基于贝叶斯网络的近海渔船碰撞事故致因分析
张曼铃,丁天明,丁超君,艾万政
浙江海洋大学船舶与海运学院,浙江 舟山 316022
摘要:为减小近海渔船碰撞事故发生的概率,保障渔船航行安全,从渔船视角出发,根据渔船碰撞事故样本数据,采用鱼刺图法分析近海渔船碰撞事故的原因,并将事故原因作为贝叶斯网络的节点进行推理和验证,通过后验概率推理和敏感性分析探究事故致因链。结果表明:人为因素对近海渔船碰撞事故的影响程度最大;其次是管理因素,其中未按规定显示号灯号型、未采取有效的避让行动、瞭望疏忽及配员不符合要求的敏感度较高,对近海渔船碰撞事故的影响程度较大。
关键词:渔船;碰撞事故;鱼刺图;贝叶斯网络;致因分析
Cause analysis of collision accidents of offshore fishing vessels based on Bayesian networks
ZHANG Manling, DING Tianming, DING Chaojun, AI Wanzheng
School of Naval Architecture and Maritime, Zhejiang Ocean University, Zhoushan 316022, China
Abstract: To reduce the probability of collisions between fishing vessels in the vicinity of the coast and ensure the safety of fishing vessel navigation, an analysis of the causes of these collisions is conducted using the fishbone diagram method from the perspective of fishing vessels. The causes of the accidents are used as nodes in a Bayesian network for inference and validation. Posterior probability inference and sensitivity analysis are employed to explore the causal chain of accidents. The results indicate that human factors have the greatest impact on collisions involving fishing vessels near the coast, followed by management factors. Among the human factors, high sensitivity is observed for failure to display proper lights and signals, failure to take effective evasive actions, negligence in lookout, and inadequate crew qualifications. These factors have a significant impact on collisions involving fishing vessels in the vicinity of the coast.
Keywords: fishing boat; collision accident; fishbone diagram; Bayesian network; causation analysis