Traffic simulator for autonomous driving research
Top 61.3% on sourcepulse
LimSim is an integrated traffic and autonomous driving simulator designed for long-term, interactive, and multi-scenario simulations in complex urban environments. It targets researchers and developers in autonomous driving and intelligent transportation systems, offering a robust platform for evaluating driving behaviors and traffic dynamics.
How It Works
LimSim focuses on long-term traffic flow generation, diverse heterogeneous driving behaviors, and sophisticated vehicle interactions within various road structures. It supports co-simulation with SUMO and CARLA, ensuring synchronized vehicle states across platforms. The simulator also includes a GUI for visualization and an evaluation module for extracting key scenarios from simulation logs.
Quick Start & Requirements
git clone https://github.com/PJLab-ADG/LimSim.git
followed by pip install -r requirements.txt
in the cloned directory..net.xml
files from CitySim Repo and placing them in networkFiles/CitySim/
.Highlighted Details
Maintenance & Community
The project is actively developed, with recent updates including LimSim++ for multimodal LLM applications and LimSimLight for OpenDrive map parsing. Contact is available via GitHub issues or email (wenlicheng@pjlab.org.cn).
Licensing & Compatibility
Released under the GNU GPL v3.0 license. This license is copyleft and may impose restrictions on linking with proprietary or closed-source software.
Limitations & Caveats
The CitySim dataset maps (freewayB, Expressway_A) are not included due to copyright and require a separate data access request and download. The GPL v3.0 license may restrict commercial use or integration into closed-source projects.
6 months ago
1 day