We study an estimation method for generating dynamic activity model based on previous trips. The dynamic model predicts the daily activity pattern of the multi-day period, which takes into account the dynamic demand: the daily preference change of an activity and the travel expenditure budget. Compared with static model, the dynamic model can show the time rhythm that can be found in the daily activity pattern. We establish an ordered logit model for dynamic activity schedule formation decision-making, and show how to calculate the travel probability in the next few days. Finally, numerical experiments are carried out to verify the effectiveness of the proposed method. Our method can provide a reasonable way to predict activity-based travel demand without a large amount of data. At the same time, it can be used for passenger flow prediction and data expansion.

CICTP