Implementing TD3 to train a Neural Network to fly a Quadcopter through an FPV Gate

TL;DR

This paper applies the TD3 reinforcement learning algorithm to train a neural network for quadcopter flight control, successfully transferring the policy from simulation to real-world flight to navigate through an FPV gate.

Contribution

First application of TD3 to train a neural network for quadcopter navigation through an FPV gate with real-world deployment.

Findings

Policy successfully navigates quadcopter through gate in real environment

Demonstrates effective transfer from simulation to real-world flight

Shows potential for reinforcement learning in autonomous drone navigation

Abstract

Deep Reinforcement learning has shown to be a powerful tool for developing policies in environments where an optimal solution is unclear. In this paper, we attempt to apply Twin Delayed Deep Deterministic Policy Gradients to train a neural network to act as a velocity controller for a quadcopter. The quadcopter's objective is to quickly fly through a gate while avoiding crashing into the gate. We transfer our trained policy to the real world by deploying it on a quadcopter in a laboratory environment. Finally, we demonstrate that the trained policy is able to navigate the drone to the gate in the real world.