Day-to-day route choice in networks with different sets for choice: experimental results

This paper studies the impact of route choice set of networks on the day-to-day traffic dynamics through conducting a group of experiments, in which four traffic networks with different sets for choice are used. It is observed that a deterministic user equilibrium (DUE) state is difficult to sustain and the probability of reaching DUE is larger in a network with smaller route choice set. Three day-to-day models are calibrated by using the experimental data. Results reveal that (1) there are linear relationships between traffic volume/flow changes obtained from the experiments and models; (2) when the same model is applied to different networks, the calibrated adjustment parameters are statistically different; and (3) the same model reveals different performances for different networks. Besides, it is shown the participants are more likely to change route choices in a network with larger route choice set and tend to switch away after experiencing a longer travel time.