Citation

Wang G, Ma S, Jia N. Transp. Res. C Emerg. Technol. 2013; 35: 156-179.

Copyright

(Copyright © 2013, Elsevier Publishing)

DOI

10.1016/j.trc.2013.06.019

PMID

unavailable

Abstract

The paper develops a new framework to model the evolution of traveler route choice under risk. It embeds prospect theory in the replicator dynamics (RD), so as to allow for individual decision behavior while investigate the network state change at the population level. First, the new model is put forward based on a linear piecewise value function, in a two-route network. Some conclusions about the equilibrium states are discussed. Second, four models are formulated in a three-route network, to investigate the day-to-day route choice evolutions under the prospect assumptions. The new framework achieves average travel times no worse than the Expected Utility Theory-based RD framework, significant route switches occur at the same time. The effect of the reference point values are investigated, robust or even tiny bettering-off in average travel times still hold. Obvious route switches occur merely when the reference points fall in a certain interval. Then, taking the fourth model (M4) as an example, the sensitivities of some parameter (-combinations) are examined, the possible implications are discussed. The new framework shows flexibility in incorporating different factors in relation to individual bounded rationality and population dynamics. The possible applications may involve modeling group choice evolutions in occurrence of unexpected incidents, from the aspect of individual risk attitude, perception and decision schemes. Some limitations of the proposed framework and other research opportunities are discussed.