为保障海上运输安全,实现有效的海上风险预警,对影响国际海运通道安全最主要的非传统安全风险的时空分布模式进行探索。海上运输非传统安全风险主要包括海盗、海上恐怖主义和海上偷渡,对这3类风险发生的影响因素进行分析并结合风险特征...为保障海上运输安全,实现有效的海上风险预警,对影响国际海运通道安全最主要的非传统安全风险的时空分布模式进行探索。海上运输非传统安全风险主要包括海盗、海上恐怖主义和海上偷渡,对这3类风险发生的影响因素进行分析并结合风险特征构建特征指标体系。基于国际海事组织全球航运信息集成系统(Global Integrated Shipping Information System,GISIS),应用K原型聚类模型,并以轮廓系数法进行模型校正,对2015—2020年全球海运通道非传统安全风险进行空间划分。试验结果表明:东马来西亚沿岸、苏禄海、苏拉威西海是马六甲海峡周边海域海上袭击的高风险区;非洲沿岸港口始终是偷渡的高风险区,拉丁美洲海上偷渡发生地逐渐南移,地中海沿岸港口海上偷渡频率显著上升。研究成果表明该方法可实现对区域内不同海上非传统安全风险的细致划分。展开更多
A novel Hybrid Clustering Algonthm (HCA) that incorporates the K-means into the canonical immune programming algorithm is proposed after analyzing the advantages and disadvantages of the classical k-means clustering a...A novel Hybrid Clustering Algonthm (HCA) that incorporates the K-means into the canonical immune programming algorithm is proposed after analyzing the advantages and disadvantages of the classical k-means clustering algorithm in the paper. The theory analyse and experimental results show, the algorithm not only avoids the local optima and is robust to imtialization, but also increases the convergent speed, meanwhile evidently restrains the degenerating phenomenon during the evolutionary process.展开更多
文摘为保障海上运输安全,实现有效的海上风险预警,对影响国际海运通道安全最主要的非传统安全风险的时空分布模式进行探索。海上运输非传统安全风险主要包括海盗、海上恐怖主义和海上偷渡,对这3类风险发生的影响因素进行分析并结合风险特征构建特征指标体系。基于国际海事组织全球航运信息集成系统(Global Integrated Shipping Information System,GISIS),应用K原型聚类模型,并以轮廓系数法进行模型校正,对2015—2020年全球海运通道非传统安全风险进行空间划分。试验结果表明:东马来西亚沿岸、苏禄海、苏拉威西海是马六甲海峡周边海域海上袭击的高风险区;非洲沿岸港口始终是偷渡的高风险区,拉丁美洲海上偷渡发生地逐渐南移,地中海沿岸港口海上偷渡频率显著上升。研究成果表明该方法可实现对区域内不同海上非传统安全风险的细致划分。
基金国家自然科学基金(the National Natural Science Foundation of China under Grant No.60461001)广西自然科学基金(the Natural Sci- ence Foundation of Guangxi of China under Grant No.0542048)
文摘A novel Hybrid Clustering Algonthm (HCA) that incorporates the K-means into the canonical immune programming algorithm is proposed after analyzing the advantages and disadvantages of the classical k-means clustering algorithm in the paper. The theory analyse and experimental results show, the algorithm not only avoids the local optima and is robust to imtialization, but also increases the convergent speed, meanwhile evidently restrains the degenerating phenomenon during the evolutionary process.