Abstract
It is discussed to change the traffic condition in the road network according to time zone as a transport policy for the decentralization of the road traffic demand at the peak time. The travel behavior change of driver corresponding to such an hourly transport policy includes the departure-time choice with the travel mode choice and the route choice. In the present study, the traffic simulation model is constructed with departure-time choice for the evaluation of the transport policy. The multi agent system is composed to describe individual driver's heterogeneity of the value of the late risk and time as an important factor on departure time choice. Moreover, the learning process of the perceived travel time is built into the departure time choice model. Therefore, a basic finding for development of the traffic simulation can be summarized for the evaluation of hourly transport policy.
