Trying to model waiting behavior of pedestrians on train station platforms via Cellular Automaton.
A more detailed explanation of the model is given in the paper and in the talk at the FEMTC 2020.
The results presented there are computed with the FEMTC2020 release from this repository.
Python 3.6 with following packages is required:
- descartes
- matplotlib
- numpy
- pandas
- scikit-fmm
- scipy
- Shapely
- VisiLibity
They can by installed by:
pip install -r requirements.txt
python waiting_ca.py <optional arguments>| option | default value | description |
|---|---|---|
| -h, --help | show this help message and exit | |
| --max_agents | 100 | max number of pedestrians |
| --init_agents | 20 | number of pedestrians at start of simulation |
| --standing_agents | 0 | number of pedestrian which will not move during simulation (only init agents) |
| --steps | 180 | number of simulation steps |
| --file | './geometries/simplified.xml' | geometry used for simulation |
| --seed | 124 | used random seed |
| --w_exit | 1 | weight of exit potential |
| --w_wall | 1 | weight of wall potential |
| --door_b, --door_c | 1, 1 | sigmoid parameter for door decay |
| --exit_b, --exit_c | 1, 1 | sigmoid parameter for exit decay |
| --wall_b, --wall_c | 1, 1 | sigmoid parameter for wall decay |
| --ped_b, --ped_c | 1, 1 | sigmoid parameter for ped decay |
| --plot | False | plot the static ff, and peds in each step |
| --output_path | 'results' | directory where the results are stored |