Modeling of commuters' day-to-day route choice behaviors based on reference-dependent approach

In order to investigate commuters' day-to-day route choice behaviors under travel time uncertainty, a novel model was presented based on reference-dependent approach, a reference-dependent route disutility function was introduced, and commuters' day-to-day route choice behaviors obeyed the decision process based on logit model, a day-to-day evolution model of route choice was proposed, in which the equilibrium for every day's route choice was kept. The stability of evolution model was analyzed and solved by a heuristic solution algorithm combining the method of successive averages with logit assignment technique. Analysis result shows that network flows and reference point come into steady state after several days' evolution, and traditional expected utility models underestimate total network user cost and overestimate the benefits from road capacity expansion or road pricing policies in amount of 1%.