A freight forecasting approach for river-sea direct transport without direct historical data

Abstract

This paper proposes an effective freight volume forecasting approach for river-sea direct transport without direct historical data. First, we collect easy-to-obtain shipping data, including channel ship flow, channel set running speed, free-flow running time, channel length, channel capacity, etc.. Then, based on these data, we use the OD matrix estimation method to obtain the freight volume matrix. Next, we forecast the future total freight volume by using the grey prediction model. Finally, we adopt the Logit model to estimate the sharing rate of river-sea direct transport, and obtain the freight volume of river-sea direct transport. This paper describes the effectiveness of this approach by using a specific route from Ningbo-Zhoushan port to Ma anshan port to forecast the river-sea direct transport freight volume.