Mixed Traffic Network Traffic Equilibrium Assignment Model And Algorithm

Traffic assignment is a crucial part of urban transportation planning. Under the assumption that every traveler can know all information of the road well in the fixed demand network, Wardrop had put forward user equilibrium system in 1952. However, traffic behavior is always stochastic , so stochastic user equilibrium system which had unified the conception of stochastic assignment and Wardrop equilibrium had been put forward by Daganzon and Sheffic in 1977. In our country, mixed traffic has played a main role in urban traffic. So SUE model that contains only one traffic mode can not describe the actual traffic state well.The thesis studies the mixed traffic assignment problems by mathematical programming method.First, this thesis introduces the traffic assignment models and theirs algorithm briefly. Second, it gives the stochastic user equilibrium flow split model based on mixed traffic network equating to the equilibrium conditions, and proves the uniqueness of mathematical programming model's solution. At the same time, because of the interactions between the traveling modes, the equilibrium flow split model with symmetric traffic interactions based on mixed traffic network is formulated. The equivalence to the equilibrium conditions and the uniqueness of the resolution to the model are proved. Moreover, genetic algorithm is tried to work out the two models. Finally, two examples are provided, and the effectiveness of this algorithm is illustrated.