Simulating travel to points of interest for demographically rich synthetic populations

Keywords: synthetic population, population allocation
Session Type: Virtual Paper Abstract
Day: Friday
Session Start / End Time: 2/25/2022 02:00 PM (Eastern Time (US & Canada)) - 2/25/2022 03:20 PM (Eastern Time (US & Canada))
Room: Virtual 6

Abstract

Synthetic populations approximate the individual-level makeup of social areas (i.e., neighborhoods, communities) and are widely used to model human mobility and activity spaces for urban/regional planning purposes. The UrbanPop spatial microsimulation framework produced by Oak Ridge National Laboratory creates synthetic populations to model human activities by merging detailed demographic profiles from census microdata (the American Community Survey’s Public Use Microdata Sample [PUMS]) with individual travel behavior, including work and school commutes. This is accomplished through simulating origin-destination (OD) flows between census block groups. Toward more precise models of human mobility/activity spaces, we explore extending the UrbanPop methodology to account for OD flows from residential block groups to specific points of interest (POI). Examining schools within the Knoxville, TN metropolitan statistical area, we develop an integer programming (IP) model of OD travel that realistically accounts for both student proximity to schools and school capacity based on total enrollment sizes by grade (childcare, preschool/K-12, post-secondary). This IP model is an adaptation of the classic Transportation Problem whereby supply is allocated to demand locations, students to schools respectively in our case, with the objective of minimizing total travel distance. Euclidean distance between block group centroids and schools is current used in this project with the future goal of finer detail through network analysis produced from (1) distance along roads; and (2) POIs. Beyond student travel behavior, these methods will be further adapted within UrbanPop to enhance understanding of commute behavior among workers and travel to leisure/social/civic activities for non-obligate (non-student/worker) populations.