Freight distribution analysis and modelling of inland waterway transport using big data

Abstract

The Yangtze River economic belt (YREB) serves as the main east–west axis of China to promote economic development and environmental protection along the Yangtze River (YR). The factors that affect the freight distribution of major types of cargo transported through the YR were analysed using automatic identification system data and ship visa data. First, a set of freight impedance functions was developed for different types of links of the waterway network, considering a number of factors (e.g. cargo type, delays at ship locks, water levels and flows at different waterway segments and upstream and downstream shipping speeds). Distance-based (DB) and time-based (TB) impedance matrices of different types of cargo were computed. A gravity model (GM) and an intervening opportunity model were then used to simulate the distributions of different types of cargo based on the computed impedance matrices. A trip length distribution method was applied to validate the estimated distribution models. The results showed that the GM with a power term outperformed the other models and that the TB models were superior to the DB models for the prediction of freight distributions over large geographies like the YREB. This work offers an in-depth understanding of the freight characteristics of inland waterways and therefore should be helpful for relevant authorities in formulating port and inland waterway plans and policies.