An open-source framework for the robust calibration of large-scale traffic simulation models

Abstract

The dynamics of urban transportation can be understood with activity-based models, which rely on synthetic travel demand data to get a comprehensive understanding of urban mobility. These data are usually derived from small population samples and surveys, which may be expensive and do not adequately cover the spatial trajectories of the users. In this paper, we explore the use of a time-dependent origin-destination (OD) matrix derived from mobile phone data for the attribution of locations in a synthetic population for the city of Lyon, France. OD matrix data can also mitigate uncertainties or outdated information in travel surveys regarding flows by time of day and between zones. The resulting population enrichment is measured in terms of fit to the input mobility data.