RoadTrafficSimulation

Introduction

This project aims to simulate road traffic using a macroscopic approach. The traffic flow is modeled using the Lighthill-Whitham-Richards (LWR) model. The simulation is done in 2D and the road network is generated using a web interface. The results are then displayed in a video file, or in multiple graphs showing the evolution of the density over time on each road.

How to run this project

Prerequisites

• Clone this repository and enter the python environment. The list of dependencies is in the requirements.txt file. You will also need to install ffmpeg to generate the video file.
• To change the parameters used by the numerical solver, edit the config.json file in the res/ folder.

Generate the road network

• Open the HTML file index.html in the app/ folder in a browser. This will open an interface to generate the road network and the traffic flow. The instructions are written on the interface webpage.
• Once the network is generated, click on the Export button to get a file called roads.json. This file contains the road network data used by the simulation. Copy this file in the res/ folder by replacing the existing file (the name of the file is important and should be roads.json).

Run the simulation

• Finally, run the main.py file. This will run the simulation and generate a video file in the out/ folder. This video file contains the simulation of the traffic flow on the road network.
• You can also choose to visualize the evolution of density over time on each road, you can also change the value of outputType from "video" to "density" in the res/config.json file (advise: do not use this option if you have a lot of roads, otherwhise it might cause lag).

References

Main sources used for this project:

• Lighthill-Whitham-Richards traffic flow model (lectures)
• Traffic flow: the Lighthill-Whitham-Richards model (python implementation)
• 2D-LWR in large-scale network with space dependent fundamental diagram