基于改进灰色聚类的洋山港大型集装箱船通航安全评价
Navigational Safety Assessment of Yangshan Port for Largescale Container Ships Based on Improved Grey Clustering
-
摘要: 为解决洋山港大型集装箱船舶通航安全评价过程中指标因素的模糊性和不确定性的问题,提升大型集装箱船舶通航安全评价的客观性。结合改进层次分析法(Analytic Hierarchy Process,AHP)与改进熵权法,采用博弈论协调二者权重,确定综合最优权重。提出一种基于改进中心点混合正弦可能度函数的灰色聚类评价模型,该模型克服了传统灰色聚类评价模型的部分缺点,且在计算可能度函数时,增加归一化步骤,解决未保证聚类系数统一的问题,符合聚类规范,保证聚类结果科学准确。以洋山港大型集装箱船为例,建立洋山港大型集装箱船评价指标体系,经计算得到的最大综合聚类系数为0.253,所属风险等级为“风险较低”,其评价结果与实际基本吻合,结果表明:利用该模型用于通航安全评价是合理可行的。Abstract: The improved Analytic Hierarchy Process (AHP) in conjunction with improved Entropy method is used to handle the ambiguity and uncertainty of the indexes in the navigational safety assessment of Yangshan Port. The optimal values of weights are decided through balancing, with the game theory, the values which would have been chosen respectively otherwise. A grey clustering evaluation model based on the sinusoidal center-point mixed possibility functions is developed, which partially overcomes the weak points of normal grey clustering models. Moreover, introduction of the normalization of weights into the calculation of the possibility function can guarantee the consistency of clustering coefficients, in turn, the accuracy of the clustering results. The assessment method is used for largescale container ships and results a maximum composite clustering coefficient 0.253, meaning low risk, a correct reflection of actual situation.