Computational framework for reinforcement learning in traffic control

NetLimiter-like bandwidth limiting and QoS for Linux

Traffic Lights Control with Deep Learning

Adaptive real-time traffic light signal control system using Deep Multi-Agent Reinforcement Learning

We developed a system that leverages on YOLO Machine Learning Model for managing the traffic flow based on the vehicle density.

This model is very useful to detecting cars, buses, and trucks in a video.

The name says everything...

Using reinforcement learning and genetic algorithms to improve traffic flow and reduce vehicle waiting times in a single-lane two-way junction simulator by coordinating traffic signal schedules.

Documentation:

Python API for the SUMO environment of Plymouth Rd.

A dynamic traffic control system using image processing

Geo-Distributed Infrastructure Emulation using Traffic Shaping

Joint Pedestrian and Vehicle Traffic Optimization in Urban Environments using Reinforcement Learning

The PiWall project is a Raspberry Pi based, secure and standalone low-level (Layer 2 OSI) network firewall with enchanced flexibility as its rules and policies may directly be defined in Python (3.4).

NOCD is a micro NOC (Network Operations Center) that aims to help people with little to no experience in networking to create and manage Linux network.

Highway Decongestion with Variable Speed Limit Control – A Decentralized, Model-Free Approach

This research focused on developing a mainline metering policy for freeways. The mainline metering policy was controlled by a DRL agent, alongside an ALINEA algorithm to control the ramp metering policy. To model and evaluate the effectiveness of these policies, we utilized Vissim, a traffic simulation software.

Implementation of the FQ-PIE algorithm in ns-3

Implementation of Universal Multi-Agent Reinforcement Learning via Policy Decoupling with Transformers (UDPET) on Multi-Agent Traffic Control

Experiments in which Deep Reinforcement Learning agents try to choose the correct traffic light phase at an intersection to maximize the traffic efficiency. (Deep Q-Learning and Independent Deep Q-Networks)