Most ABMs in practice generate tours first and then stop frequency, location, and other details are modeled for each tour. However, in reality most of the travelers first decide on the activities they want to participate in during the day. The corresponding trips and tour structure emerge from the activity participation, potential activity location, and activity sequence choices coupled with the time and space constraints imposed by inflexible activities. This study presents an approach to model daily travel by generating activities first and then forming tours. The central idea of this modeling approach is that certain inflexible or prioritized activities are scheduled first and then other activities are built (in terms of schedule and locations) around these prioritized activities. This new procedure has been incorporated in the latest version of the CT-RAMP (Coordinated Travel and Regional Activity Platform) ABM developed for the Maricopa Association of Governments (MAG).
There are four steps to this modeling approach:
1) Generation and scheduling of prioritized activities and associated tour skeletons: Decisions relating to mandatory activities, school children pickups and drop offs, joint activities, and special events participation are made at this stage.
2) Formation of day segments: The generated higher priority activities (or pegs) divide a persons’ daily schedule into different day segments classified as – between segments, part of tours, and workplace based.
3) Allocation of other activities: Other individual non-mandatory activities are allocated to one of these day segments.
4) Tour formation: Within each day segment, decisions relating to finding a most probable location for allocated activities, relative sequencing of activities, and possibility of multiple tours is evaluated.
Presentation includes model validation and sensitivity testing results to the entire synthetic population for the MAG region. Good validation results were observed with all sub-models closely replicating the targets. This approach also resulted in a relatively short run time for a large region. Sensitivity testing results with such scenarios as highway pricing shows important insights into the possible user responses with respect to trip chaining. Practical improvements resulting from this approach is an important contribution towards a further advancement of ABMs in practice.