基于模糊聚类的海上交通事故分析

Maritime Traffic Accident Analysis by Using Fuzzy Clustering

大量研究表明,80%的海上交通事故与人的因素有关。究其原因,驾驶员对碰撞危险度很难有定量的认识。模糊聚类是一种定性与定量相结合的数学分析方法。首先结合实例利用仿真技术对船舶的碰撞过程进行动态还原仿真,然后通过船舶碰撞事故分析系统进行分析得出事故相关船舶连续的最近会遇距离(DCPA)、最近会遇时间(TCPA)动态曲线,最后应用模糊聚类的分析方法建立碰撞危险度的分类模型,为海上安全主管机关调查和处理船舶碰撞事故提供一种新的途径,同时为船长、驾驶员采取正确的避碰措施提供参考。

Aiming at quantitative analysis o{ ship collision risk, ship collision simulation is carried out and dynamic curves of the ship＇s DCPA, TCPA is drawn by the ship collision accident analysis system. Classification model Risk- degree of Collision based on fuzzy clustering analysis method is then established. The result provides a new way for the maritime security supervisor organization to investigate and deal with ship coUision accidents, and it also pro- vides valuable reference for watchkeepers and captains.