Intelligent autonomous drone

Abstract

With technological development, Unmanned Aerial Vehicle (UAV) is more and more widely used. However, the application of quadrotors is still relatively limited. At present, mainstream drones, such as DJI, cannot be operated indoors. The drone will automatically shut off when there is an obstacle five to six meters away to avoid accidents, making them completely useless for indoor environment. Hence, it is quite necessary to expand the application scenarios of quadrotor drones to indoors. This dissertation comprehensively reviews the state-of-the-art techniques and technologies for indoor UAVs with a different reinforcement learning model and explores a potential way to guide UAV operation indoors. Furthermore, we demonstrated the successful case for indoor UAV applications as well as further directions and challenges.