually results from a series of factors that are related in complicated manners. Studies indicate that more than 90 percent of maritime accidents are related to human factors. This study started with the analysis of human elements that lead to ship collisions and carried out an investigation of ship collisions based on the theory of Bayesian Networks. Through analyzing sample data of ship collision cases, a Bayesian Network reflecting ship collisions was established. Computing with GeNIe software, the data and network were analyzed, and the most possible causation chain leading to ship collisions was analyzed. With Bayesian Networks, the most possible factors leading to ship collisions can be estimated, which may have significance in the prevention of ship collisions and the research of navigation safety.
展开▼
