Network work trip assignment for agent-based simulation accounting for travel time

Abstract

With advances in computational power and larger sample data sources, destination choices can be determined from the disaggregated data itself using logit models rather than from aggregated gravity models. Unfortunately, larger sample data are required as spatial resolution increases in such models, which is often limited due to privacy restrictions or survey costs. This makes higher spatial resolution work trip OD assignment inaccurate due to over-aggregation from limited sample data. Alternatively, high resolution aggregated OD employment data are available through Census sources, making detailed OD assignment possible with the gravity model. The challenge now becomes how to assign destinations to disaggregated individuals using spatially aggregated sources. This paper explores a method of assigning disaggregated destinations using aggregated sources, constrained by individual travel time using OD travel time matrices. This process was performed on a synthetic population of 4.1-million individuals in the Boston, Massachusetts area distributed across 2,727 traffic analysis zones. Travel times by mode from an origin zone were used to create a zonal radius around the origin, greatly reducing the number of possible destinations for individuals. From this subset, the possibilities were then further stratified using individual criteria of age and industry. This added travel time constraint reduces the possibility of individuals being assigned to impossible destinations where random assignment is used.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 2 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.