STUDY ON TRAVELER'S DYNAMIC LEARNING AND ITS APPLICATION TO TRAFFIC ASSIGNMENT IN URT

摘要

This paper proposed a three-step procedure method to simulate travelers' behavior: updating travel time perception, triggering alteration of choice and choosing an alternative. The models are also applied to Fuxinmen transfer station to assign the passenger flow to different departure time periods. It is concluded that travelers will get familiar with a new network or some new traffic policies after 5-6 days of traveling. This conclusion will help policy makers and operators on traffic policies and optimized operation plan.