Hybrid Transformer based Multi-agent Reinforcement Learning (HTransRL) is for drone coordination in air corridors, addressing the challenges of dynamic dimensions and types of state inputs, which cannot addressed by the traditional MARL.

Enabling UAVs to navigate corridors efficiently, aiming to minimize travel time to their destinations.

This repository contains simulations of drones movement according to ODE model by Agata Lonc, Barbara Domżał, Monika J. Piotrowska.