UXsim: Traffic Simulation in Python

About

UXsim is an open-source macroscopic and mesoscopic network traffic flow simulator developed in Python. This webpage is a technical documentation of UXsim codes. Please visit GitHub repo first.

Example

Belows are simulation result where approximately 60000 vehicles pass through a 10km x 10km grid network in 2 hours. The computation time was about 30 seconds on a standard desktop PC. The first animation shows link congestion level and the second shows sampled vehicle trajectories.

Contents

• Getting started
  • Install
  • Test scenario
• Tutorial and Examples
  • Step-by-step Tutorial
  • Tutorial on Traffic Signal
  • Advanced example
  • Signal control by PyTorch (DQN)
  • Network data import from OpenStreetMap
  • Taxi / Shared mobility
  • Dynamic Traffic Assignments: DUO, DUE, DSO
  • Huge-scale simulation using Chicago-Sketch dataset
  • Fast Simulation Mode with the Internal C++ Engine
• Simulation Mechanism
  • Models
  • Key Inputs
  • Simulation Framework
  • Traffic Flow Model
  • Further reading
• Technical Reference
  • Classes (essential)
  • Classes (support)
  • Modules

Links

• Github repo of UXsim

• Author (Toru Seo, Institute of Science Tokyo)