Abstract: A state-preference survey on drivers' route selection is conducted by designing the simulation scenario with real-time traffic condition in Vissim. A route selection forecasting model is constructed with binary logit model. The impact of habit and traffic condition on the drivers' route decision is estimated using this model. Results indicate that both the habit and traffic condition impact the drivers' route selection, however, the influence of habit is greater than the traffic condition. Besides, the effects of both factors vary with the changing traffic condition. A real-world forecasting experiment proves that the proposed model can be applied to predict the selection probability between habit route and alternative route according to traffic condition with satisfactory prediction accuracy. The results can be applied to real-time prediction and recommendation of travel route in vehicle navigation software.

Key words: engineering of traffic and transportation system, travel route selection, stated preference(SP) survey, binary Logit model, habit, traffic condition