A Direct Demand Model of Departure Time and Modefor Intercity Passenger Trips

Travel demand is well announced as a crucial component of transportation planning. This paper aims todevelop a direct demand model, denoting a more acceptable abstraction of reality, for intercity passengers indaily work and leisure trips in Tehran province. The model utilizes combined estimation across the datasource, collected in 2011, of travelers originating from the city of Tehran and heading toward two destinationclusters: intra-province and inter-province. The paper sketches a way to predict simultaneous choice ofdeparture time and travel mode under the influence of zonal (origin, destination, and residence), individualand household socio-demographic, and trip-related variables. The time frame for analysis of departure timeis [5-19] and available modes are auto, taxi, bus, and metro. Multinomial Logit (MNL) and Nested Logit(NL) models as behavioral models are selected from discrete choice family to provide appropriate directdemand structure. Besides, the paper discusses Independent Irrelative Alternative (IIA) assumption of themodels and demonstrates choice order of NL; Travelers choose departure time prior to mode at first level andthen decide on mode at second level. Finally, travel demand elasticity and marginal effect with respect totravel time, age, and auto cost are also highlighted.

Keywords:

Direct demand , departure time , mode , multinomial logit , Nested Logit