Quadcopter Flight Control using Modular Spiking Neural Networks
Sushrut Thorat · 2019
The dynamics of a quadcopter are unstable and non-linear. As a result, the quadcopter's flight relies heavily on the Flight controller. Here we present a robust control scheme which can act as the flight controller for the quadcopter. We then describe a scheme to translate this scheme into a Spiking Neural Network using a modular approach to control the quadcopter flight in realistic environmental conditions (presence of noisy wind, IMU noise, and delayed signals). (The final part was left incomplete as I graduated and shifted my focus to other questions)