The traditional method of loading and unloading time prediction of freight transport,which directly utilizes the standard time specified in the Detailed Instructions Governing Train Operation at Station,cannot properly characterize the time change under the impacts of complex factors,and achieves low prediction accuracy.This paper utilizes the data mining method to gather the relevant data on the loading and unloading time of freight transport from the railway integrated dispatching information system.It also utilizes the boosted decision tree model Xgboost to predict the loading and unloading time of freight transport.Compared with the reference model,the proposed model can achieve substantial improvement in prediction accuracy,and provide more effective support for traffic flow prediction and automatic drawing of train operation charts.
关键词
重载铁路/货运作业/装卸作业时间/决策树/Xgboost
Key words
heavy haul railway/freight transport/loading and unloading time/decision tree/Xgboost
维普
万方数据