2D Quadrotor Control Through Reinforcement Learning

Table of Contents Project Overview In this project, I created a deep deterministic policy gradient (DDPG) reinforcement learning agent in Tensorflow to learn the dynamics of a two-dimensional drone. This project is an example of model-free RL since the agent has no knowledge of the drone’s kinematics or dynamics. I created my own OpenAI Gym … Read more

Quadrotor – Dynamical Modelling and Control

Table of Contents Project Overview In this project, the non-linear dynamics of a quad-rotor drone are linearized about an equilibrium point (stable hover configuration). Feedback control is then applied to achieve the desired reference trajectories. The goals of this project are to: Model and simulate the dynamics of a 6 DOF quad-rotor drone. Apply LQG … Read more

Multi-UAV Drone Swarm

Project Overview This project was done during my summer internship at UNIQ Technologies. I created a drone swarm algorithm for multiple drone navigation. Each drone had it’s own flight path and both were equipped with collision and obstacle avoidance. Due to the confidential nature of this work, the source code is not provided. Please feel … Read more