An Experimental Study On Travelers' Bi-objective Day-to-day Route Choice Behavior

Urban traffic system is a typical complex system with various information.Its basis elements include: means of transportation,network and travelers.It's the travelers' decisionmaking behavior that constitutes the traffic flow on each road in the network.In daily travel life,travel time and monetary cost are two essential factors towards travel choice.Thus,when making the route choice,Traveler often face the trade-offs between travel time and monetary cost.The day-to-day traffic flow dynamic models depicts the dynamic evolution process of the traffic flow and simulates the travelers' route choice behavior much closer to real life compared to static model.As an effective research method,behavior experiment is often used to explore traveler's behavior and help to enrich the study of theoretical models.Based on above,after summarizing the research on the empirical studies of related problem and the dynamic traffic flow model with multi-objectives,this paper designed two experiments,one is to discover the travelers' trade-offs between the travel time and monetary cost,the other is to examines several day-to-day traffic flow dynamic models with bi-objectives.Firstly,using z-tree platform,this paper designed the travel time and monetary cost tradeoff behavior experiment by Stated Preference Survey(SP survey),and use the experiment data to explore the traveler's real trade-off between the travel time and monetary cost by considering both linear and non-linear relation between travel time and monetary cost,calculate the value of time and time-cost indifference curve and classify the travelers into three groups based on the experiment's results.Secondly,it designed and conducted a virtual day-to-day route choice experiment of 90 rounds with the participation of toll roads.However,the network traffic did not reach a stable equilibrium at last under the influence of both travel time and monetary cost.Finally,this paper introduces several bi-objective day-to-day traffic flow dynamic models considering both travel time and travel cost,and with the realized day-to day path flows and travel times from the experiment,it would be able to do the simulated calculation and regression analysis and compares the results of different models.And also explores the path preference and learning behavior of travelers.