Abstract: With the rapid growth of urban population, the travel demand of public transportation of citizens also rises, which easily causes problems such as difficulty in getting on at some stations during rush hours. It is of great significance to relieve the pressure of urban public transportation by analyzing the big data of IC card to deduce the passenger flow at the boarding station and predict the distribution of passenger flow in advance. For more targeted research on dynamic passenger flow distribution of stations, a kind of clustering analysis method is adopted based on IC card data, through the establishment of data model of bus lines and stations of rush hour traffic, to analyze the measured results, and finally measures are put forward to improve traffic operation, including dispatching station express, shuttle bus and other measures, ensuring urban residents ease and comfort in daily bus ride and at the same time providing guidance for the city traffic department on bus lines design and comprehensive transport planning.

Key words: IC card data, cluster analysis, passenger flow forecast, boarding station planning