Autonomous Landing of an Aerial Robot on a Moving Platform

12/09/2021

Unmanned Aerial Vehicles (UAVs) have had an increase of interest due to their versatility and variety of applications, however their limited battery life requires constant manual recharging. Heterogeneous multi-robot systems provide a solution to make these systems fully autonomous, by allowing an Unmanned Ground Vehicle (UGV) to serve as a recharging station for the aerial vehicle. The cooperation between aerial and ground robots allows them to solve high complexity tasks and overcome their individual limitations. This work proposes a vision-based approach that allows an aerial robot to detect a mobile charging station, perform a trajectory planning and autonomously land on the vehicle. The vision-based detection algorithm makes use of fiducial markers, a type of unique identifiable artificial marker, more specifically the ArUcO markers. Afterwards, the landing is implemented with the help of a state machine, that overviews the high-level behaviour of the UAV, and a cascaded controller, with three independent controllers for the Attitude and Altitude commands, which are then sent to the autopilot for inner-loop control. This implementation was tested both in the Simulator Gazebo 7 and on the real drone. On both systems several tests were performed, such as testing the limitations of the hardware, controller tuning experiments, tests for consistency, and a variety of flight tests. The range of test run show that the robot successfully detected and landed on a charging pad.

Authors in the community:

• 

  I

  Isabel Maria Pereira Castelo

  ist187023

Supervisors of this institution:

• 

  M

  Meysam Basiri

  ist174728

• 

  P

  Pedro Manuel Urbano de Almeida Lima

  ist12116

electrical-engineering-electronic-engineering-information-engineering - Electrical engineering, electronic engineering, information engineering

eng - English

Instituto Superior Técnico