Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/Traffic-Alpha/LLM-Assisted-Light
This repository contains the code for the paper "LLM-Assisted Light: Leveraging Large Language Model Capabilities for Human-Mimetic Traffic Signal Control in Complex Urban Environments".
https://github.com/Traffic-Alpha/LLM-Assisted-Light
emergency-vehicles intelligent-transportation-systems large-language-models reinforcement-learning traffic-signal-control
Last synced: 19 days ago
JSON representation
This repository contains the code for the paper "LLM-Assisted Light: Leveraging Large Language Model Capabilities for Human-Mimetic Traffic Signal Control in Complex Urban Environments".
- Host: GitHub
- URL: https://github.com/Traffic-Alpha/LLM-Assisted-Light
- Owner: Traffic-Alpha
- License: apache-2.0
- Created: 2023-09-15T08:45:49.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2024-04-01T08:59:18.000Z (8 months ago)
- Last Synced: 2024-08-01T05:13:20.052Z (4 months ago)
- Topics: emergency-vehicles, intelligent-transportation-systems, large-language-models, reinforcement-learning, traffic-signal-control
- Language: Python
- Homepage:
- Size: 7.86 MB
- Stars: 33
- Watchers: 2
- Forks: 2
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
- Awesome-LLM4AD - LLM-Assisted-Light
README
# LLM-Assisted Light (LA-Light)
LLM-Assisted Light: Augmenting Traffic Signal Control with Large Language Model in Complex Urban Scenarios
[Scenario_1](https://github.com/Traffic-Alpha/LLM-Assisted-Light/assets/21176109/3075d18c-a6eb-4b5c-bdc9-f79936e13dc2)
Examples of LA-Lights Utilizing Tools to Control Traffic Signals (Normal Scenario)
[Scenario_2](https://github.com/Traffic-Alpha/LLM-Assisted-Light/assets/21176109/9062f888-314d-43f8-b668-9ad46471504c)
Examples of LA-Lights Utilizing Tools to Control Traffic Signals (Emergency Vehicle (EMV) Scenario)
## Overall Framework
The LA-Light framework introduces an innovative hybrid decision-making process for TSC that leverages the cognitive capabilities of LLMs alongside traditional traffic management methodologies. This framework includes **five methodical steps** for decision-making:
- **Step 1** outlines the task planning phase where the LLM defines its role in traffic management.
- **Step 2** involves the selection of appropriate perception and decision-making tools by the LLM.
- **Step 3** utilizes these tools interact with the traffic environment to gather data.
- **Step 4** depicts the analysis of this data by the Decision Unit to inform decision-making.
- **Step 5** illustrates the implementation of the LLM's decisions and the provision of explanatory feedback for system transparency and validation
## Evaluating LA-Light
### Training and Evaluating the RL Model
For training and evaluating the RL model, refer to [TSCRL](./TSCRL/). You can use the following command to start training:
```shell
python train_rl_agent.py
```
The [RL Result](./TSCRL/result/) directory contains the trained models and training results. Use the following command to evaluate the performance of the model:
```shell
python eval_rl_agent.py
```
### Pure LLM
To directly use LLM for inference without invoking any tools, run the following script:
```shell
python llm.py --env_name '3way' --phase_num 3 --detector_break 'E0--s'
```
### Decision Making with LLM + RL
To test LA-Light, run the following script. In this case, we will randomly generate congestion on `E1` and the sensor on the `E2--s` direction will fail.
```shell
python llm_rl.py --env_name '4way' --phase_num 4 --edge_block 'E1' --detector_break 'E2--s'
```
The effect of running the above test is shown in the following video. Each decision made by LA-Light involves multiple tool invocations and subsequent decisions based on the tool's return results, culminating in a final decision and explanation.
[LLM_for_TSC_README.webm](https://github.com/Traffic-Alpha/LLM-Assisted-Light/assets/21176109/131281d9-831d-4e08-919c-2ee8ac3fd841)
Due to the video length limit, we only captured part of the first decision-making process, including:
- Action 1: Obtaining the intersection layout, the number of lanes, and lane functions (turn left, go straight, or turn right) for each edge.
- Action 3: Obtaining the occupancy of each edge. The -E3 straight line has a higher occupancy rate, corresponding to the simulation. At this point, LA-Light can use tools to obtain real-time road network information.
- Final Decision and Explanation: Based on a series of results, LA-Light provides the final decision and explanation.
## Citation
If you find this work useful, please cite our papers:
```bibtex
@article{wang2024llm,
title={LLM-Assisted Light: Leveraging Large Language Model Capabilities for Human-Mimetic Traffic Signal Control in Complex Urban Environments},
author={Wang, Maonan and Pang, Aoyu and Kan, Yuheng and Pun, Man-On and Chen, Chung Shue and Huang, Bo},
journal={arXiv preprint arXiv:2403.08337},
year={2024}
}
```
## Acknowledgments
We would like to thank the authors and developers of the following projects, this project is built upon these great open-sourced projects.
- [TransSimHub](https://github.com/Traffic-Alpha/TransSimHub)
- [LangChain](https://github.com/hwchase17/langchain)
- [stable-baselines3](https://github.com/DLR-RM/stable-baselines3)
## Contact
- If you have any questions, please report issues on GitHub.