Analysis On Traffic Assignment And Congestion Strategies With Decision Inertia

This research analyzes travelers’ behavior of urban trips deeply and focuses on the characteristics of travelers’ inertial behavior in trip makings.Models with decision inertia are established to describe the network equilibrium flow patterns.The Braess paradoxes in network design problem are explored and congestion pricing strategies are represented to alleviate traffic congestion.The main contents and conclusions are listed below.Firstly,the psychological and behavioral factors contributing to bounded rationality and decision inertia are analyzed respectively under the theory of bounded rationality.Meanwhile,the relationship between bounded rationality and decision inertia is briefly discussed,and then methods of describing travelers’ bounded rationality and decision inertia in the existing user equilibrium models are compared in detail.Secondly,given that the existing user equilibrium models can not fully characterize travelers’ inertial behavior,a model of status quo-dependent user equilibrium(SDUE)is proposed.The nonlinear complementarity problem and mathematical program of the SDUE model are formulated.Afterwards,The SDUE model can incorporate some user equilibrium models in the literatures as special cases with specific settings for parameters,such as the Wardrop user equilibrium(UE),boundedly rational user equilibrium(BRUE),bi-objective user equilibrium(BUE)and time surplus maximization bi-objective user equilibrium(TSmaxBUE).Compared with the BRUE model,the SDUE model considers route choice inertia resulting from the asymmetric preference.Moreover,Compared with the BUE model and TSmaxBUE model,the SDUE model incorporates route choice inertia resulting from satisficing behavior and misperceptions,allowing travelers to choose the dominated paths.And besides,the network equilibrium flow patterns of the SDUE may not be unique and we characterize the set of the SDUE as a non-convex set.Numerical examples are conducted to illustrate the properties of the SDUE flow distributions.Thirdly,we incorporate travelers’ decision inertia into the network design problem.Network planners may hold different attitudes towards risk and uncertainty due to non-uniqueness of the SDUE,and therefore three types of paradoxes are defined:risk-averse Braess paradox,risk-prone Braess paradox and risk-neutral Braess paradox.Furthermore,in the classical Braess network,the relationship between the Braess paradox occurrence conditions and the parameters of the SDUE model is investigated.Finally,we explore the congestion pricing strategies with travelers’ decision inertia,and pay attention to risk-averse and risk-prone congestion pricing strategies.A model of risk-averse congestion pricing is proposed aiming to optimize for the worst-case scenario,which is formulated and solved as a min-max problem.In addition,a model of risk-prone congestion pricing is also established trying to optimize for the best-case scenario,which is described as a min-min problem.Risk-averse and risk-prone congestion pricing strategies are analyzed with three simple numerical examples.