主管:中国科学院
主办:中国优选法统筹法与经济数学研究会
   中国科学院科技战略咨询研究院

中国管理科学 ›› 2025, Vol. 33 ›› Issue (10): 24-35.doi: 10.16381/j.cnki.issn1003-207x.2023.0902

• • 上一篇    下一篇

基于DKELM选择性深度集成的集装箱吞吐量预测模型研究

寇宇轩1, 李思涵2, 刘敦虎3, 李若依4, 黄静5, 肖进2,6()   

  1. 1.澳门科技大学商学院,澳门特别行政区 999078
    2.四川大学商学院,四川 成都 610064
    3.成都信息工程大学管理学院,四川 成都 610225
    4.芝加哥大学哈里斯公共政策学院,美国 芝加哥 60637
    5.四川大学公共管理学院,四川 成都 610065
    6.四川大学管理科学/运筹学研究所,四川 成都 610064
  • 收稿日期:2023-06-01 修回日期:2024-03-26 出版日期:2025-10-25 发布日期:2025-10-24
  • 通讯作者: 肖进 E-mail:xjxiaojin@126.com
  • 基金资助:
    国家自然科学基金面上项目(72171160);国家自然科学基金面上项目(71974139);国家自然科学基金面上项目(71974020);四川省“天府万人计划项目”(0082204151153);四川大学国家领军人才培育基金项目(sksyl2021-03)

A DKELM-based Selective Deep-ensemble Model for Container Throughput Forecasting

Yuxuan Kou1, Sihan Li2, Dunhu Liu3, Ruoyi Li4, Jing Huang5, Jin Xiao2,6()   

  1. 1.School of Business,Macao University of Science and Technology,Macao 999078,China
    2.Business School,Sichuan University,Chengdu 610064,China
    3.School of Management,Chengdu University of Information Technology,Chengdu 610225,China
    4.Harris School of Public Policy,The University of Chicago,Chicago 60637,U. S.
    5.School of Public Administration,Sichuan University,Chengdu 610065,China
    6.Management Science and Operations Research Institute,Sichuan University,Chengdu 610064,China
  • Received:2023-06-01 Revised:2024-03-26 Online:2025-10-25 Published:2025-10-24
  • Contact: Jin Xiao E-mail:xjxiaojin@126.com

摘要:

随着国际贸易的蓬勃发展,港口集装箱化水平不断提高,集装箱吞吐量成为衡量一个港口在全球商业管理和工业发展方面的关键指标,构建有效的时间序列预测模型来准确预测港口集装箱吞吐量至关重要。本研究提出了基于深度核极限学习机(DKELM)选择性深度集成的集装箱吞吐量预测模型(DSDE)。首先,采用优化变分模态分解法,将原始时间序列分解成若干个本征模函数分量;然后,考虑到分解后的本征模函数大都是高度非线性的,构建不同核函数的DKELM作为基准模型进行预测;进一步引入新的数据分组处理技术(eMIA-GMDH)在每个本征模函数上进行选择性深度集成;最后,整合全部预测结果得到最终的预测结果。为了验证本文构建的预测模型效果,引入4种评价指标在6个港口的集装箱吞吐量月度数据集上进行实证研究。实验结果显示,与3种基于单一深度学习的混合模型以及基于加权平均集成的混合模型相比,DSDE模型的预测性能更优。

关键词: 集装箱吞吐量, 混合预测模型, 深度神经网络, 选择性深度集成

Abstract:

With the rapid development of international trade and the continuous improvement of port containerization level, its throughput has become a key indicator to measure a port's global commercial management and industrial development. Therefore, it is very important to construct an effective time series forecasting model to accurately predict port container throughput. A DKELM-based selective deep-ensemble model is proposed for container throughput forecasting (DSDE). Firstly, the original time series is decomposed by the optimized variational modal decomposition method, and several intrinsic mode functions are obtained. Then, considering that most of the decomposed intrinsic functions are highly nonlinear, DKELM of different kernel functions is constructed as the benchmark model for prediction. Further, the new enhanced multilayered iteration algorithm of the group method of data handling (eMIA-GMDH) is introduced to carry out the selective deep ensemble on each intrinsic mode function. Finally, the final forecasting result is obtained by integrating all the forecasting results. To verify the effect of the forecasting model constructed in this paper, four evaluation indexes are introduced to carry out empirical analysis on the monthly container throughput data set of six ports. Through experiments, it can be concluded that the performance of DSDE is better than that of the three hybrid models based on a single model and the hybrid model based on the weighted average ensemble

Key words: container throughput, hybrid prediction model, deep neural network, selective deep ensemble

中图分类号: