UAV Demonstrator

This project demonstrates the use of both traditional (PID) and intelligent (RL) control methods for the Crazyflie nano-quadcopter.
We use:

• MATLAB for PID-based control and for the user interface (UI)
• Python (gym-pybullet-drones) for RL-based control The MATLAB app allows users to visualise and compare PID and RL behaviours on a variety of trajectories. You don't need Matlab to run the RL simulation. The RL part is based of gym-pybullet-drones and the Waypoints problem is implemented after PyFlyt's Waypoints.

Requirements

You need to have conda and Matlab installed to run this demonstartor. RL simulations could run without Matlab.

Installation

git clone https://github.com/norahatem/rl-crazyflie
cd rl-crazyflie

conda create -n drones python=3.10
conda activate drones

pip3 install -e .