|
|
|
Title: High-Resolution Destination Choice in Agent-Based Models
Accession Number: 01371518
Record Type: Component
Abstract: This paper describes the modeling of destination choice for discretionary activities in a multi-agent transport simulation, using MATSim as an example. MATSim is based on utility maximization. Randomness was included implicitly and in an uncontrolled way through the stochasticity of the simulation process, and sometimes through a logit choice model. Unobserved heterogeneity is now added directly to the utility function through a random error term. Importantly, those random error terms are quenched, i.e., they will be the same for repeated executions of the choice model. Real-world simulation experiments for Zurich show that this substantially improves results. High-resolution destination choice for large-scale microsimulations raises several technical issues; pragmatic engineering solutions have been developed or applied to cope with them. These solutions are described in technical detail to assist in the further development of similar microsimulations.
Supplemental Notes: This paper was sponsored by TRB committee ADB40 Transportation Demand Forecasting
Monograph Title: Monograph Accession #: 01362476
Report/Paper Numbers: 12-1988
Language: English
Corporate Authors: Transportation Research Board 500 Fifth Street, NW Authors: Horni, AndreasNagel, KaiAxhausen, Kay WPagination: 21p
Publication Date: 2012
Conference:
Transportation Research Board 91st Annual Meeting
Location:
Washington DC, United States Media Type: Digital/other
Features: Figures; References
TRT Terms: Identifier Terms: Uncontrolled Terms: Geographic Terms: Subject Areas: Planning and Forecasting; Transportation (General); I72: Traffic and Transport Planning
Source Data: Transportation Research Board Annual Meeting 2012 Paper #12-1988
Files: TRIS, TRB
Created Date: Feb 8 2012 5:06PM
|
