MICROSIMULATION OF ACTIVITY PARTICIPATION, TOUR COMPLEXITY, AND MODE CHOICE WITHIN AN ACTIVITY-BASED TRAVEL DEMAND MODEL SYSTEM

Abstract

Over the past few decades, trip-based travel demand approaches have been replaced by activity-based microsimulation travel demand techniques, which are able to capture the latent demand for activity participation, interdependency among trips, and household interactions. Activity-based models consider trips as a derived demand which arise from activity engagement behavior. This research aims to depict the daily activity-travel behavior of travelers as a result of choice decision making processes through the development of the Scheduler for Activities, Locations, and Travel (SALT) microsimulation travel demand model. The SALT model is comprised of five main components: population synthesis, time-use activity pattern recognition, tour mode choice, activity destination choice, and activity/trip scheduling. A series of advanced econometric micro-behavioral modules are developed to model behavioral mechanisms of different population groups in the region. An under-recognized issue in most of the econometric activity-based models is that they treat all out-of-home travelers, whether workers or non-workers, as undifferentiated groups, decreasing the ability to predict activity-travel decisions. To this end, an advanced disaggregated modeling framework is developed that can derive separate utility functions for both in-home and out-of-home activities for travelers with heterogeneous daily-activity patterns, along with simulation of correlation matrices. Additionally, a cluster-based technique is developed to model trip chaining, tour complexity, and tour mode choice of worker and non-worker clusters. These models capture associations between socio-demographics characteristics, trip attributes, and land use patterns in order to predict travel tour incidence and type, and mode choice. For empirical analysis of activity-travel behavior this study employs data from the large Halifax Space Time Activity Research (STAR) household time-use and travel survey, which consists of GPS-verified data for 2,778 person-days. This study also contributes by designing and conducting the first Canadian university-based travel-diary survey (EnACT), to better understand activity-travel patterns and trip making frequencies of university commuters. In addition, a synthetic pseudo-panel modeling framework is developed to explore the longitudinal activity-travel behavior of urbanities. In summary, the disaggregated modeling framework presented in this study is useful for deeper understanding of individuals’ activity-travel decisions, and may be operationalized to examine sensitive policy issues such as transportation control measures and congestion-pricing.