Development and calibration of route choice utility models: Factorial experimental design approach

Abstract

This paper provides a detailed overview on the state of practice in the area of data collection, modeling, calibration, and validation of route choice models. This paper presents a new approach for modeling route choice behavior using stated preference data coupled with rating (scoring) technique. The data collection was carried out over three stages; travelers' scores of route attributes, attributes' levels identification, and experimental templates for validation. A factorial experiment-design model was developed and validated through a two-stage procedure: the validation of the scoring technique, and the validation of the developed factorial experiment design model. Results indicated the dominant influence of travel time, and queuing time information on the route choice decisions. Other factors such as the speed, highway class, familiarity, and highway/pavement conditions were also found significantly important. The results include the percentages of the factors' contributions to the route utility, and the relationships derived to illustrate the effect of the factors' interactions on the travelers' route utility. © ASCE / MARCH/APRIL 2004.