基于海事大数据的船舶综合风险评价和船旗国检查选船标准优化研究

Research on Ship Comprehensive Risk Assessment and Optimization of Flag State Inspection and Ship Selection Standards Based on Maritime Big Data

在现行船旗国安全检查选船标准中,缺乏对船舶动态风险的评估,且船舶风险值差异较小.文中基于海事大数据,采用FP-Tree关联规则算法分析船舶风险因素的关联性,挖掘出对船舶安全影响较大的风险因素.并在现行选船标准的基础上,新增了船舶动态风险因素评价指标,构建船舶综合风险评价体系,综合动、静态风险分值和时间窗口全面评估船舶检查优先等级,达到优化船旗国检查选船方法的目的.通过实验验证,基于综合风险评价模型的船舶安全检查选船方法识别必检船的准确性提升了9.06%.

In the current standard of ship selection for flag state safety inspection, there is a lack of assessment of ship dynamic risk, and the difference of ship risk value is small. Based on maritime big data, FP-Tree association rule algorithm was used to analyze the correlation of ship risk factors, and the risk factors that have great impact on ship safety were excavated. On the basis of the current ship selection standard, an evaluation index of ship dynamic risk factors was added, and a comprehensive ship risk evaluation system was constructed. By integrating the dynamic and static risk scores and time windows, the priority level of ship inspection can be comprehensively evaluated, and the purpose of optimizing the method of ship selection for flag state inspection can be achieved. Through the experimental verification, the accuracy of ship selection method for ship safety inspection based on comprehensive risk assessment model to identify the ships that must be inspected is improved by 9.06%.