README.md

MBSE-traffic-simulation

Simple traffic simulator.

Quickstart

git clone git@github.com:mactat/MBSE-traffic-simulation.git
cd ./MBSE-traffic-simulation/
pip3 install -r requirements.txt
cd ./src

# For example with animation
python3 example.py

# For headless
cd ../tests
python3 ../src/headless.py < "input1.txt" 

Creating simple simulation

from sim import Scheduler

highway_length =          1    # in kilometers
num_of_lanes =            4    # number of lanes
average_drivers_mood =    0.85 # how propable it is that driver will not perform any action
sim_time =                20   # in minutes
inflow =                  10   # cars per minute
speed_limit =             60   # in kilometers per hour
propotion_of_autonomous = 0.5  # propotion of autonomous cars
proportion_of_trucks =    0.1  # proportion of trucks 

scheduler = Scheduler(
                        average_drivers_mood =      average_drivers_mood ,
                        num_of_lanes =              num_of_lanes, 
                        highway_length =            highway_length, 
                        speed_limit =               speed_limit,
                        propotion_of_trucks =       propotion_of_trucks,
                        proportion of autonomous =  propotion_of_trucks,
                        step_time = 1)

results, results_dict, average_speed = scheduler.simulate(time_of_sim = sim_time, inflow = inflow)

Reusing scheduler object

scheduler.num_of_lanes = 3
scheduler.average_drivers_mood = 0.97
scheduler.reset()
results, results_dict, average_speed = scheduler.simulate(time_of_sim = sim_time, inflow = inflow)

Exporting data

Exporting only data

After simulation data can be safed in form f a json file:

scheduler.safe_results_to_file("results")

Exporting whole scheduler

Scheduler can be aslo exported and imported(both data and parameters of an objects)

scheduler.safe_to_file("sheduler.pkl")

Importing scheduler

scheduler1 = Scheduler.load_from_file("test.pkl")

Visualizing output

from animation import createAnimation

createAnimation(
    results_list = [results_dict1],            # results from simulation - multiple can provided for compering simulations
    animation_speed= 20,                       # animation speed
    reduce_data = 1,                           # how much reduce the data, usefull in large datasets
    highway_length=highway_length,             # length in kilometers
    num_of_lanes=[num_of_lanes],
    names = ["Simple test"] ,                  # if not provided default
    export_gif_path = "../static/testcase.gif" # if not provided, animation will be shown in the form of plot
    )

Note: All parameters provided to animation are arrays, so it is possible to visualize multiple simulation in one animation.

Results

As a results the simulator is evaluating number of cars which were able to pass the highway (flow) and the average speed of the cars compared to the speed limit:

Animation time: 23.90/24.0s Real time: 3.98/4.00min
Flow: 13/60
Average speed: 32.3/110 km/h.

Functions of the simulator

Basecases(overtaking, changing lanes)

Basic simulation

Changing speed of simulation

Changing number of lanes

Comparing different simulations

Simulation with autonomous cars

Results without autonomous: 35/160
Results with only autonomous: 140/160
Results with 50/50: 74/160

Animation time: 179.00/180.0s Real time: 2.98/3.00min
Results without autonomous vehicles:
Flow: 38/120 vehicles passed the highway.
Average speed: 23.3/110 km/h.

Results with only autonomous vehicles:
Flow: 147/120 vehicles passed the highway.
Average speed: 109.5/110 km/h.

Results with 50/50 autonomous vehicles:
Flow: 120/120 vehicles passed the highway.
Average speed: 81.5/110 km/h.

Simulation with autonomous cars and plots

Experiments with autonomous cars and plots