摘要
为科学合理预测中欧班列需求量,提高班列运营效率,针对单一预测模型的局限性,提出基于多元线性回归预测、多变量灰色预测和BP神经网络预测的Shapley值组合预测模型,并采用灰色关联分析法选取中欧班列需求量的影响因素指标,以2015-2021年中欧班列开行数据为例验证模型的有效性,最后将ARIMA模型预测的影响因素值输入到Shapley值组合预测模型预测2022-2026年中欧班列需求量,判断中欧班列未来5年发展趋势。结果表明,Shapley值组合预测模型比单一模型更具有精准度优势,应用价值更高,未来5年中欧班列需求量呈现先增加后降低再增加的趋势,为政府和铁路运输管理部门提供相关决策依据,有利于中欧班列健康高效发展。
In order to forecast the demand of China-Europe railway trains scientifically and reasonably and improve the operation efficiency of railway trains,a combined prediction model of Shapley values based on multiple linear regression prediction,multivariate grey prediction and BP neural network prediction is proposed to overcome the limitations of a single prediction model.Grey correlation analysis method is used to select the factors influencing the demand of China-Europe freight trains.The opening data of China-Europe trains from 2015 to 2021 is taken as an example to verify the validity of the model.The influence factors predicted by ARIMA model are input into Shapley combined prediction model to predict the demand of China-Europe freight trains in 2022-2026,and to judge the development trend of China-Europe freight trains in the next 5 years.The results show that the combined prediction model of Shapley values has more accuracy advantages and higher application value than the single model.In the next five years,the demand of China-Europe railway trains will increase first,then decrease and then increase,which provides relevant decisionmaking basis for the government and railway transportation management departments,and is conducive to the healthy and efficient development of China-Europe railway trains.
作者
李艳丽
曹永新
李利军
郭湛
LI Yanli;CAO Yongxin;LI Lijun;GUO Zhan(School of Management,Shijiazhuang Tiedao University,Shijiazhuang 050043,China;School of Economics and Law,Shijiazhuang Tiedao University,Shijiazhuang 050043,China;China Railway Group Safety Research Center,Beijing 100081,China)
出处
《综合运输》
2024年第3期145-152,共8页
China Transportation Review
基金
国铁集团安全研究项目:铁路运行环境安全综合评估体系及典型环境安全风险评估技术方法研究(FYF2023SJ012)。
关键词
中欧班列需求量
组合预测
多变量灰色预测
BP神经网络
多元线性回归
Demand for China Europe freight train
Combined forecasting
Multiple linear regression
Multivariate grey prediction
BP neural network
