Understanding of Day-to-Day Route Choice Behavior: Experiments and Simulations

Abstract

Dissension arises on whether the day-to-day route choice behavior will cause an equilibrium distribution of traffic flow on the road network, and the travelers’ decision-making mechanism of route choice behavior is still in the exploratory stage. This paper focuses on the ‘equilibrium dissension’ and the ‘decision making mechanism’ under the condition of historical experience and traffic information by conducting human-computer interaction experiments and multi-agent simulations. The results of experiments support the conclusion that ‘no convergence to equilibrium had been found’. Moreover, the simulations with the existing mechanism that perceptions of travel time being the criterion of cognition and the logit discrete choice model being the criterion of route selection support the conclusion as well, it is also found through the simulations that dissension may due to exact treatment of treatment on the probability of discrete choice model. At the same time, comparisons between experiments and simulations found that the existing mechanism was not sufficient to reflect the fact that travelers tend to choose the shorter route more, and the shorter the more when difference of routes’ length exist, The modified mechanism proposed by this paper reflect the fact better. This study is beneficial for understanding the traveler’s route choice behavior and the causes of traffic congestion.