This repository contains the implementation of the final project for the course Computational Control taught by Prof. Saverio Bolognani at ETH Zurich(Spring 2025).
The objective of the project was to regulate speed limits in a small urban traffic network with the goal of minimizing congestion and maximizing throughput.
To address this problem, several control strategies were implemented and compared, ranging from classical approaches to modern learning-based methods:
- Proportional Control (P-Control)
- Model Predictive Control (MPC)
- Data-Driven Predictive Control (DeePC)
- Reinforcement Learning (RL)
All approaches were tested and validated using the SUMO Traffic Simulator.
Due to copyright restrictions, only my code implementation can be shared in this repository.
The datasets, simulator configuration files, and parts of the official project description are not included.