https://github.com/harshramani00/trafficsignaloptimizationforsmartcitiesusingdeep-reinforcementlearning
A Deep Reinforcement Learning-based system for optimizing traffic signal control in smart cities. This project simulates a real-time traffic network and applies Q-Learning and Deep Q-Networks (DQN) to dynamically manage traffic flow, reduce congestion, and prioritize emergency vehicles using SUMO.
https://github.com/harshramani00/trafficsignaloptimizationforsmartcitiesusingdeep-reinforcementlearning
deep-learning deep-q-learning dqn python qlearning reinforcement-learning simulation sumo
Last synced: 2 months ago
JSON representation
A Deep Reinforcement Learning-based system for optimizing traffic signal control in smart cities. This project simulates a real-time traffic network and applies Q-Learning and Deep Q-Networks (DQN) to dynamically manage traffic flow, reduce congestion, and prioritize emergency vehicles using SUMO.
- Host: GitHub
- URL: https://github.com/harshramani00/trafficsignaloptimizationforsmartcitiesusingdeep-reinforcementlearning
- Owner: harshramani00
- Created: 2025-02-02T21:25:24.000Z (8 months ago)
- Default Branch: main
- Last Pushed: 2025-04-11T00:40:26.000Z (6 months ago)
- Last Synced: 2025-06-20T04:38:35.107Z (4 months ago)
- Topics: deep-learning, deep-q-learning, dqn, python, qlearning, reinforcement-learning, simulation, sumo
- Language: Jupyter Notebook
- Homepage:
- Size: 11.1 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# 🚗 Traffic Signal Optimization using Deep Q-Learning
---
## 📅 Project Overview
As urban populations grow, traffic congestion presents serious challenges to efficiency and quality of life. Traditional traffic light systems struggle with real-time adaptation. This project proposes a scalable, intelligent alternative by using reinforcement learning to control signal phases based on real-time traffic conditions.
---
## 🪀 Key Features
- Intelligent traffic light control using Deep Q-Learning (DQL)
- SUMO-based simulation environment
- Emergency vehicle prioritization
- Real-time traffic flow optimization
- Comparison with traditional Q-Learning
---
## 🔄 Algorithms Used
- **Q-Learning**: For simple environments with reduced state spaces
- **Deep Q-Networks (DQN)**: For complex, dynamic traffic scenarios
---
## 💡 Technologies & Tools
- Python
- SUMO (Simulation of Urban Mobility)
- TraCI (Traffic Control Interface)
- TensorFlow / Keras
- NumPy, Matplotlib
---
## 📈 Results Summary
| Model | Reward | Vehicles Waiting | Avg Speed (m/s) | Avg Wait Time (s) |
|--------------------|----------|------------------|------------------|--------------------|
| Baseline Simulation | -15255.36 | 405 | 4.2 | 3493.33 |
| Q-Learning | -2327.62 | 5 | 4.6 | 79.22 |
| Deep Q-Learning | 2774.90 | 0 | 6.5 | 15.66 |
---
## 🌐 How to Run
```bash
# Step 1: Install dependencies
pip install -r requirements.txt
# Step 2: Run environment simulation
python RL_env.py
# or use Jupyter Notebooks for step-by-step walkthrough
jupyter notebook dql-simulation-1.ipynb
```
---
## 🚀 Future Enhancements
- Train on larger datasets like Kinetics or real-world city data
- Optimize the model for real-time, on-device deployment
- Deploy as a web application for live simulations and demos
---
## 🤝 Contributors
- Aryan Patil ([GitHub](https://github.com/aryanator))
- Harsh Anilkumar Ramani
- Saiteja Kalam
- Chandra Mourya
---