Autonomous Intelligent Control for UAV with Self Docking, Charging and Learning Capabilities
The use of drones as autonomous systems offers several difficulties, primarily because of them being such highly unstable dynamic system. My research aims at solving some of these issues and also enabling the ability for a drone to self learn to fly by means of trial and error with Reinforcement Learning. This also entails the ability to self charge and hence also be aware of its position information, meaning having the ability to self localize. This research shows the development of such a drone and a localization system which provides self charging ability to a drone to perform sampling for Reinforcement Learning methods.