Day-to-day Dynamics And Congestion Pricing In Transportation Network With Travelers' Behavioral Characteristics

With rapidly increasing of car ownership,the traffic congestion becomes a severe problem leading to social economic loss and environment pollution in the city.To solve this issue,several traffic demand management policies are proposed,such as peak shifts,vehicle restrictions and congestion pricing.Particularly,congestion pricing,from the economic perspective,could effectively induce users to alter travel modes and travel routes.Study on individuals' daily traveling,which is the main purpose of urban traffic,could enhance implementation effect of traffic demand management policies.Day-to-day traveling is a process with dynamic learning and repeated decision-making which would influence by individuals' behavioral characteristics.Integrating day-to-day traveling and congestion pricing,this dissertation mainly carries out the following work:1)Study on individuals' day-to-day route choice behavior considering route preferences.In the dissertation,a dynamic-updating mechanism for route preference based on Dogit model is introduced into the traditional day-to-day route choice model.Effects of individual route preference on day-to-day route flow evolution track are analyzed and three different learning strategies on perceived route travel time are compared—exponential smoothing leaning and peak-end rule learning.In addition,we bring in dynamic pricing based on a trial-and-error implementation method and discuss the evolution of route preference and traffic flows and present various route choices of users with different value of time.2)Study on individuals' day-to-day route choice considering risk attitude and value of time.Under the framework of cumulative prospect theory,this paper introduced a trial-and-error dynamic congestion pricing scheme,and separated the time and toll into two dimensions.It further assumed that the individuals' daily time reference point were related to their risk attitude and perceived travel time,and the daily toll reference points were dependent on their value of time and saved travel time.All of reference points dynamically varied with the day-to-day perceived travel time.With this setting,a dynamic user equilibrium model was proposed with consideration of two reference points,i.e.,time reference point and toll reference point.Various route choice behavior of users with multiple value of time and the weight of pricing prospect value is analyzed.