Abstract:In order to guide the operation, planning and management of urban rail transit, considering the congestion in rail carriage and the cost of early arrival and late arrival, a dynamic equilibrium model of travel in rail transit corridor was established. The model algorithm and the influence of value of travel time on passenger travel distribution was studied. If passengers weigh the congestion cost and delay cost, and choose the best rail schedule to minimize the total personal travel cost, the user equilibrium state of travel costs will be achieved. For the first time, the equilibrium solution of user travel time distribution in rail transit corridor was calculated iteratively based on Frank-Wolfe algorithm. The results show that, in the equilibrium state, the passengers at different destination points are continuously distributed among different shifts, and the number of passengers reaches the peak on the train with the least cost of delay. When the cost of late arrival increases to a certain extent, no passengers choose the late arrival frequency. When reducing the unit cost of early arrival time, passengers only weigh the congestion cost to choose the frequency. It not only has a certain theoretical significance for rail transit management, but also has a certain practical guiding significance for individual travel choice behavior.