Analysing Parking Behaviour in Amsterdam using Agent Based Modelling (Internship Project 2019 - CWI, Amsterdam)
This repo contains all information regarding the evaluation of the agent-based parking model SimPark as a parking policy tool. This evaluation will be done with a case where on-street parking places were moved off-street in the Frans Halsbuurt. This case will be split up in three parts.
Parking situations Frans Halsbuurt:
- May 2018 - 567 parking places on-street
- Oct 2018 - 567 parking places on-street and 600 parking places off-street
- Feb 2019 - 81 parking places on-street and 600 parking places off-street
All public data that is needed for the scripts and the model is stored in this folder.
In this folder all documents created for/on meetings etc. are stored.
All papers used for this research are stored in this folder.
These three parking situations will be simulated with the agent-based parking model SimPark. All agents (drivers) will be modeled separately and will drive there own trips. These trips will be obtained through OViN data. This data contains all trips a person makes during one day of the year. To generate agents for the simulation trips are generated in scripts/schedules_generator.py and outputs an xml file for SimPark. This xml file is stored in data/generated/xml.
To generate the parking facilities and parking regulations parking facility and regulation data from the City of Amsterdam is used. This data is processed in scripts/parking_generator.py and outputs an xml file for SimPark. This xml file is stored in data/generated/xml.
The road network from the Frans Halsbuurt for SimPark is obtained via OpenStreetMap. This xml file is stored in data/generated/xml.
The vehicles used in SimPark are stored in data/generated/xml.
All files in data/generated/xml are input files for SimPark.