This study aims to develop a feasible and practical method for predicting medium- and short-term metro passenger volume. A regression model for was built by input days after metro opening,operation mileage, months and weekdays as independent variables and metro passenger volume as dependent variables. The results show that the model can model the regularity of month-based and weekday-based changes and presents cyclical characteristics of passenger flow in different months and weeks.Days after metro opening,operation mileage, Friday,Saturday and Sunday of weekdays are positively related to the metro passenger volume, while all the twelve months has negative effect on volume. In general,operation mileageaffected the volume most significantly, followed by days after metro opening,weekdays and months rankly. The mode performed well when it is used to predict a newly opened metro line. The findings can be a guideline for metro equipment optimization, operation organization in metro-opened cities and for passenger volume prediction in newly metro-opened cities.

CICTP