基于模糊C均值聚类的多属性船舶综合质量分级方法  

Multi-attribute ship classification method based on fuzzy C mean clustering algorithm

作　　者：鲍君忠[1] 王利东[2] 王西召 

机构地区：[1]大连海事大学航海学院,江苏无锡214000 [2]大连海事大学数学系,辽宁大连116026 [3]中船重工第七〇二研究所,江苏无锡214000

出　　处：《大连海事大学学报》2016年第3期44-48,共5页Journal of Dalian Maritime University

基　　金：国家自然科学基金资助项目(61203283);辽宁省自然科学基金资助项目(2014025004)

摘　　要：为进一步提高我国海事管理机构的船舶管理水平与管理资源的使用效能,通过确定船舶评价指标,聚合5名海事领域专家对各指标重要性的判断矩阵,形成船舶综合质量评价指标权重.运用模糊C均值聚类(FCM)对船舶进行分级,构建新的船舶综合质量评价体系,并利用MATLAB进行算例验证.结果表明,FCM可降低人为设定船舶综合质量分级阈值的主观性问题,提高了海事管理机构对船舶综合质量管理的科学性.In order to enhance ship management level and application performance of management resource for China MSA,the fuzzy C mean clustering algorithm（ FCM） was proposed to rank ships and to construct a new evaluation system by establishing assessment indexes of ships,and judgment matrix on importance of each index from five maritime experts formed the comprehensive quality evaluation index weights of the ship. Some cases were verified by using MATLAB. Results show that FCM can reduce subjectivity in threshold setting and make it more scientific in ship management.

关 键 词：船舶 综合质量管理 评价方法 模糊C均值聚类(FCM) 

分 类 号：U698[交通运输工程—港口、海岸及近海工程]