An Improved Gravity Model for Passenger Flow Distribution Prediction of New Urban Rail Transit Lines

Abstract

The structural features of the rail transit network in a city are significantly affected when a new urban rail transit line is additionally opened. The traditional prediction method to estimate the passenger flow distribution based on historical passenger flow data is no longer applicable. Although the existing gravity model can be used to predict the flow distribution of passengers after a new line is opened, it cannot show the characteristics of passenger flow in a station since the average travel impedance calculation method is employed and it doesn't take the station attributes into consideration. In order to address this problem, a new gravity model integrating station attributes is proposed in this paper. The model can be used to predict the passenger flow distribution with fitted parameters, and different travel characteristics are considered in this model. Taking Chengdu Metro as an example, all the stations are clustered according to their attributes, and the prediction results are calculated with the traditional dual-constrained gravity model and the proposed gravity model, respectively. In the end, we have made a comparison, which shows the advantages of the proposed model when the station attributes are considered.