中文核心期刊
CSCD来源期刊
中国科技核心期刊
RCCSE中国核心学术期刊

Journal of Chongqing Jiaotong University(Natural Science) ›› 2020, Vol. 39 ›› Issue (12): 31-36.DOI: 10.3969/j.issn.1674-0696.2020.12.06

• Traffic & Transportation Engineering • Previous Articles     Next Articles

Ship Selection Model Based on Principal Component Analysis and Extreme Learning Machine

ZHENG Zhongyi, MU Jiaqi   

  1. (Navigation College, Dalian Maritime University, Dalian 116026, Liaoning, China)
  • Received:2019-03-29 Revised:2019-10-09 Online:2020-12-18 Published:2020-12-18
  • Supported by:
     

基于主成分分析-极限学习机的选船模型

郑中义,牟家奇   

  1. (大连海事大学 航海学院,辽宁 大连 116026)
  • 作者简介:郑中义(1964—),男,河北安国人,教授,博士,主要从事海上交通工程方面的研究。E-mail: dlzzyi@sina.com 通信作者:牟家奇(1994—),男,山东烟台人,硕士研究生,主要从事海上交通工程方面的研究。E-mail: mujq0064@dlmu.edu.cn
  • 基金资助:
     

Abstract: In order to solve the problem that the new inspection regime of the Tokyo MOU is difficult to identify the inspection order of the inspected ships, the method of principal component analysis and extreme learning machine was used to study the selection of the inspected ships. Based on the port state control data of the Tokyo MOU, the targeting model was analyzed in case study and the predicted results were compared with the actual data. The research results show that the proposed ship selection model is based on principal component analysis to reduce the dimension and complexity of the index samples and the extreme learning machine is used to fit and predict the detention and defects of the inspected ships, which can reduce the inspection volume by half with the accuracy rate of 90%. The validity of the proposed model is verified, which can provide decision support for the port state control for ship selection.

 

Key words: traffic and transportation engineering, inspection ship, ship selection model, principal component analysis, extreme learning machine

摘要: 针对东京备忘录新检查机制难以划分可检船检查优先顺序的问题,利用主成分分析和极限学习机相结合的方法进行了可检船的选船研究。基于东京备忘录的港口国监督数据,对建立的模型进行实例分析,将预测结果与实际数据进行对比。研究结果表明:建立的选船模型基于主成分分析,降低了指标样本的维度和复杂度,并利用极限学习机对可检船的滞留和缺陷情况进行拟合和预测,能够在90%的准确率下减少一半的检查量,验证了模型的有效性,可为港口国监督选船提供决策支持。

关键词: 交通运输工程, 可检船, 选船模型, 主成分分析, 极限学习机

CLC Number: