Day-to-day Dynamical Evolution Of Network Flows With Considering Departure Time Choice

Urban traffic congestion is a special kind of network flows pattern in a certain time-space context,and the network flows are the aggregation result of multiple trip choice behavior of the travelers.Based on the analysis of traveler's departure time and route choice process,a network flows evolution model was proposed.The evolution laws of network flows were searched from the day-to-day dynamical evolution of network flows.The paper analyzed the chaotic characteristics of a simple network,and further explored the flows evolution laws of the complex network,the traveler's bounded rationality was considered to study the evolution laws of network flows.More details are listed as follows:Firstly,in order to investigate how the travelers' departure time choice behavior influenced day-to-day dynamic evolution of traffic flow in transportation network,a simple two-link network with two intervals was considered as the object.Considering the influence of historical information and real-time information,a dynamic evolution model of network flows with two intervals was presented.Based on the numerical experiment results,the evolution laws of network flows were analyzed.The chaotic characteristics of unstable state in system evolution were studied by the chaotic judgment method based on Lyapunov exponent.Then,the paper further considered the complex network with multi-period and multi-path.Based on the day-to-day traffic demand and within-day traffic flow update rule,a within-day and day-to-day double dynamic network flows evolution was modeled.According to the numerical experiments results,the day-to-day evolution laws of network flows were analyzed.At the same time,the effect of different combinations of traveler's characteristics on the stability of network flows evolution was studied.Finally,it was assumed that travelers adjust departure time and route choices according to updating historical information,with considering traveler s bounded rationality.The behavior decision-making theory was introduced into the dynamic travel choice behavior research,and the day-to-day dynamic model of network flows was developed based on the cumulative prospect theory.Based on the numerical experiment results,the evolution laws of network flows were analyzed from the day-to-day dynamic evolution process of network flows,and the critical point of the evolution of network flows to stable and unstable state was obtained.From the perspective of the overall efficiency of the network system,the difference between the bounded rational traffic assignment and the perfect rational one was discussed.This paper studied the evolution of network flows under the influence of traveler's departure time and route choice behavior.From macroscopic view,it can observe whether the evolution of network flows reaches a stable state and get the characteristics of the unstable state.From microscopic view,it can also analyze the relationship between traveler's choice behavior and the evolution of network flows so as to better understand the traveler's travel choice behavior mechanism and the evolution laws of network flows.The research results can provide a theoretical basis for traffic managers to solve the problem of urban traffic congestion.