Abstract: With the rapid development of urbanization, more and more people accommodate in urban areas and the number of private cars increases sharply, which make urban road congestion more and more serious. Traffic congestion concerns the public, researchers and authorities, for which a variety of measures from punishing to incentive perspectives have been conceived and carried out in practice. However the results of governance are quite unsatisfactory. An evolutionary congestion game model is built in this paper, lifting the veil on the process of different governance measures functioning on urban congestion problem and trying to give an explanation for observed results. The whole community is divided into two groups, one showing preference towards taking public transport and the other preferring private cars. The quantified government congestion measures are numerically examined which trigger trade-off matrix changes, and dynamic trends of two groups' ESS strategies in different situations. It is found that both the incentive and punitive traffic congestion policies can alleviate the congestion problem, but the punitive measures are more effective than the incentive measures in the rapidity and persistence.

Key words: traffic congestion, collective rationality, evolutionary game theory, congestion governance