摘要
为有助于对港口规划和资源分配进行科学管理,以深圳港2012—2017年共24个季度的集装箱吞吐量数据为例,对各季度数据进行加权灰色关联分析,通过各季度吞吐量的影响度排序得出其发展规律;将三次指数平滑法与马尔科夫模型相组合,并应用于港口集装箱季度吞吐量预测中。结果表明:与传统的三次指数平滑模型、灰色预测模型相比,ES-Markov模型使相对误差降至5%以下,预测精度大幅提高,因此,该模型可很好地适应港口集装箱吞吐量季节性波动的发展变化规律。未来可将该组合模型应用于港口吞吐量预测中,为港口发展决策提供理论依据。
In order to help in scientific management of port planning and resource allocation,the container throughput data of Shenzhen Port for a total of 24 quarters from 2012 to 2017 are taken to demonstrate the process of data analysis and prediction with an ES-Markov model.The weighted grey correlation analysis is performed on each quarterly data,and the rule of development is obtained by sorting the influence degree of each quarter's throughput.The cubic exponential smoothing method is combined with the Markov model to predict the quarterly container throughput of the port.The analysis shows that compared with the traditional cubic exponential smoothing model or gray prediction model,the ES-Markov model reduces the relative error to less than 5%,and the prediction accuracy is greatly improved.The model can well get with inherent seasonal fluctuation of port container throughput and provide theoretical basis for decision making in port planning.
作者
王振振
苌道方
朱宗良
罗天
WANG Zhenzhen;CHANG Daofang;ZHU Zongliang;LUO Tian(Institute of Logistics Science&Engineering,Shanghai Maritime University,Shanghai 201306,China)
出处
《中国航海》
CSCD
北大核心
2019年第4期125-130,共6页
Navigation of China
基金
国家自然科学基金(71602114)
上海市科委科研项目(16040501500
17595810300)
关键词
集装箱吞吐量
季度数据
三次指数平滑法
马尔科夫模型
预测
container throughput
quarterly data
cubic exponential smoothing method
Markov model
prediction
