MAS-Bench: a benchmarking for parameter calibration of multi-agent crowd simulation

Shusuke Shigenaka (), Shunki Takami (), Yuki Tanigaki (), Shuhei Watanabe () and Masaki Onishi ()
Additional contact information
Shusuke Shigenaka: National Institute of Advanced Industrial Science and Technology
Shunki Takami: National Institute of Advanced Industrial Science and Technology
Yuki Tanigaki: Osaka Institute of Technology
Shuhei Watanabe: Preferred Networks Inc.
Masaki Onishi: National Institute of Advanced Industrial Science and Technology

Journal of Computational Social Science, 2024, vol. 7, issue 2, No 36, 2145 pages

Abstract: Abstract Multi-agent simulation (MAS) has attracted significant attention for the prevention of pedestrian accidents and the spread of infectious diseases caused by overcrowding in recent years. In the MAS paradigm, each pedestrian is represented by a single agent. Control parameters for each agent need to be calibrated based on pedestrian traffic data to reproduce phenomena of interest accurately. Furthermore, observing all pedestrian traffic at large-scale events such as festivals and sports games is difficult. In such cases, parameter optimization is essential so that the appropriate parameters can be determined by solving an error minimization problem between the simulation results and incomplete observed pedestrian traffic data. We propose a benchmark problem, namely MAS-Bench, to discuss the performance of MAS parameter calibration methods uniformly. Numerical experiments demonstrate the baseline performance of four well-known optimization methods on six different error minimization problems that are defined on MAS-Bench. Moreover, we investigate the validity of the error function in the calibration by evaluating the correlation between the calibration and estimation scores. These scores are error functions relating to the available and unavailable observations, respectively.

Keywords: Multi-agent crowd simulation; Parameter calibration; Meta-heuristic optimization; Benchmarking (search for similar items in EconPapers)
Date: 2024
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s42001-024-00302-6 Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:jcsosc:v:7:y:2024:i:2:d:10.1007_s42001-024-00302-6

Ordering information: This journal article can be ordered from
http://www.springer. ... iences/journal/42001

DOI: 10.1007/s42001-024-00302-6

Access Statistics for this article

Journal of Computational Social Science is currently edited by Takashi Kamihigashi

More articles in Journal of Computational Social Science from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().