针对已有PSC选船模型采用层次分析法、模糊综合评价方法和BP神经网络等方法存在的诸如不能反映各风险因素之间相关性、专家主观因素带来偏差,样本需求量过大和收敛速度慢等问题,以2009/16/EC巴黎备忘录目标船选船机制NIR为研究对象,提出一种新的基于支持向量机理论的PSC选船模型,并选取巴黎备忘录“THETIS”检船数据库的部分船舶信息进行实证分析.结果表明,使用基于支持向量机理论的PSC选船模型测试的船舶样本结果与其公布的船舶实际情况一致,使用该算法可十分有效地在样本少且检查资源有限时对船舶进行快速分类,对PSC选船的实际操作具有实用价值.%In view of the problems of deviations caused by the inconsiderateness of correlation between risk factors and experts' subjective factors, need for oversize samples and slow convergence etc. with the methods such as Hierarchy analysis, fussy comprehensive assessment and BP neural network, adopted by existing PSC ship selecting model, following the Paris MOU NIR(New Inspection Regime) of latest PSC rules of 2009/16/EC, a novel rapid risk evaluation model based on SVM (Support Vector Machine) method is proposed. It has been validated by its application in the"THETIS" inspection database of Paris-MOU. The results show that effective rapid ship classifications can be carried out by the algorithm. The sample test results show a good coincidence with published actual inspection results. The algorithm is perfectly suitable to do a rapid ship classification in the condition of limited inspection resource and has practical value for PSC targeting.
展开▼
