Pedestrian Dynamic - Contractile Particle Method

What to Install

• python3 -m pip install numpy
• python3 -m pip install matplotlib
• Download and install OVITO from: https://www.ovito.org/

Versions

python 3.8

Configuration

Everything is configured by modifying config.json. Available configuration keys are:

• dynamic_file: dynamic file filepath
• exit_file: exit file filepath por all exit times
• animation_file: animation file filepath
• N: number of pedestrians in the room
• L: room side, room will be of L x L
• d: exit width
• rmin: minimum pedestrian radius
• rmax: maximum pedestrian radius
• vdmax: maximum desired velocity
• tau: time taken for a particle to reach its maximum radius
• beta: velocity formula exponent
• use_seed: if true use fixed seed, if false use nanoseconds
• seed: fixed seed value
• dt_print_mult: integer timestep multiplier for simulation prints to file. dt_print = dt * dt_print_mult
• dt_anim_mult: integer timestep multiplier for animation prints to file. dt_anim = dt * dt_print_mult * dt_anim_mult
• plot: determines whether to plot or not single analysis, must be true or false

Particle generator

To generate initial particle positions by creating dynamic_file. Generates N particles with random positions, rmin and 0 velocity. If N cannot be reached in a number of iterations, resets and tries again. Run python3 generator.py [dynamic_filename] [N]. If provided, dynamic_filename and N parameters overwrite config params

Simulation

To generate executable and run pedestrian dynamic simulation

1. Run ./prepare.sh in root to generate executables (only required once).
2. Run ./target/tp5-simu-1.0/pedestrian-dynamic.sh -Dn=N -Dd=d -Ddynamic=dynamic.txt -Dexit=exit.txt. Parameters from config.json can be overwritten by using n, d, dynamic and exit properties.

Output will be appended to dynamic_file, showing time and particle position and velocity for each timestep dt_print. Particle exit times will be printed to exit_file with a precision of dt.

Animation Tool

Generates simulation_file using information from dynamic_file. Run python3 animator.py [dynamic_file] [N] [d], using the parameters from config.json. If provided, params overwrite dynamic_file and d from config.

To view the animation, you must open simulation_file with Ovito: ./bin/ovito simulation_file.

Particle color shows particle radius, where particles with rmin are red (many contacts) and particles with rmax are black (fewer contacts).

Column Mapping

Configure the file column mapping as follows:

• Column 1 - Radius
• Column 2 - Position - X
• Column 3 - Position - Y
• Column 4 - Particle Identifier
• Column 5 - Color - R
• Column 6 - Color - G
• Column 7 - Color - B

Analysis Tools

analyzerSame.py

Generate plots and metrics given simulation files as input. Run python3 analyzerSame.py [file_1 file_2 ...], using parameters from config.json.

If one or more filenames are provided, analysis will be performed individually and then condensed for multiple simulations. This can be used to provide multiple repetitions for a given set of parameters. If plot is false, then no graphs are plotted.

Plots shown are:

• n = f(t), n: number of particles that left

If multiple files are provided, some more plots are shown:

• Discharge = f(n), taking dn=10
• n = f(t) for each simulation file
• t = f(n) for each simulation file
• n = f(t avg) with horizontal error bars for each value
• t avg = f(n) with vertical error bars for each value

same.sh

This script can be used to run pedestrian-dynamic simulation multiple times with the same params, given N, d and a number of repetitions. ./same.sh N d rep

The script runs rep simulations with the parameters N and d. Then, it runs analyzerSame.py with the output datafiles as parameters.

analyzerSame.py

Generate plots and metrics given multiple simulation files as input. Run python3 analyzerMult.py [file_1 file_2 ...]. Filenames must be in the format name-N-....

Aanalysis will be performed individually and then condensed for multiple simulations. This can be used to provide multiple repetitions for a multiple sets of parameters. If plot is false, then no graphs are plotted.

Plots shown for each (N,d) pair are:

• Discharge = f(n), taking dn=10
• n = f(t avg) with horizontal error bars for each value
• t avg = f(n) with vertical error bars for each value

Plots shown as a comparison between sets of parameters are

• Discharge = f(n) for each (N,d) pair
• Discharge = f(t) for each (N,d) pair
• Average discharge in stationary sector for each d value with vertical error bars
• Average discharge in stationary sector adjusted via Beverloo's Law (with corresponding error plot)
• Average discharge in stationary sector adjusted via Beverloo's Law with the vertical error bars and a continous adjustment (with corresponding error plot)

multiple.sh

This script can be used to run pedestrian-dynamic simulation multiple times with different params, given a number of repetitions. ./multiple.sh rep

The script runs rep simulations for each of the following (N;d) params: (200;1.2), (260;1.8), (320;2.4), (380;3.0). Then, it runs analyzerMult.py with the output datafiles as parameters.