| 摘要: |
| 船舶航行安全是实现内河航运高质量发展的重要前提，Apriori算法可挖掘船舶事故强关联规则，对于事故致因分析与预防具有重要指导价值。针对Apriori算法参数依赖先验知识的不足，本文引入了一种参数自适应Apriori算法，借助多项式曲线拟合方法确定支持度和置信度阈值；结合长江干线航道2015年01月-2025年04月船舶事故数据样本，分析了船舶事故基本特征，研究了致因属性和基础属性之间的潜在映射关系，得到了强关联规则。研究结果表明：长江航道船舶事故以碰撞事故、一般事故为主，且与船舶类型、事故致因、船舶总吨等变量密切相关；事故多伴随人员伤亡，小吨位船舶事故更易造成人员伤亡；事故多发生在第二季度、0400-0800时段，且基础属性和致因属性之间存在一定的关联关系。本文研究结果可为内河海事安全监管、船舶事故预防和应急管理决策提供技术支撑。 |
| 关键词: 长江航道 船舶事故 数据挖掘 参数自适应 关联规则 Apriori算法 |
| DOI： |
| 分类号:U698 |
| 基金项目:国家自然科学基金项目（面上项目，重点项目，重大项目） |
|
| Analysis of Ship Accident Characteristics in the Yangtze River Waterway Based on a Parameter-Adaptive Apriori Algorithm |
|
JIANG Zhonglian1, YANG Di1, LI Jia2, FENG Honglin3, WANG Yu4
|
|
1.State Key Laboratory of Maritime Technology and Safety,Wuhan University of Technology;2.Yangtze River Communications Administration,Ministry of Transport;3.Transport Planning and Research Institute,Ministry of Transport;4.Institute of Logistics Science and Engineering,Shanghai Maritime University
|
| Abstract: |
| Maritime navigation safety is an important prerequisite for the high-quality development of inland waterway transportation. The Apriori algorithm can effectively mine association rules from ship accident investigation reports, which provide valuable insights for accident analysis and prevention. To reduce the dependence of the classic Apriori algorithm on prior parameter settings, a parameter-adaptive Apriori method based on polynomial curve fitting was introduced to automatically determine the support and confidence thresholds. By leveraging ship accident data from the Yangtze River main waterway spanning from January 2015 to April 2025, the basic characteristics of inland ship accidents are explored, and the potential mapping relationships between causal attributes and fundamental attributes are investigated to extract strong association rules. The results demonstrate that collisions and general accidents are the most common types of ship accidents in the Yangtze River waterway, and are closely related to variables such as vessel type, causal factors, and gross tonnage. Accidents are often accompanied by casualties, and small-tonnage vessels are more likely to result in injuries or fatalities. In addition, accidents occur more frequently in the second quarter and during the time period of 04:00-08:00. Significant associations were also identified between basic attributes and causal attributes. The present findings provide technical guidance for inland maritime safety management, and support decision-making of accident prevention and emergency management. |
| Key words: Yangtze River Waterway ship accidents data mining parameter adaptation association rules Apriori algorithm |
