船舶事故时空特征及致因关联规则挖掘方法

Mining method for spatio-temporal features and causal correlation rules of ship accidents

为准确掌握洋山水域船舶事故特征,提高通航安全和减少船舶事故的发生,选取2016—2020年洋山水域船舶事故数据,采用关联规则对事故特征进行时空挖掘,构建关联规则网络图,基于Apriori算法挖掘洋山水域特定通航环境下的事故潜在风险因素。结果表明:洋山水域船舶事故多发生于南支航道延伸段、漕泾东支线航道等水域;500总吨以下船舶与事故关联程度较高;易在大潮汛期间发生自沉、碰撞等事故;春冬两季为事故潜在高发时段。依据上述结果,给出针对性的管理建议,为洋山水域海事管理机构更有效实施现场监管、保障船舶通航安全提供参考。

In order to accurately grasp the features of ship accidents in Yangshan waters,improve navigation safety,and reduce the occurrence of ship accidents,the data of ship accidents in Yangshan waters from 2016 to 2020 are selected,the correlation rules are adopted to mine the spatio-temporal features of ship accidents,the correlation rule network graphs are constructed,and the potential risk factors of accidents in the specific navigation environment of Yangshan waters are mined based on the Apriori algorithm.The results show that:ship accidents in Yangshan waters mostly occur in the South Branch Channel extension,Caojing East Branch Channel and other waters;ships below 500 gross tons have a higher correlation degree with accidents;it is easy to cause self-sinking and collision accidents during the high tide season;spring and winter are the seasons of the potential high incidence of accidents.Based on the above results,targeted management suggestions are given to provide reference for the maritime management agencies in Yangshan waters to implement on-site supervision more effectively and ensure the safety of ship navigation.