NicklasXYZ / sumotde

SUMO Travel Demand Estimation: Estimation of Dynamic Origin-Destination Matrices v0.1.0

A codebase containing different tools for estimating origin-destination matrices using SUMO.

Overview

Currently the main implementation resides in the /fileserver directory:

• The /fileserver/main.py contains code that can be executed to:
  • Generate a test traffic network
  • Generate synthetic scenario data using a generated test traffic network
  • Run all or a specific estimation algorithm of choice given given synthetic scenario data as input
• The /fileserver/static directory contains experimental data and final algorithm outputs
• The /fileserver/images directory contains different plots generated with python scripts /fileserver/plot1.py and /fileserver/plot2.py

Assuming SUMO (>=1.10.0) and all Python dependencies have been intalled, the /fileserver/images/main.py can be executed:

# Run database migrations
python manage.py makemigrations && python manage migrate

# Run estimation script
python main.py

Note

To be able to run the estimation procedure the duaIteratev3.py script contained in this current directory will have to be copied to the same directory containing the original duaIterate.py installed together with SUMO. The duaIterate.py script usually resides in directory /usr/share/sumo/tools/assign/.

Running the following command should do it:

cp duaIteratev3.py /usr/share/sumo/tools/assign/.

Implementation Notes

To be consistent acorss methods and the data used by them data is handled internally:

• Vehicle link counts are derived from vehicle exist times recorded in vehicle route files
• Vehicle link counts are aggregate meaning that counts are aggregated across lanes if multiple lanes exist on a link