集装箱港口群的吞吐量预测及运价博弈决策改进模型  

A Model for Predicting Throughput of Container Port Group and Improving Decision-Making in Pricing Game

作　　者：耿粲 钟铭 GENG Can;ZHONG Ming(College of Transportation Engineering,Dalian Maritime University,Dalian 116026,China)

机构地区：[1]大连海事大学交通运输工程学院,辽宁大连116026

出　　处：《中国航海》2022年第2期100-108,共9页Navigation of China

基　　金：国家社科基金后期资助项目“‘通实力’研究”(19FZZB013);国家社科基金重大项目(13&ZD170)。

摘　　要：基于内部恶性竞争和新冠肺炎疫情导致运营收入下降的实际情况,如何提高集装箱港口群的总运营收入,成为集装箱港口群决策者需解决的重要问题。为了提高港口群总运营收入,压制港口群内部恶性价格竞争,提出一种基于改进GM(1,1)模型、logit模型和伯特兰德-纳什均衡模型构成的二阶预测博弈模型。首先利用三点平滑法改进GM(1,1)模型,对未来集装箱吞吐量进行预测;然后利用logit模型根据集装箱港口效用模拟集装箱货运人对集装箱港口的选择行为;最后利用伯特兰德-纳什均衡模型,分别求出集装箱港口群的非合作博弈和合作博弈收入。结果显示,对于集装箱吞吐量的预测准确度比改进前提高0.68%,且合作博弈的总运营收入和非合作博弈总运营收入比原始收入分别提高52.3%和35.5%。研究表明,二阶预测博弈模型能对集装箱吞吐量做出更准确地预测,且能有效地提高集装箱港口群的运营收入,为港口群的管理者制定合理的运价奠定理论基础。To inhibit internal vicious price competition within a container port group and improve the income of the group as a whole,a second order prediction game model is developed based on GM(1,1)model,Logit model and Nash-Bertrand Model.The GM(1,1)model is improved with three point smoothing and used to predict future container throughput.Logit model is used to simulate the behavior of shipper in port selection.Nash-Bertrand Model is used to calculate the income of the port group for cooperative game and non-cooperative game respectively.The calculation shows that with presented method,the prediction accuracy improved 0.68%and the overall group income increased 52.3%and 35.5%for cooperative game and non-cooperative game respectively.

关 键 词：集装箱港口群 博弈模型 GM(1 1)模型 伯特兰德-纳什均衡 

分 类 号：F552[经济管理—产业经济]