SUMO Live Data Loop

This python package allows the repeated execution of Eclipse SUMO in the context of regularly updated traffic data such as speeds from floating car data or counts and speeds from induction loops.

Currently it requires a Postgres database as a backend which provides regular updates. The configuration is done with a schema python file which encapsulates the database access (gives details on the queries, see src/default_schema.py for a template) and a config file in ini format (see default.cfg).

There are two major loops implemented. One for detector data correction and one for the simulation run.

Setup

Database inspection

A good idea is to use DBeaver which is cross platform and can handle other databases than oracle as well: https://dbeaver.io/download/.

Python

• Create a virtualenv (python -m venv sumo_ldl_env) and activate it (source sumo_ldl_env/bin/activate or sumo_ldl_env\Scripts\activate).
• Install the requirements python -m pip install -r requirements.txt.

python3 -m venv sumo_ldl_env
. sumo_ldl_env/bin/activate
python3 -m pip install -U pip
git clone https://github.com/DLR-TS/SUMOLiveDataLoop
python3 -m pip install -r SUMOLiveDataLoop/requirements.txt
python3 -m pip install eclipse-sumo
cd SUMOLiveDataLoop
python3 -m pip install -e .

Differences to the original dsp

• checkData.py has been removed (primary focus: PSM interaction)
• scenarioClient.py and scenarioTrigger.py are not integrarted (focus: running different scenarios, may be re-added later)

Good to know

• Traffic demand (Routes) for different days fo the weekday (Monday, Tuesday...) can be set under "routesPrefix" in the configurtion file, for example, ./%(region)s/infra/pkw/pkw_Mon, ./%(region)s/infra/pkw/pkw_Tue and so on.
• Currently, only one calibrator.rou.xml is used even when traffic demand on different weekdays is considered.
• Only routes, extracted from the given traffic demand, on the edges where detector data are available are considered by the calibrator.