Latest revision as of 08:11, 2 February 2021

Abstract

The simulation of large-scale scenarios requires high-performance simulations. MATSim, an agent-based transport simulation, is increasingly reaching limits. The traditional approach is to scale the scenarios, i.e. simulating only 10% of a population instead of 100%. This paper suggests MacroSim, a macroscopic mobility simulation module for MATSim, to overcome the current per- formance limits within MATSim. It uses volume-delay functions to estimate travel times for links based on experienced usages of these links. This allows to decouple agents and thus allows a parallelization of the mobility simulation within MATSim per design. A preliminary implementation of MacroSim showed promising results (7 to 50 times faster than the current mobility simulation depending on the scenario size). Given its limitations - most important no back propagation of traffic congestion - MacroSim is suggested as a complementary mobility simulation to the current implementation for cases where scenario size and simulation performance are more important than precise traffic dynamics. ISSN:1877-0509

Original document

The different versions of the original document can be found in:

• http://dx.doi.org/10.3929/ethz-b-000124001

• http://dx.doi.org/10.3929/ethz-b-000165874

• http://hdl.handle.net/20.500.11850/124001

• http://hdl.handle.net/20.500.11850/165874 under the license http://creativecommons.org/licenses/by-nc-nd/4.0/

• https://doi.org/10.1016/j.procs.2017.05.406 under the license cc-by-nc-nd

• https://api.elsevier.com/content/article/PII:S1877050917310827?httpAccept=text/xml,

https://api.elsevier.com/content/article/PII:S1877050917310827?httpAccept=text/plain,

http://dx.doi.org/10.1016/j.procs.2017.05.406 under the license https://www.elsevier.com/tdm/userlicense/1.0/

• https://www.sciencedirect.com/science/article/pii/S1877050917310827,

https://dblp.uni-trier.de/db/conf/ant/ant2017.html#BoschC17,

https://www.research-collection.ethz.ch/handle/20.500.11850/165874,

http://hdl.handle.net/20.500.11850/165874,

https://doi.org/10.1016/j.procs.2017.05.406,

https://academic.microsoft.com/#/detail/2625726666

DOIS: 10.1016/j.procs.2017.05.406 10.3929/ethz-b-000165874 10.3929/ethz-b-000124001