Research On The Optimal Toll Rate Of Distance-based Urban Road Dynamic Congestion Pricing

Congestion pricing,as an important instrument on transportation demand management(TDM),is of great significance in ameliorating urban traffic congestion,and has been highly focused by transport authorities.Since the first implementation of congestion pricing in Singapore in 1975,there has been a number of countries/cities implemented the congestion charging policy,which has achieved remarkable results in alleviating urban traffic congestion.One of the key problems in congestion pricing is to determine the optimal toll rate.All the implemented congestion charging schemes currently adopt a flat toll-charge method to determine the toll rate.However,this flat toll-charge method may induce the equity as well as the efficiency problems,due to the overlook of travel distances.Besides,the traffic flows are determined based on deterministic and static equilibrium state,without the consideration of stochastic day-to-day dynamics and the disequilibrium evolution process of the traffic network.In order to solve these problems,this thesis studies the distance-based congestion pricing problem considering stochastic day-to-day dynamics.Firstly,the thesis made a more reasonable division of the congestion pricing schemes,analyzed the shortcomings of the existing congestion pricing schemes,and discussed the advantages and feasibilities of the next generation of distance-based pricing scheme.The purpose is to provide a reference for the transport authorities to make a more reasonable congestion charging policy.Secondly,a piecewise linear toll function is formulated to approximate the nonlinear distance-based toll in terms of the travel distance in the cordon area,thus a general function is formulated.When the interval of a linear function is appropriate,the piecewise linear function can approximate the nonlinear distance-based toll with good performance.Thirdly,after the analysis of existing theories of the dynamic traffic assignment,two types of stochastic day-to-day dynamics models including the convergence-oriented model and the behavior-oriented model are proposed.These two types of models also represent two trends of day-to-day dynamics models in the literature.Fourth,a stochastic day-to-day dynamics based minimax regret model is proposed considering the long-term disequilibrium state of the traffic network,and the proposed model considers the network performance on each day during the whole planning period,thus is a robust optimization model.Finally,the computing complexity due to the implicit flow map function is analyzed,and a two-stage ABC algorithm is developed for the highly complex problem considering the implicit day-to-day flow map function with a distance toll.The result of the case study validates the proposed model as well as the algorithm.The purpose of this thesis is to provide the theoretical basis and decision-making evidence for the transport authorities to make a more reasonable congestion pricing policy.